["Analyzing Hinge Flow in an Optically Clear Mechanical Heart Valve Using Particle Image Velocimetry Natalia Briseno Advisor: Dr. Bellofiore Department of Biomedical Engineering, San Jose State University BME 291 MS Final Report Briseno 1 Introduction 2 Briseno 2 Introduction Aortic stenosis is a type of heart valve disease that describes the stiffening of the aortic valve as a result of accumulating calcium deposits in the area. A stiff and unyielding aortic valve will not open and close properly, disrupting normal blood flow in the heart. The primary treatment option for aortic stenosis is valve replacement, where surgery is performed to implant either a mechanical or bioprosthetic heart valve. Over 200,000 aortic valve replacements are performed per year worldwide 1 . Typically, patients with shorter life expectancy will opt for bioprosthetic valves, since these types of valve replacements create a more natural flow of blood into the heart. However, mechanical heart valves (MHV) are the preferred option for patients who have a life expectancy of over 10 years after implantation. This preference is due to the benefits of MHVs, which include easier availability, lower cost, and higher durability 2,3 . Currently, the most popular MHV designs include the bileaflet and trileaflet valves. Bileaflet MHVs have two leaflets, whereas Trileaflet MHVs have three leaflets. An example of a bileaflet valve, the St. Jude Bileaflet valve, is shown in Figure 1. Because of their market history and extensive clinical data, bileaflet valves are more commonly implanted than trileaflet valves. Figure1: St. Jude MHV with the General Components Labeled 14 Unfortunately, MHVs have been shown to produce non-physiological flow patterns, which can cause platelet activation and thereby induce thrombosis 4,5 . Therefore, following implantation of an MHV, patients will be required to take anticoagulant drugs for the rest of their life to prevent thrombosis from occurring. This medication increases the risk of blood loss as the anticoagulative properties of the medication prevents the formation of clots on cuts and the healing of injuries. As such, researchers attempt to improve or revolutionize the design of MHVs to induce a more physiologically relevant flow through the aortic valve into the heart. Bioprosthetic valves in comparison, do not require the patient to take anticoagulant therapy, however, they are prone to structural valve degeneration. Patients will then need to undergo a valve replacement procedure if they outlive their valve implant, which poses significant risks and is highly invasive,", "of clots on cuts and the healing of injuries. As such, researchers attempt to improve or revolutionize the design of MHVs to induce a more physiologically relevant flow through the aortic valve into the heart. Bioprosthetic valves in comparison, do not require the patient to take anticoagulant therapy, however, they are prone to structural valve degeneration. Patients will then need to undergo a valve replacement procedure if they outlive their valve implant, which poses significant risks and is highly invasive, potentially leading to serious complications during recovery. In order to improve the design of bileaflet MHVs to reduce the risk of clotting, it is important to understand the hemodynamic behavior in all regions of the valve. Any regions where non-physiological flow is observed could contribute to clot formation when implanted into a patient 6 . Currently, particle image velocimetry (PIV) is used to image flow in mechanical heart Briseno 3 valves 7 . However, it is impossible to visualize the flow field in certain key locations, such as the hinge region and closing flow, as the metal valve housing blocks the view of the camera. These regions still hold important information about non-physiological flow that could contribute to clot formation. Therefore, it is important to develop optically clear valves to characterize flow in these areas. By characterizing the flow in these areas, we can understand and validate the safety of a specific valve design to be implanted in a patient. The goal of this project is to design, prototype, and test an optically clear SJM valve and test section. In the future, the methodology developed can be used to conduct research on both existing heart valve designs as well as improved designs. The specific aims of the project are: designing and prototyping the heart valve and test section assembly, validating the model by comparing flow properties around the model with flow properties of an SJM valve, and conducting PIV analysis around the hinge region of the valve. The data collected from PIV will be used to characterize flow and justify clotting around the hinge region. In order to improve the performance of mechanical heart valves, it is essential to understand the mechanisms by which clots are forming, and how the valve design is contributing to the clots. Currently, there are limited methods to observe the flow in specific regions of the valve, such as the hinge region, to prove the", "around the hinge region of the valve. The data collected from PIV will be used to characterize flow and justify clotting around the hinge region. In order to improve the performance of mechanical heart valves, it is essential to understand the mechanisms by which clots are forming, and how the valve design is contributing to the clots. Currently, there are limited methods to observe the flow in specific regions of the valve, such as the hinge region, to prove the hypothesis that the hinge design is contributing to non-physiological flow, and therefore, clotting. A similar hypothesis is also seen with the closing flow of mechanical heart valves, in which the flow may become turbulent or disruptive due to the rigid nature of the components of the MHV. In both cases, either the abrupt close or hinge movement can lead to thrombosis or hemolysis. By designing and prototyping an optically clear mechanical heart valve and test section, PIV can be used to study the flow around these regions more closely. Any unnatural flow patterns observed can be attributed to causing the formation of clots. Moreover, this process could be applied to novel valve designs to similarly compare the flow directly through the valve and determine the physiological relevance of the flow through the valve compared to existing valve designs. This research fills a gap in knowledge available through current literature, allowing for further innovation on bileaflet valve designs to improve their safety and effectiveness for patients. With successful completion of this project, a cost-effective method of fabrication of a clear heart valve and test section assembly using equipment available at SJSU will be developed. The hinge flow within the heart valve will also be characterized and validated to the St Jude MHV. Future research teams will be able to replicate this method to prototype both new and existing valve designs for use with PIV, as well as provide a basis for novel hinge designs to possibly reduce clotting. This will increase the body of knowledge on flow in heart valves, especially specific flow that increases risk of thrombogenicity. As more high quality research is available, product designers will be able to identify new valve designs that reduce non-physiological flow, and therefore reduce the risk of clotting. This will increase the speed of innovation in mechanical heart valves, and allow for a greater variety of novel valve geometries to be tested. Briseno", "novel hinge designs to possibly reduce clotting. This will increase the body of knowledge on flow in heart valves, especially specific flow that increases risk of thrombogenicity. As more high quality research is available, product designers will be able to identify new valve designs that reduce non-physiological flow, and therefore reduce the risk of clotting. This will increase the speed of innovation in mechanical heart valves, and allow for a greater variety of novel valve geometries to be tested. Briseno 4 Literature Review Current mechanical heart valves are known to cause clotting in patients, requiring them to be on anticoagulant medication for the rest of their lives. In order to improve the design and quality of mechanical heart valves in the future, it is important to understand the mechanism by which clots form. Many research groups hypothesize that stagnant flow and leakage around the hinge region of heart valves lead to blood damage, and therefore clotting 13,15,20 . Currently, there are clinical, in vitro, and computational studies that have shown that flow in the valve hinges could be contributing to clot formation 13 . However, there continues to be a gap in research that utilizes imaging systems to characterize the flow inside a mechanical heart valve. Figure 2: Diagram of Particle Image Velocimetry Setup for Analysis of Flow in Mechanical Heart Valves: a) flow simulator and b) test section 15 PIV is a commonly used method to understand various aspects of flow in a device. This type of imaging is especially useful for measuring velocity in high resolution through vector fields 21 . In the case of PIV utilized for analysis of MHV flow, the setup consists of a pump that simulates human aortic flow, a test section that holds the valve, a laser, and a CCD camera, as shown in Figure 2 15 . Fluid is run through the system, and is seeded with particles that the camera captures over time. The particle movement over time is tracked to develop a velocity field map in the area that is being imaged. In this specific study, the model was scaled up, with a heart valve with fixed leaflets in the system. This model observed cloning flow of the heart valve, and found unsteady flow in the hinge region 15 . By implementing PIV in conjunction with an optically clear heart valve, hinge flow and closing flow can be observed, which", "over time. The particle movement over time is tracked to develop a velocity field map in the area that is being imaged. In this specific study, the model was scaled up, with a heart valve with fixed leaflets in the system. This model observed cloning flow of the heart valve, and found unsteady flow in the hinge region 15 . By implementing PIV in conjunction with an optically clear heart valve, hinge flow and closing flow can be observed, which provides new insights on the performance of MHVs. There are three main examples of PIV being successfully utilized in this way. First, Klusak et al. used PIV to measure velocity in leakage jets caused by the hinge flow of scaled up bileaflet Briseno 5 mechanical heart valve models 15 . They were able to identify unsteady flow from the hinge region, which they hypothesized could lead to increased clotting. Vennemann et al. used PIV to evaluate the closing flow in a novel trileaflet heart valve 20 . Due to the clear valve, they were able to identify reduced jetting during closing flow, which could lower the risk of clotting in trileaflet valves compared to bileaflet. Finally, Jun et al. measured hinge flow in a standard St. Jude Medical bileaflet MHV 13 . For the first time in MHV research, they were able to obtain a complete flow field image of the hinge region of the valve. This work is also preceded by research done by Leo et al., although they only obtained a partial flow field 17 . Based on these papers, it is clear that PIV in combination with an optically clear valve provides an opportunity to characterize specific types of flow, especially hinge and closing flow, that is not possible with conventional PIV methods and valves. a) b) Figure 3: Transparent Heart Valves from Existing Literature - a) 3D Printed VeroClear Heart Valve 20 and b) Acrylic Valve Housing 13 There are two main considerations that must be taken when developing a PIV setup, the composition of the test fluid, and the material and design of the valve and test section. These materials must be optimized to limit optical distortion, as any distortion will lead to a less accurate velocity map output. To limit optical distortion, a solid-liquid system with matching refractive indices must be identified. Current literature on optically clear heart valves used with PIV provide a", "Housing 13 There are two main considerations that must be taken when developing a PIV setup, the composition of the test fluid, and the material and design of the valve and test section. These materials must be optimized to limit optical distortion, as any distortion will lead to a less accurate velocity map output. To limit optical distortion, a solid-liquid system with matching refractive indices must be identified. Current literature on optically clear heart valves used with PIV provide a starting point on possible solid-liquid systems. Vennemann et al. utilized a 3D printing approach, with the Objet Eden350V 3D printer and a transparent printing material known as VeroClear, with a refractive index of 1.48 20 . The transparent 3D printed valve and test section are shown in Figure 3a. They used this along with a matched solution of glycerol, water, and sodium iodide (19/23.5/57) wt%, which had a matching refractive index of 1.48, and Rhodamine-B tracer particles 20 . Jun et al. utilized an acrylic valve housing along with a mixture of 100% saturated Sodium Iodide solution, glycerin and water (79:20:1 by volume) that matched the acrylic refractive index at 1.49 13 . The acrylic valve housing is shown in Figure 3b. Briseno 6 Unfortunately, both methods described previously require solid-liquid systems with high refractive indices, 1.48 and 1.49. This requires large amounts of sodium iodide (NaI), which is both costly and creates a solution that is difficult to work with due to the high concentration. In order to identify possible alternatives with lower refractive indices, this literature review was expanded to review other possible solid-liquid systems, including those used in applications outside of MHVs. PIV is also used for vascular models, and there are two main examples of this application, both with different solid-liquid systems. Antonowicz et al. used a system consisting of Clear resin from Formlabs and 45 wt% glycerin solution, which had some distortion that required digital correction 2 . The clear resin model is shown in Figure 4b. However, similar to the previous systems using acrylic, this method requires a high refractive index solution to enable PIV and would therefore not be suitable for this application. Another solid-liquid system utilized Sylgard-184 to model blood vessels, and created a matching liquid with water, glycerol and sodium iodide salt (47.38/36.94/15.68 wt%) with RI=1.414 23 . The Sylgard-184 vascular model is shown in Figure 4a. Sylgard-184 is an ideal", "correction 2 . The clear resin model is shown in Figure 4b. However, similar to the previous systems using acrylic, this method requires a high refractive index solution to enable PIV and would therefore not be suitable for this application. Another solid-liquid system utilized Sylgard-184 to model blood vessels, and created a matching liquid with water, glycerol and sodium iodide salt (47.38/36.94/15.68 wt%) with RI=1.414 23 . The Sylgard-184 vascular model is shown in Figure 4a. Sylgard-184 is an ideal material for this application since its refractive index (1.41 to 1.43) is much lower than the polymers mentioned in the previous paragraph, making it easier to find a liquid with matching refractive index. In addition, Sylgard-184 can be molded easily, making it more suitable for modeling different geometries. a) b) Figure 4: Transparent Vascular Models from Existing Literature - a) Sylgard-184 Vascular Model 22 and b) Model Printed With Formlabs Clear Resin 2 Out of all of the solid materials mentioned above, Sylgard-184 is most ideal in this case due to its low refractive index (1.41 to 1.43) and that it can be molded. Based on this choice, existing literature was reviewed on molding Sylgard-184 using 3D printed molds. Currently, Fused Deposition Modeling (FDM) and Stereolithography (SLA) printers are available for use in the BME labs at SJSU, so the literature review was focused on these methods. Generally, FDM printing creates parts with rough surfaces that are not suitable for mold making, due to layer lines from printing. However, utilizing FDM printing along with chemical smoothing is a promising methodology. Ferretti et al. describes a method for creating smooth Briseno 7 molds using FDM printing of polyvinyl butyral (PVB) 10 . PVB has similar material properties to other commonly used filament materials, such as PLA and ABS. However, it is unique because the surface of 3D printed PVB parts can be smoothed with isopropyl alcohol (IPA), significantly improving the surface finish of the mold. Figure 5 shows the improvement in surface finish after IPA smoothing of PVB using the Polyshear IPA smoothing device. The Polyshear device was used as it provides controlled IPA vapor. This leads to more consistent smoothing compared to other smoothing methods, such as dipping or brushing the part in IPA. This method was used by Jiang et al. to successfully create a smooth mold that produced a suitable silicone phantom for PIV analysis 12 . Figure", "finish of the mold. Figure 5 shows the improvement in surface finish after IPA smoothing of PVB using the Polyshear IPA smoothing device. The Polyshear device was used as it provides controlled IPA vapor. This leads to more consistent smoothing compared to other smoothing methods, such as dipping or brushing the part in IPA. This method was used by Jiang et al. to successfully create a smooth mold that produced a suitable silicone phantom for PIV analysis 12 . Figure 5: 3D Printed PVB Part - a) before IPA smoothing and b) after IPA smoothing 10 Another option for fabricating a test section assembly would be using 3D printing and lost-core casting. Falk et al utilized a lost-core casting technique to create a vascular model for PIV out of Sylgard-184 with internal cavities 9 . Lost-core casting uses dissolvable material, such as polyvinyl alcohol (PVA), to print a positive model of the inner feature, as seen in Figure 6a. The core is placed in an acrylic box and silicone is poured around it. Once the silicone is cured, the inner part is dissolved, leaving the negative impression of the model, as seen in Figure 6b. Unfortunately, the dimensions of the negative impression can be inaccurate 7,9 . Therefore, this method is more suitable in producing vascular models with organic shapes rather than geometric models with precise dimensions. Despite this, these papers provide an example of Sylgard-184 phantoms being used under physiological flow conditions, and provide insight into the types of barbed connectors and thickness of Sylgard-184 to use. Briseno 8 (a) (b) Figure 6: Lost-core casting method for intracranial aneurysm using PVA core and Sylgard-184 casting 9 (a) 3D-printed positive model, (b) Resulting Sylgard part As an alternative to FDM printing, SLA resin printing provides high detail parts and was accessible to the group in the BME lab. However, there is limited formal literature about fabricating resin molds for Sylgard-184. One resource in particular, a Formlabs white paper, shows a promising method for fabricating silicone parts with resin molds for silicone earpieces 11 . The white paper includes procedures on designing, printing, and cracking eggshell molds, which are molds with a thin outer wall that can be cracked to retrieve the casted silicone part. Figure 7a shows a resin eggshell mold for a custom silicone earpiece, and Figure 7b shows the resulting part created with this mold. Based on the", "One resource in particular, a Formlabs white paper, shows a promising method for fabricating silicone parts with resin molds for silicone earpieces 11 . The white paper includes procedures on designing, printing, and cracking eggshell molds, which are molds with a thin outer wall that can be cracked to retrieve the casted silicone part. Figure 7a shows a resin eggshell mold for a custom silicone earpiece, and Figure 7b shows the resulting part created with this mold. Based on the white paper, the group believed pursuing a resin printing process with eggshell molds could be used to create parts with precise features. Given the procedure\u2019s effectiveness in producing a part without smoothing or accounting for overhang, the group considered this method to fabricate the valve housing. Due to the high print quality of resin printing, the group also considered printing regular (non-eggshell) molds out of resin. Briseno 9 (a) (b) Figure 7: Resin eggshell molds for silicone earpieces 11 (a) Filled resin eggshell mold and (b) Resulting silicone part After defining the method to create the test section and valve assembly, a blood mimicking refractive index matched liquid must be developed. Blood is generally considered a shear-thinning, non-Newtonian fluid, however, in tubes with diameter greater than 1mm, it behaves as a Newtonian fluid 4,24 . Based on the expected shear rate of the system, the viscosity of a Newtonian fluid can be adjusted to match that of blood. Glycerol is used to alter the viscosity of the solution, and a review of literature shows that a ratio of 62/38 water to glycerol can represent the median reported viscosity of blood at high shear rates 5 . In addition to the systems reported in the previous examples, review articles are available that summarize different solid-liquid systems that could be utilized for PIV, as well as charts that describe the refractive indices and viscosities of liquid solutions at different concentrations 1,5,21 . Briseno 10 Previous Experiments: Since last year, this project has already been in progress as part of a Senior design project with group members Nikitah Fernandes and Sunayana Pai. Specifically, completing the first aim. In a group, there were several attempts to create separate valve and test sections, such that the silicone valve housing would sit in place in the test section. An example of silicone valve housing and its mold is shown in Figure 8, where the housing was", "Briseno 10 Previous Experiments: Since last year, this project has already been in progress as part of a Senior design project with group members Nikitah Fernandes and Sunayana Pai. Specifically, completing the first aim. In a group, there were several attempts to create separate valve and test sections, such that the silicone valve housing would sit in place in the test section. An example of silicone valve housing and its mold is shown in Figure 8, where the housing was overall too flexible to successfully hold leaflets. This flexibility also contributed to concerns about how the valves would perform when in the test section with fluid pressure. For this reason, the group shifted focus to designing the test section and identifying how this part could securely hold a flexible valve housing and provide stability. (a) (b) Figure 8: 3-Part PVB Mold - (a) Mold and (b) Resulting Part Based on the flexibility of the valve housings and their inability to hold leaflets, the group decided to combine the valve housing and test section into one part. This ensures that the valve housing is as rigid as possible, and does not separate from the test section when exposed to physiological flow. The combined test section and valve mold was printed using PVB filament and smoothed using 99% IPA, as this process provides the best surface finish and dimensional accuracy. This final model is shown in Figure 9, along with its associated molds. The Sylgard was degassed prior to pouring it into the molds using a vacuum chamber, successfully eliminating all bubbles in the final part. In addition, the mold was designed to avoid leaks, and had an open top to prevent an incomplete fill and guarantee a perfect surface finish on this side. These improvements led to a final prototype that included all required material characteristics and dimensions. In addition, the process is easily replicable, with an overall timeline of 1 week for a complete part, including 3D printing, smoothing, filling, and demolding the parts. Briseno 11 (a) (b) Figure 9: Final Combined Valve and Test Section Prototype - (a) Molds and (b) Parts Additionally, the design for the impression of the valve was updated with a 1.30 mm fillet along its bottom inner edge to enable printing without overhang issues. The updated valve housing design can be seen in Figure 14c. In order to determine the final quality of the", "week for a complete part, including 3D printing, smoothing, filling, and demolding the parts. Briseno 11 (a) (b) Figure 9: Final Combined Valve and Test Section Prototype - (a) Molds and (b) Parts Additionally, the design for the impression of the valve was updated with a 1.30 mm fillet along its bottom inner edge to enable printing without overhang issues. The updated valve housing design can be seen in Figure 14c. In order to determine the final quality of the molded parts, key dimensions were measured and compared to the initial drawings. Figure 10 shows labels of the two test section parts and key dimension locations, both for each test section half and overall. Table 2 shows each key dimension, the expected value and the measured dimension on the final part. Most dimensions are measured in multiple locations, with the left side being closest to the valve as indicated in Figure 10b. Table 1: Dimension and Fit Verification Measurements Dimension Name Expected Dimension Value Measured Dimension (mm) Left Middle Right Small Side Thickness 31.75mm 27.70 27.87 27.94 Big Side Thickness 31.75mm 26.22 26.25 26.22 Big Side Width 52.83mm 52.36 52.44 52.64 Small Side Width 52.83mm 52. 52. 52.2 Briseno 12 35 81 8 Big Side Length 76.20mm 75.95 N/A 75.84 Small Side Length 76.20mm 76.00 N/A 75.93 Inner Diameter of Test Section 27.43mm 27.15 N/A 27.11 (a) (b) (c) Briseno 13 Figure 10: Key Dimensions to Be Measured (a) Big and Small Side of Test Section Valve Assembly, (b) Labeled Overall Dimension and Left Side Label, and (c) Key Thickness, Length and Width Dimensions (Same for Both Sides of Test Section) The results from Table 1 show a general decrease in size from the expected dimensions to the final measured dimensions on the part. This decrease was most prominent in the thickness of the part, with both the small and big sides having a much smaller thickness than expected. This was caused by a lack of measurement when filling up the molds to ensure that the part thickness was consistent. In the future, this can be avoided by either designing a fill line into the molds to ensure that the part is filled to the correct point, or filling the molds by weight to ensure that the correct quantity of Sylgard is added. However, a slightly different thickness will not contribute significantly to the function of the test section and", "was caused by a lack of measurement when filling up the molds to ensure that the part thickness was consistent. In the future, this can be avoided by either designing a fill line into the molds to ensure that the part is filled to the correct point, or filling the molds by weight to ensure that the correct quantity of Sylgard is added. However, a slightly different thickness will not contribute significantly to the function of the test section and can be accepted. The other notable issue was in the thickness across the part, particularly for the small side of the test section. In this case, the dimensions are slightly different across the part, increasing from the left to right side. This is likely due to the mold being placed on a surface that was not level while the Sylgard cured. In the future to ensure this does not happen, a level can be used to ensure that the parts are flat during curing. However, this small deviation only affects the outer thickness of the test section, and does not have any effect on the inner diameter of the test section hole, or the valve housing itself. For the rest of the dimensions, measurements were generally within 1 millimeter, which is within the tolerances for the part. In the future, if it was desired to correct this, the molds can be adjusted to be slightly smaller to adjust for printer error and increase the dimensions of the part to the expected values. In parallel, the group was also researching the amount of Sodium Iodide (NaI) that needed to be added to a base solution of water and glycerol to achieve a target RI.. Figure 11 shows the variable amounts of NaI added to a base solution of 62:38 water and glycerol (% weight) and its associated measured RI. From the values, a polynomial line of best fit was calculated, resulting in the polynomial equation y = 0.1172x 2 + 0.1310x + 1.3831. This equation serves as a direct correlation between the weight percent of NaI (x) added to the base solution and the expected RI value (y) of the resulting liquid. Using this equation, a solution with a specific RI could be created by solving for the amount of NaI that would need to be added to the base solution. The R 2 value for the best fit equation was", "in the polynomial equation y = 0.1172x 2 + 0.1310x + 1.3831. This equation serves as a direct correlation between the weight percent of NaI (x) added to the base solution and the expected RI value (y) of the resulting liquid. Using this equation, a solution with a specific RI could be created by solving for the amount of NaI that would need to be added to the base solution. The R 2 value for the best fit equation was 0.998. This value was calculated using Matlab, and indicates how closely the calculated equation fits with the measured data. In this case, the R 2 value is very close to 1, indicating that the equation is a good fit for the data. This will allow for close approximations for the NaI needed to be added to the base solution to achieve a specific RI. However, the RI of the solutions created still needs to be verified using a refractometer. Briseno 14 Figure 11: Graph of Refractive Index Solutions with Best Fit Line Plotted This RI matched liquid would then be used to measure the distortion of fabricated silicone parts. Because the RI of Sylagrad is typically 1.4-1.41 from literature, we created a solution of 51:31:19 of water, glycerol, NaI (% weight) that had a measured refractive index of 1.41. Although 1.41 is a target refractive index value for Sylgard, the actual refractive index can vary, so it is important to confirm visually as well. Figure 12 shows the piece of Sylgard-184 being dipped into the liquid. In the photo, the material disappears when submerged in the liquid, confirming that the refractive index of both materials matches. Figure 12: Image of Sylgard-184 Dipped Into Matched Liquid To determine the amount of distortion resulting from the test section and valve housing assembly, photos were taken of the test section in air, water, and in the RI matched liquid. Figure 13 shows the photos after cropping to focus on the area around the valve housing. In Figure 13a, the test section in air has a large amount of distortion, including both radial distortion Briseno 15 under the circular part of the test section, and general distortion around the valve housing, particularly in the hinge region. In Figure 13b in water, the distortion is improved, but still visible, particularly in the valve housing region. Finally, in Figure 13c with the RI matched liquid (RI", "photos after cropping to focus on the area around the valve housing. In Figure 13a, the test section in air has a large amount of distortion, including both radial distortion Briseno 15 under the circular part of the test section, and general distortion around the valve housing, particularly in the hinge region. In Figure 13b in water, the distortion is improved, but still visible, particularly in the valve housing region. Finally, in Figure 13c with the RI matched liquid (RI = 1.41, 50:31:19 water:glycerol:NaI by weight %), very little distortion is observed visually. In this figure, the main cause of distortion is the bubbles in the RI matched liquid, around the outside of the valve housing. (a) (b) (c) Figure 13: Test Section and Valve Assembly Distortion Photos (a) Air, (b) Water and (c) RI Matched Liquid In order to quantify the distortion of the part in the RI matched liquid, the Matlab Camera Calibration Toolbox was utilized to identify the pixel values of the corner points on the checkerboard in the reference image and in the part and RI matched liquid photo. The code for this data analysis can be found in Appendix 1. The Matlab Camera Calibration Toolbox detection outputs an array of pixel values for the corner points for both images. An overlay of the detected corner points on the original images can be seen in Figure 14a and b. Next, the area of each checkerboard square was estimated by multiplying the lengths of the top and left sides of each square. Then, the percent change in area was calculated for each square using the percent error formula: abs(actual area - ref area)/ref area. Based on this calculation, Figure 14c shows a surface plot of the percent error in area for each square, and Figure 14d shows a histogram with a distribution of the number of points over the percent area. The histogram shows that most points had a percent error less than 10%, with a few outliers between 10 and 20%. From reviewing the photo in Figure 14c of the part in RI matched liquid, a few bubbles can be seen, and the location of the high percent error squares matches the location of these bubbles. Since the bubbles are not expected to persist under physiological pressure in the mock circulation loop, these higher percent error values can be discounted when considering the overall distortion error", "had a percent error less than 10%, with a few outliers between 10 and 20%. From reviewing the photo in Figure 14c of the part in RI matched liquid, a few bubbles can be seen, and the location of the high percent error squares matches the location of these bubbles. Since the bubbles are not expected to persist under physiological pressure in the mock circulation loop, these higher percent error values can be discounted when considering the overall distortion error caused by the solid-liquid system. The remaining 5-10% error can easily be corrected by the PIV software during data analysis. Briseno 16 (a) (b) (c) (d) Figure 14: Distortion Testing Results (a) Corner points (reference checkerboard), (b) Corner points (assembly in liquid), (c) Colormap of Percent Distortion by Area of Each Checkerboard Square and (d) Distribution of percent area distortion The group also conducted flow tests to ensure that the parts would work in conjunction with the MCL. During the flow test, the group was unable to fill the loop up without the test section assembly leaking. The location of the leak is labeled in Figure 15, and was also observed on the other side of the test section in the same location. The leakage began as the loop was Briseno 17 being filled, and the rate of leakage increased as more water was added into the loop. Because of this, the pump could not be turned on and pressure and flow data was not collected. It was unclear visually why the leakage was occurring, although it was likely related to the size and shape of the resin printed connectors in relation to the test section assembly. Based on these results, the connector design will need to be updated to prevent leakage. In addition, the acrylic plates and clamps used did not provide enough even pressure to hold the test section assembly together. Therefore, an alternate setup of acrylic plates and clamps will also need to be developed for use in the MCL. The final issue observed was that the weight of the filled Tygon tubing pulled down on the connectors, which put more pressure on the silicone and could also have contributed to leakage. Figure 15: Leak Location of Test Section Assembly in MCL During this project, the group developed a rapid prototyping process to fabricate an optically transparent mechanical heart valve and test section for use with PIV", "and clamps will also need to be developed for use in the MCL. The final issue observed was that the weight of the filled Tygon tubing pulled down on the connectors, which put more pressure on the silicone and could also have contributed to leakage. Figure 15: Leak Location of Test Section Assembly in MCL During this project, the group developed a rapid prototyping process to fabricate an optically transparent mechanical heart valve and test section for use with PIV analysis. The part fabrication process developed is low cost, can be completed using materials available in the lab, and can be completed within a week from design to final part. To develop this process, the group went through many iterations of FDM and resin-printed molds for silicone prototypes. Resin-printed molds resulted in curing issues with the Sylgard, whereas PVB molds resulted in better quality silicone valve housings. However, valve housings made from silicone were not as rigid as expected, deviating the original plan of fabricating a separate test section and valve housing. As such, the group decided to pursue a combined test section and valve that would provide structure to the valve housing in order to better hold leaflets. Second, a blood-mimicking liquid with a matching refractive index to Sylgard-184 was also identified. Finally, distortion Briseno 18 tests were then conducted to quantify the distortion created by the solid-liquid system. Minimal distortion of 5 to 10% was observed in the resulting system using the Sylgard 184 parts along with the refractive index matching liquid, indicating that PIV will be successful. The group was then able to begin testing to validate that the part is suitable for use in the MCL. A leakage test without flow was conducted, and indicated that the test section assembly should be suitable for PIV. Then, the part was inserted into the MCL, and leakage issues were observed at the connection point between the resin connectors and test section. This testing revealed that although the prototyping method was successful, more work is needed to improve the design to reduce leakage in the MCL. Briseno 19 Materials and Methods: Aim #1: Prototype a Combined Test Section and Valve Using a Sacrificial Mold In this study, a combined test section and valve was fabricated based on the dimensions of a St. Jude bileaflet mechanical heart valve (MHV) and adapted to fit the existing mock circulation loop (MCL) for", "and test section. This testing revealed that although the prototyping method was successful, more work is needed to improve the design to reduce leakage in the MCL. Briseno 19 Materials and Methods: Aim #1: Prototype a Combined Test Section and Valve Using a Sacrificial Mold In this study, a combined test section and valve was fabricated based on the dimensions of a St. Jude bileaflet mechanical heart valve (MHV) and adapted to fit the existing mock circulation loop (MCL) for PIV testing. Compared to previous year\u2019s proposal, the design was altered to reduce the required volume of silicone elastomer (Sylgard 184, Dow Chemicals). This change was implemented to reduce overall material costs associated with the experiment. The completed design is shown in Figure 16. The Part also features perpendicular flat sides to allow for imaging of the fluid without distorting the laser and image captured by the camera. Figure 16: The combined test section and valve (top) and the complete setup (bottom) Briseno 20 The mold was designed to be sacrificial, ensuring that critical dimensions were maintained and allowing the part to be cast as a single piece. SolidWorks models of the combined mold are shown in Figures 17. Pressure ports were positioned 1 inch upstream and 3 inches downstream of the valve as shown in the part drawing in Figure 18. The mold featured 1 mm-thick walls for easy dissolution and was designed to prevent the need for supports, as supported areas proved difficult to smooth. The design also features no areas for infill. When in a vacuum, any areas with infill would expand and crack, creating bubbles in the sylgard. Figure 17: The mold for the Combined Test Section and Valve cut through to see the center portion Briseno 21 Figure 18: Drawing for the Mold, with measurements and section view A separate test section featuring an aortic sinus was fabricated using a previously designed sinus geometry developed by a previous student linked in Appendix 2. A key design modification included the addition of an internal ledge upstream of the aortic sinus, as shown in Figure 19, to prevent the valve from shifting into the sinus region during operation. This ensured proper valve positioning and preserved physiological flow dynamics within the test section. All external features of the test section, aside from the internal sinus impression, remained consistent with those of the combined test section and valve. These", "by a previous student linked in Appendix 2. A key design modification included the addition of an internal ledge upstream of the aortic sinus, as shown in Figure 19, to prevent the valve from shifting into the sinus region during operation. This ensured proper valve positioning and preserved physiological flow dynamics within the test section. All external features of the test section, aside from the internal sinus impression, remained consistent with those of the combined test section and valve. These included flat sidewalls to minimize optical distortion and pressure ports positioned for flow measurement. The corresponding molds were designed similarly, incorporating thin walls for easy dissolution and step printing ledges to prevent supports. Figure 19: Complete Aortic Test Section Briseno 22 Fabrication Process: Molds were fabricated using a Bambu Labs P1S 3D printer with Polymaker PolySmooth PVB Filament (1.75 mm, blue). The mold was oriented to print in one part, as shown in Figure 20. This configuration also eliminated the need for supports anywhere inside the mold. A 2-layer wall thickness was used to promote efficient isopropyl alcohol (IPA) flow during dissolution and a skirt was added to promote bed-adhesion. The seam line was also placed to the back of the object, such that it would not run through the hinge region of the valve, or any other areas where the flow will be directly imaged. Emphasis was placed on surface smoothness to prevent defects that could interfere with PIV measurements. Print settings are detailed in Appendix 3. During printing, it was critical to avoid pausing the print, as thermal contraction can cause plastic to shrink if cooling is not well-controlled. Resuming the print after a pause can result in a visible layer line at the interface, which transfers to the PDMS part and interferes with optical imaging. To further minimize dimensional changes, printed molds were allowed to cool completely within the printer\u2019s controlled chamber. Removing the part prematurely could lead to uneven cooling, introducing distortions in geometry. In this case, maintaining dimensional accuracy was essential to ensure fit and functionality. Figure 20: orientation of mold in slicer software Briseno 23 Once printed, mold components underwent IPA vapor smoothing to eliminate surface roughness. For the parts creating the pressure ports, solvent welding with 99% IPA was used to attach the pressure ports to the main body of the mold. Sylgard 184 base and curing agent were mixed at a 10:1", "lead to uneven cooling, introducing distortions in geometry. In this case, maintaining dimensional accuracy was essential to ensure fit and functionality. Figure 20: orientation of mold in slicer software Briseno 23 Once printed, mold components underwent IPA vapor smoothing to eliminate surface roughness. For the parts creating the pressure ports, solvent welding with 99% IPA was used to attach the pressure ports to the main body of the mold. Sylgard 184 base and curing agent were mixed at a 10:1 ratio, with a total mass of approximately 680 grams, calculated using the equation provided in Appendix 4. Following thorough mixing, the solution was divided into multiple containers, allowing each portion sufficient space to expand up to four times its original volume during vacuum degassing. The mixture was placed in a vacuum chamber to eliminate trapped air prior to casting. PDMS was poured into the mold in 3-inch increments, with degassing performed after each pour to ensure uniformity and minimize air entrapment. After the final degassing step, the mold was topped off with any remaining Sylgard to slightly overfill. After curing for 48 hours, or once fully solidified, wire cutters were used to cut off the bottom of the mold and completely open the inner tube. The entire mold was then submerged in an IPA dissolution tank, shown in Figure 21. A peristaltic pump circulated IPA through the center opening, with tubing positioned to allow IPA to flow through the center tube. The assembly was left to dissolve for 2 hours. Figure 21: IPA dissolution Tank The test section was removed from the dissolution tank, and the softened PVB wall sections were peeled away and discarded as shown in figure 22. To ensure complete dissolution of the center tube and pressure ports, the undissolved regions were selectively targeted by sealing open areas of the part with stoppers. A similar process was used for the fabrication of the aortic test section. Briseno 24 Figure 22: Process for dissolving removing the PVB from silicone part Fabrication of Additional Parts: Connectors were required to interface both the main inlet/outlet and the pressure ports of the test section with the mock circulation loop. The main connectors are shown in Figure 23, with its corresponding drawing in Figure 24. Each connector retained the basic geometry of standard tubing connectors from McMaster-Carr, with modified dimensions to ensure a secure fit with the silicone test section and a", "24 Figure 22: Process for dissolving removing the PVB from silicone part Fabrication of Additional Parts: Connectors were required to interface both the main inlet/outlet and the pressure ports of the test section with the mock circulation loop. The main connectors are shown in Figure 23, with its corresponding drawing in Figure 24. Each connector retained the basic geometry of standard tubing connectors from McMaster-Carr, with modified dimensions to ensure a secure fit with the silicone test section and a thinner cross section to reduce disruption of flow. The pressure ports were similarly designed based off of McMaster-Carr standard 1/4th in. tube connectors. A ledge was added at the base of each pressure port opening to prevent the connectors from protruding into the inner flow channel of the test section, ensuring accurate pressure readings without disrupting internal flow, as shown in figure 25. A holder was also fabricated to keep the test section in place and level below the camera and aligned with the laser, without disturbing optical view, shown in figure 26. Briseno 25 Figure 23: Connectors for the Test Section to the MCL Figure 24: The Dimensions of the Connector from Figure 21 Figure 25 pressure ports connector dimensions Briseno 26 Figure 26: Holder for the Test Section in MCL setup Matching Refractive Index Liquid: Due to the high cost of sodium iodide used in previous experiments, an alternative blood analog solution was developed using sodium chloride (NaCl), water, and glycerol. In this formulation, NaCl increased the density of the solution, while glycerol was used to raise viscosity to match that of blood for flow simulation. A graph in Figure 27 shows the relationship between the amount of NaCl added to a base solution of water and glycerol, which was used to determine the final concentration. Due to the mess created by the salt, and its corrosion of materials used in the MCL, the amount of salt used was limited to the amount needed for a refractive index of 1.4. The final working solution was prepared at a volume ratio of 54:34:12 (water:glycerol:NaCl by %weight). Briseno 27 Figure 27: RI for varying amounts of NaCL in a base solution of water and glycerol Verification of Dimensions and Optical Clarity: After curing, a visual inspection was conducted to confirm the absence of bubbles or contaminants, using a grid paper test to assess optical clarity. By placing a grid behind", "was limited to the amount needed for a refractive index of 1.4. The final working solution was prepared at a volume ratio of 54:34:12 (water:glycerol:NaCl by %weight). Briseno 27 Figure 27: RI for varying amounts of NaCL in a base solution of water and glycerol Verification of Dimensions and Optical Clarity: After curing, a visual inspection was conducted to confirm the absence of bubbles or contaminants, using a grid paper test to assess optical clarity. By placing a grid behind the model, transparency was evaluated to determine its suitability for PIV imaging. Comparative images were taken using the original sodium iodide solution (RI = 1.41) and the new NaCl-based solution (RI = 1.4), showing that visual distortion was subjectively minimal and sufficiently close for PIV applications, shown in figure 28. Briseno 28 Figure 28: Aortic test section in air vs. NaCL solution(top) and aortic test section in air vs. the NaI solution (bottom) Aim #2: Prototype Leaflets with Sylgard 184 Using a Thin Rigid Frame for Structural Support. Leaflet fabrication underwent 13 design iterations, as listed in Table 2. The first iteration consisted of a fully Sylgard leaflet, formed by injecting degassed Sylgard into a dissolvable PVB Briseno 29 mold using a syringe. Iterations 2 through 5 focused on testing various material combinations for the leaflet body and supporting frame. These included different pairings of Sylgard 184, Dow 732 Sealant, RTV silicone glue, and caulking materials. In all cases, the frame was 3D printed using PETG filament to provide structural support. Concurrently, iterations 6 through 8 were developed without any frame to assess whether the frame was necessary for proper leaflet performance. All leaflet designs from these initial trials are shown in Figure 29. Table 2: leaflet iterations Material Combination Rigid accurate 1 (Control) entirely sylgard leaflet N Y 2 frame + caulking + sylgard hinges Y N 3 frame + RTV glue + sylgard hinges Y N 4 frame + 723 sealant + sylgard sylgard Y N 5 frame + sylgard Y N 6 Caulking + sylgard hinges N N 7 RTV glue + sylgard hinges N N 8 723 sealant + sylgard hinges N N 9 Fuller frame + sylgard Y Y 10 thick edges only frame + sylgard Y Y 11 Fixed leaflets + frame + sylgard Y Y 12 PVB smoothed leaflets Y Y 13 Fixed PVB smooth leaflets Y Y Briseno 30 Figure 29: the", "+ 723 sealant + sylgard sylgard Y N 5 frame + sylgard Y N 6 Caulking + sylgard hinges N N 7 RTV glue + sylgard hinges N N 8 723 sealant + sylgard hinges N N 9 Fuller frame + sylgard Y Y 10 thick edges only frame + sylgard Y Y 11 Fixed leaflets + frame + sylgard Y Y 12 PVB smoothed leaflets Y Y 13 Fixed PVB smooth leaflets Y Y Briseno 30 Figure 29: the making of the first 8 leaflet iterations Results from the first eight iterations indicated that leaflets fabricated without a frame lacked the rigidity required to maintain their shape and dimensions. Non-Sylgard materials proved difficult to manipulate into geometries with sufficient accuracy or consistency. Based on these outcomes, later designs focused on combining a PETG frame with a Sylgard leaflet body to balance flexibility and dimensional stability. Functional testing was performed by inserting each leaflet into the combined test section and valve, followed by connection to the mock circulation loop. Simulated physiological conditions (60 bpm, 70 mL stroke volume) were gradually approached by slowly increasing stroke volume. However, all leaflets tested in these trials dislodged from the valve housing at stroke volumes near 30 mL, suggesting insufficient anchoring or dimensional mismatch. Subsequent iterations focused on two alternative leaflet designs: solid PVB leaflets with varying hinge dimensions (Figure 30) and fixed-position leaflets (Figure 31) to be used if the PVB models failed under flow. Testing of the PVB leaflets revealed that hinge dimensions were critical to function. Leaflets printed with oversized hinge features remained stuck and failed to open, while those with standard dimensions were ejected from the valve at moderate flow rates. To better understand these outcomes, a dimensional analysis was conducted to quantify changes in leaflet geometry following printing and IPA vapor smoothing. Results are presented in Table 3 and showed that dimensional change was minimal. A scaling factor of 1.01% was determined to produce accurate final dimensions post-smoothing, and all future leaflets were printed accordingly. Briseno 31 Figure 30: Different dimensions of PVB leaflets tested Figure 31; fixed closed leaflets made form PETG frame and PDMS Table 3: leaflet dimension analysis after printing and smoothing Sample CAD dim Print dim -->/2 Smooth dim -->/2 difference %error Note: diff of print 1 0.45 0.892 0.446 0.898 0.449 0.001 0.2222222222 0.004 2 0.46 0.911 0.4555 0.915 0.4575 0.0025 0.5434782609 0.0045 3", "1.01% was determined to produce accurate final dimensions post-smoothing, and all future leaflets were printed accordingly. Briseno 31 Figure 30: Different dimensions of PVB leaflets tested Figure 31; fixed closed leaflets made form PETG frame and PDMS Table 3: leaflet dimension analysis after printing and smoothing Sample CAD dim Print dim -->/2 Smooth dim -->/2 difference %error Note: diff of print 1 0.45 0.892 0.446 0.898 0.449 0.001 0.2222222222 0.004 2 0.46 0.911 0.4555 0.915 0.4575 0.0025 0.5434782609 0.0045 3 0.465 0.922 0.461 0.926 0.463 0.002 0.4301075269 0.004 4 0.47 0.93 0.465 0.938 0.469 0.001 0.2127659574 0.005 5 0.475 0.94 0.47 0.946 0.473 0.002 0.4210526316 0.005 Briseno 32 6 0.48 0.951 0.4755 0.958 0.479 0.001 0.2083333333 0.0045 7 0.4545 0.9 0.45 0.907 0.4535 0.001 0.2200220022 scale 1.01 0.0045 8 0.459 0.911 0.4555 0.915 0.4575 0.0015 0.3267973856 scale 1.02 0.0035 avg difference: 0.0015 0.0044 % diff: 0.15% 0.44% Despite adjustments, fixed-position leaflet designs also failed to remain in place as stroke volume increased. Observations suggested that the root cause of leaflet ejection was not the leaflet design itself, but expansion and contraction of the entire combined valve and test section block under pulsatile flow. This deformation likely disrupted the seating and alignment of the leaflet components during operation. General Fabrication Process: The finalized leaflet fabrication process is outlined in Figure 32. First, the leaflet mold and PETG frame were 3D printed at a scale factor of 1.01\u00d7 to account for minor dimensional changes following smoothing. The printed mold was then smoothed using 99% isopropyl alcohol (IPA) by filling the internal volume and allowing the IPA to evaporate naturally. Once smoothed, a thin layer of degassed Sylgard 184 was poured into the mold, followed by placement of the PETG frame into position. The assembly was placed in a vacuum chamber to remove trapped air. A second layer of Sylgard was then added to encapsulate the frame, and the mold was vacuum degassed again. Finally, any remaining cavities were filled with Sylgard as needed to complete the leaflet structure. Briseno 33 Figure 32: Final leaflet fabrication process: 3D print mold + frame at 1.01% scale \u2192 smooth with 99% IPA (fill mold + natural evaporation) \u2192 add first layer of degassed sylgard \u2192 add frame \u2192 vacuum \u2192 second layer sylgard \u2192 vacuum \u2192 fill remaining areas as needed. Test Section and Valve Brace: To address the issue of leaflet dislodgement caused by", "again. Finally, any remaining cavities were filled with Sylgard as needed to complete the leaflet structure. Briseno 33 Figure 32: Final leaflet fabrication process: 3D print mold + frame at 1.01% scale \u2192 smooth with 99% IPA (fill mold + natural evaporation) \u2192 add first layer of degassed sylgard \u2192 add frame \u2192 vacuum \u2192 second layer sylgard \u2192 vacuum \u2192 fill remaining areas as needed. Test Section and Valve Brace: To address the issue of leaflet dislodgement caused by expansion and contraction of the test section under pulsatile flow, a structural brace was developed to minimize deformation. The brace, shown in Figure 33, was 3D printed using PETG and designed to wrap around the test section while leaving a window open at the valve region for laser access and camera imaging. Its design was similar to the existing test section caddy, but provided more direct structural reinforcement by enclosing the outer wall of the silicone model. The two halves of the brace were fastened together using heat-set inserts and plastic screws, as shown in Figure 34, to reduce potential laser reflection. Despite this reinforcement, the brace did not resolve the issue, as subsequent tests using PVB leaflets continued to show instability. Figure 33: brace over the test section with plastic screws, with windows for laser and camera Briseno 34 Figure 34: CAD model of the test section brace, made in 3 parts As a next step, focus shifted toward stabilizing the valve itself rather than the entire test section. A previously fabricated Sylgard valve, molded last year using a sacrificial PVB core, was selected and placed inside a newly designed PETG brace shown in Figure 35. This second brace featured a dedicated window to allow unobstructed laser access to the hinge region and camera visibility during PIV imaging. The proposed plan involved filling the brace with Sylgard around the exposed window region using a three-part PVB mold to encapsulate both the valve and brace. The resulting component would then be inserted into the aortic test section with leaflets properly seated. However, due to time constraints, this final approach could not be tested. Figure 35: brace for the silicone valve (right) nd the silicone vae itself (left) Briseno 35 Aim #3: Validate the Model by Comparing Pressure Gradients Between Referenced Bileaflet Valve and Silicone Model To validate the solid-liquid system developed through a rapid prototyping process, a series of tests", "both the valve and brace. The resulting component would then be inserted into the aortic test section with leaflets properly seated. However, due to time constraints, this final approach could not be tested. Figure 35: brace for the silicone valve (right) nd the silicone vae itself (left) Briseno 35 Aim #3: Validate the Model by Comparing Pressure Gradients Between Referenced Bileaflet Valve and Silicone Model To validate the solid-liquid system developed through a rapid prototyping process, a series of tests were conducted to assess its suitability for particle image velocimetry (PIV). This system, consisting of a Sylgard 184 aortic sinus model and a custom-formulated NaCl-glycerol-water solution with a matched refractive index, was designed to enable optically clear flow visualization while maintaining anatomical accuracy. Validation focused on ensuring that this combination provided sufficient optical clarity and structural compatibility for accurate PIV measurements. Multiple types of heart valves were tested within the system to evaluate its versatility. Each valve was successfully imaged using PIV, confirming the system\u2019s ability to support a range of valve geometries. In one representative test, a monoleaflet valve was visualized using both the newly developed Sylgard-based system and a previously used acrylic test section (Figure 36) for comparison. Figure 37 presents side-by-side PIV images from both setups. The results demonstrated that the new solid-liquid system yielded comparable optical quality, validating its effectiveness for PIV applications in future hemodynamic studies. Figure 36: Preexisting Test Section with St. Jude MHV Briseno 36 Figure 37: PIV processed data form a monoleaflet valve from the previou test section made from acrylic and water (right) and the new test section made from PDMS and RI matching liquid (left) both during systeole Aim #4: Conduct PIV at the Hinge Region of the Silicone Model and Assess Areas of Non-Physiological Flow Revised Experimental Aim and Testing Conditions: Originally, the goal was to evaluate flow behavior at the hinge region of a mechanical heart valve under physiological conditions using the silicone model and PIV system. As described in the initial proposal and illustrated in Figure 38, the laser would be directed at incremental vertical planes across the hinge, starting 15 mm from the top of the test section and moving downward. This setup aimed to capture high-shear jetting flow during early diastole, a phase associated with valve closure and potential recirculation, which can contribute to thrombus formation. However, due to persistent instability of the leaflets during", "physiological conditions using the silicone model and PIV system. As described in the initial proposal and illustrated in Figure 38, the laser would be directed at incremental vertical planes across the hinge, starting 15 mm from the top of the test section and moving downward. This setup aimed to capture high-shear jetting flow during early diastole, a phase associated with valve closure and potential recirculation, which can contribute to thrombus formation. However, due to persistent instability of the leaflets during preliminary hinge flow trials, the fourth experimental aim was revised to focus on assessing bulk flow through the aortic sinus model using the St. Jude Regent mechanical heart valve (MHV). The objective was to identify signs of non-physiological flow under various simulated cardiac conditions. Briseno 37 Figure 38: The Laser with Respect to the Test Section as it Moves Along the Hinge Region The modified testing strategy involved inserting the St. Jude Regent MHV into the aortic sinus test section and conducting PIV under different flow conditions while maintaining a constant heart rate of 60 beats per minute (bpm). Two variables were altered in a matrix of five test cases: stroke volume (SV) and peak aortic pressure (ppk). The baseline test condition was set to 60 bpm, 70 mL stroke volume, and 140 mmHg ppk, which represents a typical healthy adult at rest. Stroke volume was then varied to 50 mL and 90 mL, still at 140 mmHg, to mimic hypovolemic (low blood volume) and high cardiac output states, respectively. In a separate set of tests, peak aortic pressure was adjusted to 125 mmHg and 180 mmHg while maintaining stroke volume at 70 mL, simulating mild hypotension and hypertension. All five test conditions are summarized in the graph shown in Figure 39. Briseno 38 Figure 39: matrix for experimental design with stroke Volume 70 and peak aortic pressure of 140 were the baseline for both variables These tests aimed to determine whether the MHV remained hemodynamically stable and whether any abnormal flow features, such as recirculation, asymmetry, or high-velocity jets, emerged under varying cardiovascular states for different patient physiologies. Flow fields obtained through PIV at each condition were evaluated for deviations from expected physiological behavior (compared to the baseline conditions of 60 bpm, 70SV, 140 ppk), helping to identify whether the valve exhibited compromised performance in any specific pressure or volume setting. Experimental Setup: The mock circulation loop (MCL), shown", "determine whether the MHV remained hemodynamically stable and whether any abnormal flow features, such as recirculation, asymmetry, or high-velocity jets, emerged under varying cardiovascular states for different patient physiologies. Flow fields obtained through PIV at each condition were evaluated for deviations from expected physiological behavior (compared to the baseline conditions of 60 bpm, 70SV, 140 ppk), helping to identify whether the valve exhibited compromised performance in any specific pressure or volume setting. Experimental Setup: The mock circulation loop (MCL), shown in Figure 40, was designed to replicate physiological cardiac conditions using a programmable Vivitro SuperPump system. The setup includes three compliance chambers positioned throughout the circuit to simulate arterial compliance and dampen pressure fluctuations. The Vivitro pump was programmed using a predefined physiological waveform provided by Vivitro, shown in Figure 41, which mimics the pressure and flow profile of a healthy cardiac cycle. However, because the aortic valve position in this setup was located farther upstream from the pump outlet compared to a clinical scenario, the pressure waveform recorded at the valve site displayed significant noise. To mitigate this, an additional compliance chamber was introduced upstream of the valve to smooth the signal. Ideally, a custom waveform would be required to fully compensate for this geometric shift, but due to time limitations, this was not implemented. Particle Image Velocimetry (PIV) measurements were acquired by directing the laser through the midplane of the valve, as illustrated in Figure 42, allowing visualization of flow patterns in the central orifice region. Briseno 39 Figure 40: MCL setup in BME labs at SJSU Figure 41: waveform for 60 bpm uploaded to the Vivitro superpump Briseno 40 Figure 42: The laser passes through the test section and valve through the midplane The entire mock circulation loop was filled with the refractive index (RI) matching liquid formulated from NaCl, glycerol, and water to ensure optical clarity for PIV imaging. Flow rate was measured downstream of the aortic sinus using a Transonic ultrasonic flow sensor, providing real-time data on volumetric flow through the system. Pressure measurements were obtained using two transducers positioned 1 inch upstream and 3 inches downstream of the mechanical valve, representing left ventricular pressure and aortic pressure, respectively. These measurements were used to calculate the pressure gradient across the valve during each cardiac cycle. Peak aortic pressure readings were adjusted between tests to set and validate the desired pressure conditions for each experimental case,", "aortic sinus using a Transonic ultrasonic flow sensor, providing real-time data on volumetric flow through the system. Pressure measurements were obtained using two transducers positioned 1 inch upstream and 3 inches downstream of the mechanical valve, representing left ventricular pressure and aortic pressure, respectively. These measurements were used to calculate the pressure gradient across the valve during each cardiac cycle. Peak aortic pressure readings were adjusted between tests to set and validate the desired pressure conditions for each experimental case, ensuring consistency and accurate replication of physiological and pathological states. Camera Calibration and DaVis Software Settings: To calibrate the camera, a custom calibration tool with a preset grid pattern was placed inside the flow loop at the same location and orientation as the test section, as shown in Figure 43. Using DaVis software, the system captured the calibration pattern to determine the spatial resolution of the imaging setup. This process produced a scale factor of 51.13 pixels per millimeter, shown in Figure 44. Briseno 41 Figure 43: the distortion correction tool is placed inside the test section to calibrate the camera (bottom) and the camera is adjusted to focus over the grid (top right and left) and assign a pixel/mm scale value Figure 44: the scale factor reported by the software is 51.13 pixel/mm During PIV acquisition, the high-speed camera was configured to capture at 1000 frames per second. The time delay between the two laser pulses (\u0394 t) was set to 400 microseconds, allowing sufficient particle displacement for accurate velocity estimation without introducing excessive motion blur. A dual-head laser system, operating at a nominal maximum power of 30 W per head, was used to illuminate the flow field. Each recording session lasted 6 seconds to ensure that at least five full cardiac cycles at 60 bpm were captured for analysis. 6 seconds was also the maximum that could be captured in this setup, due to restrictions on image capturing on the camera. The system would work by rapidly taking images on the camera which would be Briseno 42 transferred to the computer. Because the rate at which the camera transferred images to the computer was slower than the speed at which the images were taken, the images would backup onto the camera 64gb storage, which limited our total amount of collected data to 6 seconds, which was approximately 70gb for each data set. Briseno 43 Results: Flow Rate: Flow", "on the camera. The system would work by rapidly taking images on the camera which would be Briseno 42 transferred to the computer. Because the rate at which the camera transferred images to the computer was slower than the speed at which the images were taken, the images would backup onto the camera 64gb storage, which limited our total amount of collected data to 6 seconds, which was approximately 70gb for each data set. Briseno 43 Results: Flow Rate: Flow rate data were first trimmed using Excel to isolate the initial complete cardiac cycle, beginning at the onset of systole, which was identified by a sharp upward deflection in the waveform. Corresponding pressure data were trimmed in parallel to maintain temporal alignment. The processed datasets were then imported into LabChart for further analysis, as shown in Figure 45. To reduce signal noise and improve the clarity of waveform features, a smoothing function was applied using a Bartlett (triangular) window. This method, shown in Figure 46, assigns greater weight to the center of each window while tapering at the edges, preserving the general waveform shape while minimizing short-term fluctuations. This type of filtering is well suited for physiological signals such as flow rate curves. Figure 45: the flow rate graph fro 1 second for SV70 ppk 125 before(left) and after smoothing (right) Figure 46: software settings to smooth the curve in figure 45 Using LabChart\u2019s arithmetic functions, the negative portion of each flow rate waveform was isolated to identify reverse flow areas, where the flow was moving backwards through the Briseno 44 flow sensor. The integration tool shown in figure 47 was used to calculate the area under the curve during the negative phase, representing regurgitant volume, the volume of blood that reverses direction through the valve during each cardiac cycle. This reverse flow contributes to valve closure. The absolute value of the negative integral was recorded at the end of each cycle across six consecutive cardiac cycles. These values are reported in Table 4, which also includes the mean regurgitant volume and standard deviation for each test condition. Figure 47: integral settings se to create a graph of the total ml per beat for each beat across the data collected Table 4: average regurgitant volume for every test condition across 6 beats Units in ml 1st beat 2nd beat 3rd beat 4th beat 5th beat 6th beat Average Regurgitant Volume", "each cycle across six consecutive cardiac cycles. These values are reported in Table 4, which also includes the mean regurgitant volume and standard deviation for each test condition. Figure 47: integral settings se to create a graph of the total ml per beat for each beat across the data collected Table 4: average regurgitant volume for every test condition across 6 beats Units in ml 1st beat 2nd beat 3rd beat 4th beat 5th beat 6th beat Average Regurgitant Volume Stnd. Dev. Physio 60_SV70_Ppk125 7.505 7.024 7.526 7.115 8.744 6.768 7.447 0.699 Physio 60_SV70_Ppk180 7.550 6.144 7.004 7.304 9.439 5.136 7.096 1.449 Briseno 45 Physio 60_SV50_Ppk140 8.043 10.831 8.977 6.326 7.469 8.009 8.276 1.523 Physio 60_SV70_Ppk140 6.951 8.620 6.012 7.819 6.263 8.262 7.321 1.076 Physio 60_SV90_Ppk140 6.600 8.662 6.585 9.545 6.371 7.561 7.554 1.299 Each test condition\u2019s average regurgitant volume per beat, was compared to the baseline condition\u2019s (60 bpm, 70 stroke volume, 140 peak pressure) average regurgitant volume per beat using Welch\u2019s t test. Results are summarized in Table 5. Welch\u2019s t test was selected for its robustness in comparing two independent groups with unequal variances or sample sizes. To confirm the validity of the test, the assumption of normality was assessed using the Shapiro-Wilk test, which is appropriate for small sample sizes (n less than 50). All datasets satisfied the normality requirement. Table 5: two-tailed t-est to check for significant difference between test conditions and baseline Average Regurgitant Volume (L) Stnd. Dev. t-stat p-value Stnd. Error Physio 60_SV70_Ppk125 7.447 0.699 7.447 0.8158 0.5239 Physio 60_SV70_Ppk180 7.096 1.449 7.096 0.7669 0.7368 Physio 60_SV50_Ppk140 8.276 1.523 8.276 0.2415 0.7613 Physio 60_SV90_Ppk140 7.554 1.299 0.3381 0.7425 0.6887 Two tailed p values were calculated to assess whether regurgitant volume significantly differed from the baseline, regardless of the direction of change. This approach accounted for the possibility that deviations in flow could either increase or decrease due to changes in valve Briseno 46 performance from hemodynamic conditions. Table 5 highlights p values greater than 0.05 in red, indicating a lack of statistically significant difference from the baseline and suggesting that the observed variation may reflect random fluctuations rather than physiologically meaningful changes. Pressure Gradients: Pressure data were preprocessed by aligning the onset of systole with time zero to maintain consistency with the corresponding flow rate waveforms. The pressure gradient across the valve was calculated by subtracting the aortic pressure from the left ventricular", "performance from hemodynamic conditions. Table 5 highlights p values greater than 0.05 in red, indicating a lack of statistically significant difference from the baseline and suggesting that the observed variation may reflect random fluctuations rather than physiologically meaningful changes. Pressure Gradients: Pressure data were preprocessed by aligning the onset of systole with time zero to maintain consistency with the corresponding flow rate waveforms. The pressure gradient across the valve was calculated by subtracting the aortic pressure from the left ventricular pressure at each time point, producing a calculated \u0394 P waveform shown on Channel 4 in Figure 48. This waveform was compared to the pressure gradient recorded directly from the pressure sensor to assess agreement between the two measurement approaches in Channel 1. As shown in Figure 49, the calculated \u0394 P values were significantly higher in magnitude, and the overall waveform curve was also significantly different from \u0394 P sensor. Figure 48: from top to bottom: \u0394 P from sensors, left ventricular pressure, aortic pressure, and \u0394 P calculated Briseno 47 Figure 49: Baseline pressure change change values from the sensor vs. calculated Peak \u0394 P values obtained from the sensor were recorded over six consecutive cardiac cycles for each condition. The average and standard deviation for each case are listed in Table 6. Statistical comparisons were performed between each test condition and the baseline (60 bpm, 70 stroke volume, 140 peak pressure) using Welch\u2019s t test, with results summarized in Table 7. The Shapiro-Wilk test confirmed normality in all datasets, validating the use of Welch\u2019s method. Table 6: Average peak \u0394 P from the sensor for each for each test condition 1st beat 2nd beat 3rd beat 4th beat 5th beat 6th beat Avg. peak \u0394 P across 6 beats Stnd. Dev. Physio 60_SV70_Ppk125 25.565 25.372 25.329 25.542 25.384 25.409 25.434 0.097 Physio 60_SV70_Ppk180 36.545 36.464 36.792 36.748 36.724 36.531 36.634 0.137 Briseno 48 Physio 60_SV50_Ppk140 31.64 34.108 31.842 29.705 31.095 31 31.565 1.453 Physio 60_SV70_Ppk140 28.388 28.43 28.493 28.6 28.507 28.338 28.459 0.094 Physio 60_SV90_Ppk140 28.212 28.375 28.311 28.437 28.133 28.324 28.299 0.110 Table 7:two-tailed t-est to check for significant difference between test conditions and baseline Avg. peak dPressure across 6 beats Stnd. Dev. t-stat p-value Stnd. Error Physio 60_SV70_Ppk125 25.434 0.097 -55.0361 0.0 0.05498 Physio 60_SV70_Ppk180 36.634 0.137 120.8123 0.0 0.06766 Physio 60_SV50_Ppk140 31.565 1.453 5.2265 0.0033 0.59422 Physio 60_SV90_Ppk140 28.299 0.110 -2.7213 0.022 0.05904 Two", "31.842 29.705 31.095 31 31.565 1.453 Physio 60_SV70_Ppk140 28.388 28.43 28.493 28.6 28.507 28.338 28.459 0.094 Physio 60_SV90_Ppk140 28.212 28.375 28.311 28.437 28.133 28.324 28.299 0.110 Table 7:two-tailed t-est to check for significant difference between test conditions and baseline Avg. peak dPressure across 6 beats Stnd. Dev. t-stat p-value Stnd. Error Physio 60_SV70_Ppk125 25.434 0.097 -55.0361 0.0 0.05498 Physio 60_SV70_Ppk180 36.634 0.137 120.8123 0.0 0.06766 Physio 60_SV50_Ppk140 31.565 1.453 5.2265 0.0033 0.59422 Physio 60_SV90_Ppk140 28.299 0.110 -2.7213 0.022 0.05904 Two tailed p values were used to assess the significance of differences in peak \u0394 P. Values below 0.05 are highlighted in green in Table 7, indicating statistically significant changes from baseline. PIV Velocities: Particle image velocimetry (PIV) was performed for each test condition to visualize and quantify flow through the aortic sinus model. Table 8 summarizes the minimum, maximum, and average velocities, along with standard deviations, measured across all conditions. Vortices were also consistently observed during systole, forming near the valsalva of the sinus. Representative velocity field images highlighting these vortices are shown in Figure 50, demonstrating the characteristic recirculating patterns that develop in the sinus during forward flow. Briseno 49 Figure 50: mid-systole for the St. Jude Regent valve with vortex formation near the walls of the sinus for changing stroke volumes (bottoom) and changing peak aortic pressures (top) Briseno 50 Briseno 51 Discussion: Regurgitation: This study evaluated the regurgitant volumes of a St. Jude Medical bileaflet mechanical heart valve under various physiological conditions using a silicone test section modeled after the aortic sinus. The average regurgitant volumes and associated standard deviations for each condition are summarized in Table 5. Results showed that regurgitant volume varied with changes in both stroke volume and peak aortic pressure. The highest regurgitant volume occurred under the lowest stroke volume condition (SV 50), suggesting that reduced forward flow may allow for increased backflow through the valve. In contrast, increasing peak pressure from 125 to 180 mmHg at a constant stroke volume of 70 mL resulted in lower regurgitant volumes, indicating that elevated closing pressures may promote improved leaflet coaptation and reduce leakage. These findings are somewhat consistent with previously published in vitro studies. Durko et al. reported regurgitant volumes ranging from 2 to 10 mL for mechanical valves, which align with the results observed in this study. 10 On the contrary however, Milo et al. documented a regurgitant volume of approximately 3.45 mL", "to 180 mmHg at a constant stroke volume of 70 mL resulted in lower regurgitant volumes, indicating that elevated closing pressures may promote improved leaflet coaptation and reduce leakage. These findings are somewhat consistent with previously published in vitro studies. Durko et al. reported regurgitant volumes ranging from 2 to 10 mL for mechanical valves, which align with the results observed in this study. 10 On the contrary however, Milo et al. documented a regurgitant volume of approximately 3.45 mL for the St. Jude Medical valve under comparable conditions. 23 Variability in the regurgitant volume, as indicated by the standard deviations across test conditions, may be attributed to differences in leaflet dynamics or flow turbulence. The highest variability was observed at the highest peak pressure (180 mmHg), suggesting that elevated pressures may introduce instability in valve behavior. Although the silicone test section offers a controlled environment for evaluating valve performance, it does not fully replicate the compliance and complex anatomical structure of the human cardiovascular system. Future work may incorporate more anatomically detailed models or computational simulations to better characterize the hemodynamic performance of mechanical valves under physiological conditions. Pressure Change Calculated vs. Sensor: A noticeable discrepancy was observed in the pressure waveforms generated from the experiment. Specifically, four curves were analyzed and compiled in the results section, shown in Figure 48: the \u0394 P from the sensor (Channel 1), left ventricular pressure (Channel 2), aortic pressure (Channel 3), and the calculated \u0394 P (Channel 4), which was obtained by subtracting the aortic pressure from the left ventricular pressure. The calculated pressure gradient, however, produced an unexpectedly high peak value, exceeding 100 mmHg. This was inconsistent with expectations, as sensor-based \u0394 P readings consistently peaked around 20 mmHg under all test conditions. In addition to the inflated magnitude, the shape of the calculated \u0394 P curve appeared abnormal. Typically, the pressure gradient is expected to reach a peak just before the left ventricular and aortic pressure curves reach their maximum values, then return toward zero during valve opening when pressures in the ventricle and aorta equalize. However, in the Briseno 52 calculated curve, the values remained well below zero for most of the cycle. This suggests a misconfiguration during data acquisition. It is likely that the left ventricular pressure and \u0394 P channels were misassigned or mislabeled during the LabVIEW data export process. Supporting this assumption is the observation that", "the left ventricular and aortic pressure curves reach their maximum values, then return toward zero during valve opening when pressures in the ventricle and aorta equalize. However, in the Briseno 52 calculated curve, the values remained well below zero for most of the cycle. This suggests a misconfiguration during data acquisition. It is likely that the left ventricular pressure and \u0394 P channels were misassigned or mislabeled during the LabVIEW data export process. Supporting this assumption is the observation that the curve labeled as left ventricular pressure has a higher than expected minimum, while the calculated \u0394 P curve remains negative for an extended period, which is unexpected. Further verification of channel assignments in the raw data would be necessary to resolve this issue and ensure valid interpretation of pressure measurements. Peak Pressure Change: The pressure drops across the St. Jude Medical bileaflet mechanical heart valve varied with changes in stroke volume and peak aortic pressure, as summarized in Table 7. The highest average pressure drop was measured under the test condition with peak aortic pressure of 180 mmHg, while the lowest occurred under a peak aortic pressure of 125 mmHg. This pattern reflects the influence of peak aortic pressure on transvalvular flow, where increased driving pressure results in higher flow velocities and, consequently, larger pressure differentials across the valve, so both of these results were expected. The values also had p-values of 0, indicating that the difference in these test conditions to the baseline were entirely due to the changes in peak aortic pressure rather than data fluctuations. Changes in stroke volume also affected the pressure drop. The test condition with 50ml SV produced the greatest variability, with a standard deviation of 1.453 mmHg. This variability may result from unstable leaflet behavior or irregular flow development at reduced stroke volumes, which can lead to transient pressure gradients and fluctuations in valve performance. 5 When peak pressure was held constant at 140 mmHg, the pressure drop decreased as stroke volume increased. Stroke volumes of 50, 70, and 90 mL produced pressure drops of 31.565, 28.459, and 28.299 mmHg, respectively. This inverse relationship suggests that higher stroke volumes may facilitate smoother flow through the valve, reducing energy dissipation and turbulence. Previous studies have similarly reported that bileaflet valves maintain more stable pressure profiles under moderate to high flow conditions, while lower cardiac outputs are associated with less efficient performance. 13 Low", "at 140 mmHg, the pressure drop decreased as stroke volume increased. Stroke volumes of 50, 70, and 90 mL produced pressure drops of 31.565, 28.459, and 28.299 mmHg, respectively. This inverse relationship suggests that higher stroke volumes may facilitate smoother flow through the valve, reducing energy dissipation and turbulence. Previous studies have similarly reported that bileaflet valves maintain more stable pressure profiles under moderate to high flow conditions, while lower cardiac outputs are associated with less efficient performance. 13 Low standard deviations observed under most conditions indicate a high degree of consistency and repeatability in the flow environment. The inclusion of a silicone test section modeled after the aortic sinus likely contributed to this stability by promoting gradual flow expansion and the formation of sinus vortices. These features are known to support pressure recovery and enhance flow transitions downstream of mechanical heart valves. 21 PIV Velocity Observations: Velocity, pressure, and regurgitation data revealed consistent relationships across test conditions. Cases with higher transvalvular pressure gradients also exhibited higher peak velocity magnitudes. For instance, the SV70_Ppk180 condition, which recorded the highest pressure drop (36.634 mmHg), also produced the highest maximum velocity (1.79 m/s) and the Briseno 53 largest velocity standard deviation (0.378 m/s), indicating more intense and variable flow through the valve. The SV50_Ppk140 condition, which resulted in the highest regurgitant volume (0.497 L), also showed elevated velocity fluctuations. This observation suggests that lower stroke volumes may lead to unsteady shear layers and incomplete valve closure, contributing to chaotic flow and increased backflow potential. 13 Test conditions with reduced stroke volumes produced lower average velocities but showed relatively larger fluctuations in flow, pointing to more turbulent and less directed velocity fields. In contrast, conditions such as SV70_Ppk140 demonstrated moderate peak velocities and the lowest standard deviations, suggesting more stable and uniform flow behavior through the valve. PIV Vortex Formation: Vortex structures consistently formed within the sinuses of Valsalva, particularly along the lateral walls during valve closure and early diastole. These vortices play a functional role in aiding valve leaflet closure and promoting coronary perfusion, but they also generate localized recirculation zones that can trap blood and increase the risk of thrombosis. 5, 21 In test conditions such as SV70_Ppk125, the observed vortices were relatively weak and symmetric, while in others, including SV90_Ppk140, vortex asymmetry was more pronounced. This asymmetry may result from shifts in flow dominance or subtle geometric mismatches within the", "walls during valve closure and early diastole. These vortices play a functional role in aiding valve leaflet closure and promoting coronary perfusion, but they also generate localized recirculation zones that can trap blood and increase the risk of thrombosis. 5, 21 In test conditions such as SV70_Ppk125, the observed vortices were relatively weak and symmetric, while in others, including SV90_Ppk140, vortex asymmetry was more pronounced. This asymmetry may result from shifts in flow dominance or subtle geometric mismatches within the system. The mechanical valve was implanted at a fixed angle relative to the aortic sinus, which may have influenced the observed flow asymmetries. Valve orientation with respect to sinus geometry is known to affect flow jet direction and shear distribution, particularly in bileaflet valves where skewed jets can intensify vortex formation on one side while weakening it on the other. 16 As a result, small deviations in alignment can significantly alter downstream flow fields, impacting both pressure recovery and the development of stagnation zones. These effects highlight the importance of precise valve positioning in both surgical practice and experimental modeling. Limitations of the Study: Several factors limited the scope and accuracy of this study, including blood analog restrictions, \u0394 P concerns, leaflet function, and valve orientation. Although the refractive index-matched solution enabled high-quality optical access for PIV, the physical properties of the NaCl-glycerol mixture differed significantly from those of blood. While the solution matched the refractive index of Sylgard 184, it did not replicate the non-Newtonian viscosity or particulate nature of blood, which may have influenced flow characteristics, particularly in low-shear regions and near valve surfaces. Future work could incorporate xanthan gum in the blood analog. An additional limitation was identified in the pressure data due to suspected mislabeling or misassignment of sensor channels during data export from the LabVIEW software. The calculated pressure gradient curve exhibited unphysiological behavior, including an abnormally Briseno 54 high peak and prolonged negative values, suggesting that the left ventricular pressure and \u0394 P signals may have been inadvertently swapped. As a result, the accuracy of the calculated transvalvular pressure gradient is uncertain, limiting confidence in comparisons between sensor and calculated pressure data. Another limitation was the rigidity of the silicone test section. The model successfully replicated the anatomical geometry of the aortic sinus, but it did not mimic the compliance of native vascular tissue. This lack of wall elasticity likely affected pressure wave propagation", "suggesting that the left ventricular pressure and \u0394 P signals may have been inadvertently swapped. As a result, the accuracy of the calculated transvalvular pressure gradient is uncertain, limiting confidence in comparisons between sensor and calculated pressure data. Another limitation was the rigidity of the silicone test section. The model successfully replicated the anatomical geometry of the aortic sinus, but it did not mimic the compliance of native vascular tissue. This lack of wall elasticity likely affected pressure wave propagation and flow behavior during phases of rapid deceleration, such as early diastole, where arterial compliance plays a critical role in modulating reverse flow and pressure recovery. Attempts to analyze hinge-level flow were hindered by challenges in leaflet fabrication and stability. Despite iterative efforts using various materials and bracing configurations, none of the prototypes functioned at physiological conditions. Leaflets were either ejected at physiological stroke volumes or failed to open due to friction or dimensional inaccuracies. As a result, the original aim of investigating hinge jet behavior through PIV could not be fulfilled, and the study was redirected to focus on bulk flow patterns and regurgitant behavior. The fixed orientation of the mechanical valve within the test section presented an additional constraint in evaluating flow dynamics. Bileaflet valves are known to exhibit orientation sensitive behavior, particularly in the formation of asymmetric jets and vortices. The valve was positioned at a constant angle throughout all experiments, for consistency and simplicity for initial trials. Future work could study flow dynamics through the aortic sinu with different orientations of the valve. However, given the current setup, certain positions of the valve in the aorta could not be explored without disassembly. This limitation was compounded by the fixed placement of the camera and laser system, which were mounted directly above the model. Moreover, the valve is restricted to its current position to prevent the influence of gravity on the opening and closing of the leaflets. The only variable that can be changed would be the orientation of the aortic sinus in the silicone model. However, the setup would need to be changed to study flow through the sinus when a valsalva is directly in line with the hinges of the valve. Any attempt to vary valve orientation would require repositioning both the laser and the camera to maintain proper imaging alignment and valve orientation which was not feasible within the scope and time frame of", "The only variable that can be changed would be the orientation of the aortic sinus in the silicone model. However, the setup would need to be changed to study flow through the sinus when a valsalva is directly in line with the hinges of the valve. Any attempt to vary valve orientation would require repositioning both the laser and the camera to maintain proper imaging alignment and valve orientation which was not feasible within the scope and time frame of the study. Consequently, the impact of valve orientation on sinus flow and hinge level dynamics remains unexplored in the current experimental setup. Time constraints further limited the completion of certain proposed refinements. The final version of the laser-accessible valve brace, as well as follow-up testing using fixed-leaflet configurations, could not be implemented within the study timeline. These untested components may hold potential for improving leaflet stability and enabling more detailed analysis of localized flow phenomena in the hinge region of the St. Jude Regent MHV. Briseno 55 Budget: Item Cost Per Item Quantity Total Cost Sylgard-184 (1.1 lb g) $200 1 $200 Polymaker PolySmooth PVB Filament $37 1 $37 99% Isopropyl Alcohol (16oz) $10 1 $10 Large Sealing Bin $29 1 $29 Peristaltic Pump $29 1 $29 Glycerol (1 gal) $27 1 $27 1/4in Tubing $16 1 $16 FormLabs Clear Resin (V4.1) $149 1 $149 PIV setup \u2013 \u2013 \u2013 Morton Iodized Salt (26oz) $3.5 3 $10.5 Formlabs Resin Printer - - - 3D filament Printer - - - Total $507.5 Briseno 56 Conclusion: This study demonstrated the feasibility of using a rapid prototyping approach to develop a solid-liquid testing platform for evaluating mechanical heart valve performance under physiologically relevant conditions. A silicone aortic sinus model, paired with a refractive index-matched NaCl-glycerol solution, enabled high-resolution particle image velocimetry (PIV) and reliable pressure and flow measurements. The St. Jude Medical bileaflet valve was evaluated under a range of stroke volumes and peak pressures to assess transvalvular pressure gradients, regurgitant volume, and flow dynamics. The results revealed clear correlations between hemodynamic loading conditions and valve behavior, including changes in velocity profiles, regurgitant flow, and vortex formation within the sinus region. The developed test system allowed for consistent, repeatable measurements and captured physiologically relevant flow features, supporting its utility in future in vitro valve testing applications. This work contributes to the broader effort of improving experimental models for evaluating cardiovascular device performance in", "peak pressures to assess transvalvular pressure gradients, regurgitant volume, and flow dynamics. The results revealed clear correlations between hemodynamic loading conditions and valve behavior, including changes in velocity profiles, regurgitant flow, and vortex formation within the sinus region. The developed test system allowed for consistent, repeatable measurements and captured physiologically relevant flow features, supporting its utility in future in vitro valve testing applications. This work contributes to the broader effort of improving experimental models for evaluating cardiovascular device performance in physiologically realistic settings. By replacing expensive materials such as sodium iodide with a cost-effective NaCl-glycerol solution, and by using an optically clear, anatomically inspired silicone model, the system balances affordability, reusability, and imaging compatibility. The modular nature of the design, including the interchangeable valve holders and RI-matched liquid, positions it as a versatile platform for studying a wide range of prosthetic valves. Moreover, because the process can be rapidly prototyped, the silicone model can easily be adapted to different anatomies or testing setups. Beyond experimental testing, the system may serve as a validation tool for computational models as well, helping to bridge the gap between simulation and real-world physiological behavior. Briseno 57 Appendix List Appendix 1 - Matlab Code For Data Analysis The Matlab Code used for data analysis can be downloaded at the following link: Matlab Code for Data Analysis Appendix 2 - Arctic Sinus Model Solidworks files for Sinus and all part iterations Appendix 3 - Specific Print Settings for Dissolvable Molds printed with PVB Bambu Slicer Process Presets for P1S Appendix 4 - equations (1) Specific gravity of Sylgard 184 at 25* = 1030 kg/m3 Specific gravity = density of object / density of liquid (generally water) Density of sylgard = 0.000103 g/mm^3 Volume of Combined test section and valve from Solidworks = 323.96 cubic centimeters Density = mass/volume Mass of sylgard = 33.37 grams (2) Equation of RI from graph: y = 0.1172x 2 + 0.1310x + 1.3831 Target RI = 1.41 1.41 = 0.1172x 2 + 0.1310x + 1.3831 X = 19% weight of the base solution of water and glycerol added to achieve an RI of 1.41 Appendix 5 - Past Project Proposal and Report Fall 2023 Project Proposal Spring 2024 Senior Project Final Report Appendix 5 - Raw + Analyzed Data Pressure and Flow Rate data PIV data stored separately, but images and videos can be found here Briseno 58 References 1. Agrawal,", "+ 1.3831 Target RI = 1.41 1.41 = 0.1172x 2 + 0.1310x + 1.3831 X = 19% weight of the base solution of water and glycerol added to achieve an RI of 1.41 Appendix 5 - Past Project Proposal and Report Fall 2023 Project Proposal Spring 2024 Senior Project Final Report Appendix 5 - Raw + Analyzed Data Pressure and Flow Rate data PIV data stored separately, but images and videos can be found here Briseno 58 References 1. Agrawal, Y. K., R. Sabbagh, S. Sanders, and D. S. Nobes. Measuring the Refractive Index, Density, Viscosity, pH, and Surface Tension of Potassium Thiocyanate (KSCN) Solutions for Refractive Index Matching in Flow Experiments. J. Chem. Eng. Data 63:1275\u20131285, 2018. 2. Antonowicz, A., K. Wojtas, \u0141. Makowski, W. Orciuch, and M. Koz\u0142owski. Particle Image Velocimetry of 3D-Printed Anatomical Blood Vascular Models Affected by Atherosclerosis. Materials 16:, 2023. 3. Bax, J. J., and V. Delgado. Bioprosthetic Heart Valves, Thrombosis, Anticoagulation, and Imaging Surveillance \u2217 . JACC Cardiovasc. Interv. 10:388\u2013390, 2017. 4. Bellofiore, A., E. M. Donohue, and N. J. Quinlan. Scale-up of an unsteady flow field for enhanced spatial and temporal resolution of PIV measurements: application to leaflet wake flow in a mechanical heart valve. Exp. Fluids 51:161\u2013176, 2011. 5. Bluestein, D., Rambod, E., & Gharib, M. (2000). Vortex shedding as a mechanism for free emboli formation in mechanical heart valves. Journal of Biomechanical Engineering, 122(2), 125\u2013134. https://doi.org/10.1115/1.429634 6. Brindise, M. C., M. M. Busse, and P. P. Vlachos. Density and Viscosity Matched Newtonian and non-Newtonian Blood-Analog Solutions with PDMS Refractive Index. Exp. Fluids 59:, 2018. 7. Choudhary, S. K., S. Talwar, and B. Airan. Choice of prosthetic heart valve in a developing country. Heart Asia 8:65\u201372, 2016. 8. Cozzi, F., G. Felisati, and M. Quadrio. Velocity Measurements in Nasal Cavities by Means of Stereoscopic PIV - Preliminary Tests. J. Phys. Conf. Ser. 882:, 2017. 9. Dow. Sylgard 184 Product Description.pdfat <https://www.dow.com/en-us/pdp.sylgard-184-silicone-elastomer-kit.01064291z.html#overview> 10. Durko, A., Budde, R. P. J., Geleijnse, M. L., & Kappetein, A. P. (2018). Recognition, assessment and management of the mechanical complications of acute myocardial infarction. Heart, 104(14), 1216\u20131223. https://doi.org/10.1136/heartjnl-2017-311473 11. Falk, K. L., R. Medero, and A. Rold\u00e1n-Alzate. Fabrication of Low-Cost Patient-Specific Vascular Models for Particle Image Velocimetry. Cardiovasc. Eng. Technol. 10:500\u2013507, 2019. 12. Ferretti, P., G. M. Santi, C. Leon-Cardenas, M. Freddi, G. Donnici, L. Frizziero, and A. Liverani. Molds with Advanced Materials for Carbon Fiber Manufacturing with", "Durko, A., Budde, R. P. J., Geleijnse, M. L., & Kappetein, A. P. (2018). Recognition, assessment and management of the mechanical complications of acute myocardial infarction. Heart, 104(14), 1216\u20131223. https://doi.org/10.1136/heartjnl-2017-311473 11. Falk, K. L., R. Medero, and A. Rold\u00e1n-Alzate. Fabrication of Low-Cost Patient-Specific Vascular Models for Particle Image Velocimetry. Cardiovasc. Eng. Technol. 10:500\u2013507, 2019. 12. Ferretti, P., G. M. Santi, C. Leon-Cardenas, M. Freddi, G. Donnici, L. Frizziero, and A. Liverani. Molds with Advanced Materials for Carbon Fiber Manufacturing with 3D Printing Technology. Polymers 13:, 2021. 13. Grigioni, M., Daniele, C., Morbiducci, U., D'Avenio, G., & Balducci, A. (2004). The power-law mathematical model for blood damage prediction: Analytical developments and physical inconsistencies. Artificial Organs, 28(5), 467\u2013475. https://doi.org/10.1111/j.1525-1594.2004.00015.x 14. Formlabs. Formlabs Eggshell Mold Instructions.pdf Briseno 59 15. Jiang, L., and F. Hassanipour. Bio-Inspired Breastfeeding Simulator (BIBS): A Tool for Studying the Infant Feeding Mechanism. IEEE Trans. Biomed. Eng. 67:3242\u20133252, 2020. 16. Jimenez, J. H., & Sotiropoulos, F. (2010). The effect of implantation orientation of a bileaflet mechanical heart valve on kinematics and hemodynamics in an anatomic aorta. Journal of Biomechanics, 43(12), 2294\u20132300. https://doi.org/10.1016/j.jbiomech.2010.02.018 17. Jun, B. H., N. Saikrishnan, and A. P. Yoganathan. Micro particle image velocimetry measurements of steady diastolic leakage flow in the hinge of a St. Jude Medical\u00ae regent TM mechanical heart valve. Ann. Biomed. Eng. 42:526\u2013540, 2014. 18. Kim, W., Choi, H., Kweon, J., Yang, D. H., & Kim, Y. (2020). Effects of pannus formation on the flow around a bileaflet mechanical heart valve. PLoS ONE , 15 (6), e0234341. https://doi.org/10.1371/journal.pone.0234341 19. Klusak, E., A. Bellofiore, S. Loughnane, and N. J. Quinlan. High-Resolution Measurements of Velocity and Shear Stress in Leakage Jets From Bileaflet Mechanical Heart Valve Hinge Models. J. Biomech. Eng. 137:111008, 2015. 20. Lamson, T. C., G. Rosenberg, D. B. Geselowitz, S. Deutsch, D. R. Stinebring, J. A. Frangos, and J. M. Tarbell. Relative blood damage in the three phases of a prosthetic heart valve flow cycle. ASAIO J. 39:M626\u201333, 1993. 21. Leo, H. L., & Sotiropoulos, F. (2006). Computational analysis of aortic flow disturbances caused by mechanical heart valve implantation. Journal of Biomechanical Engineering, 128(4), 588\u2013594. https://doi.org/10.1115/1.2345413 22. Leo, H.-L., Z. He, J. T. Ellis, and A. P. Yoganathan. Microflow fields in the hinge region of the CarboMedics bileaflet mechanical heart valve design. J. Thorac. Cardiovasc. Surg. 124:561\u2013574, 2002. 23. Milo, S., & Vaturi, M. (2001). Assessment of physiological prosthetic heart valve regurgitation. European", "flow cycle. ASAIO J. 39:M626\u201333, 1993. 21. Leo, H. L., & Sotiropoulos, F. (2006). Computational analysis of aortic flow disturbances caused by mechanical heart valve implantation. Journal of Biomechanical Engineering, 128(4), 588\u2013594. https://doi.org/10.1115/1.2345413 22. Leo, H.-L., Z. He, J. T. Ellis, and A. P. Yoganathan. Microflow fields in the hinge region of the CarboMedics bileaflet mechanical heart valve design. J. Thorac. Cardiovasc. Surg. 124:561\u2013574, 2002. 23. Milo, S., & Vaturi, M. (2001). Assessment of physiological prosthetic heart valve regurgitation. European Journal of Cardio-Thoracic Surgery, 19(6), 1025\u20131030. https://doi.org/10.1016/S1010-7940(01)00790-6 24. Nath, S. S., and S. Parashar. Perioperative management of patients with prosthetic heart valves-A narrative review. Ann. Card. Anaesth. 25:254\u2013263, 2022. 25. Raghav, V., S. Sastry, and N. Saikrishnan. Experimental Assessment of Flow Fields Associated with Heart Valve Prostheses Using Particle Image Velocimetry (PIV): Recommendations for Best Practices. Cardiovasc. Eng. Technol. 9:273\u2013287, 2018. 26. Vennemann, B. M., T. R\u00f6sgen, T. P. Carrel, and D. Obrist. Time-Resolved Micro PIV in the Pivoting Area of the Triflo Mechanical Heart Valve. Cardiovasc. Eng. Technol. 7:210\u2013222, 2016. 27. Wright, S. F., I. Zadrazil, and C. N. Markides. A review of solid\u2013fluid selection options for optical-based measurements in single-phase liquid, two-phase liquid\u2013liquid and multiphase solid\u2013liquid flows. Exp. Fluids 58:108, 2017. 28. Yin, W., I. B. Krukenkamp, A. E. Saltman, G. Gaudette, K. Suresh, O. Bernal, J. Jesty, and D. Bluestein. Thrombogenic performance of a st. Jude bileaflet mechanical heart valve in a Briseno 60 sheep model. ASAIO J. 52:28\u201333, 2006. 29. Yousif, M. Y., D. W. Holdsworth, and T. L. Poepping. A blood-mimicking fluid for particle image velocimetry with silicone vascular models. Exp. Fluids 50:769\u2013774, 2011. 30. Zakaria, M. S., F. Ismail, M. Tamagawa, A. F. A. Aziz, S. Wiriadidjaja, A. A. Basri, and K. A. Ahmad. Review of numerical methods for simulation of mechanical heart valves and the potential for blood clotting. Med. Biol. Eng. Comput. 55:1519\u20131548, 2017.", "1 Mock Circulatory Loop: Use of Sylgard 184 and Replication of Physiological Conditions In Fulfillment of the Requirements for the Degree Master of Science in Biomedical Engineering By: Bryan A. Medina Lab Partner: Harvey Yamada May 2025 Department of Biomedical Engineering - San Jose State University Technical Advisor: Dr. Alessandro Bellofiore - Associate Professor and Chair - Department of Biomedical Engineering Reading Committee: Dr. Alessandro Bellofiore, Dr. Patrick Jurney 2 Table of Contents 1. Introduction 3 2. Literature Review 7 2.1. Cardiovascular Function 7 2.2. Compliance 8 2.3. Resistance 10 2.4. Hypotension & Hypertension 11 2.5. Mock Circulatory Loop Research 13 2.6. Sylgard 184 and Photo Image Velocimetry (PIV) 17 3. Objective 18 4. Materials & Methods 20 4.1. Materials 20 4.2. Methods 24 5. Results 28 6. Discussion 50 6.1. MCL Data: Acrylic Test Section 50 6.2. Sylgard Molds & PIV 54 6.3. MCL Data: Sylgard Test Section 57 6.4. MCL Data & Physiological Data 58 6.5. MCL Compliance Chamber Selection 62 7. Conclusion 72 8. References 74 9. Appendix 80 3 1. Introduction According to Mathers, Murray, and the World Health Organization, cardiovascular disease (CVD) is the number one contributing cause of death on a global scale9. The American Heart Association published a report showing that in 2020, 207.1 of 100,000 people died of heart disease and stroke33. The report also went on to mention that the total number of deaths due to CVD globally rose 18.71% from 2010 to a total of 19.05 million deaths in 202033. Cardiovascular diseases encompass a variety of heart and vascular diseases and conditions that include: arrhythmia, cardiomyopathy, pulmonary and aortic stenosis, atrial fibrillation, congenital heart disease, heart attack, heart failure, rheumatic heart disease, ischemic stroke, atherosclerosis, aneurysm, peripheral artery disease, and Raynaud\u2019s disease28. The CDC has referenced data from the National Center for Health Statistics and \u201cone person dies every 33 seconds in the United States from cardiovascular disease\u201d 4. For this reason, cardiovascular disease is and will continue to be an ongoing problem that requires several modes of treatment depending on the patient and the severity of the CVD. When dealing with CVD, one potential course of treatment is the transplantation of viable organs. In general, the donation or allocation of an organ for transplant is considered to be a cost-effective solution for CVD and especially for end-stage organ failure. However, the scarcity of organs for transplantation limits the availability", "this reason, cardiovascular disease is and will continue to be an ongoing problem that requires several modes of treatment depending on the patient and the severity of the CVD. When dealing with CVD, one potential course of treatment is the transplantation of viable organs. In general, the donation or allocation of an organ for transplant is considered to be a cost-effective solution for CVD and especially for end-stage organ failure. However, the scarcity of organs for transplantation limits the availability of this treatment, with an average of 15-30% of patients dying while being on a waitlist16. In particular, the number of transplants in 2018 in both the US and Europe reached an average of \u201c80 people receiving a transplant each day\u201d16. Despite the \u201clegislative, regulatory, and humanitarian services'' provided in countries such as the United States and Europe, there still remains a myriad of obstacles for organ transplantation besides the scarcity of organs16. These can include ethical concerns, religious beliefs, lack of consent or 4 understanding, race, as well as the condition of the organ for transplantation16. Alternative solutions to treating CVD include the use of artificial organs or long- and short-term cardiovascular assist devices (CADs)36. These devices include ventricular assist devices (VADs), total artificial hearts (TAHs), prosthetic heart valves (PHVs), and intra-aortic balloon pumps (IABPs)36. VADs are devices intended to reproduce \u201cthe native cardiac cycle and generate pulsatile flow\u201d through the use of a diaphragm and a heart valve12. Unfortunately, the size and weight of early VADs decreased the mobility of patients who used these devices12. VADs can further be separated into 3 categories based on the placement of the device in the heart. Of these three types of ventricular assist devices, there are right ventricular assist devices (RVADs), left ventricular assist devices (LVADs), and biventricular assist devices (BiVADs)12. TAHs, unlike VADs, completely replace heart function over long periods of time with the intent to treat end-stage congestive heart failure (CHF)12. A TAH functions by pneumatically pumping blood to the body through 4 mechanical valves and is made from polyurethane due to the material\u2019s tensile properties23,24. The driver for the TAH is typically placed on the bedside of the patient but a portable version of the driver can be worn by the patient as a backpack that weighs approximately 14 pounds23. PHVs are meant to replace the native valve function by preventing reverse flow of blood into the atria", "failure (CHF)12. A TAH functions by pneumatically pumping blood to the body through 4 mechanical valves and is made from polyurethane due to the material\u2019s tensile properties23,24. The driver for the TAH is typically placed on the bedside of the patient but a portable version of the driver can be worn by the patient as a backpack that weighs approximately 14 pounds23. PHVs are meant to replace the native valve function by preventing reverse flow of blood into the atria or ventricles during ventricular systole and diastole, respectively13. PHVs are also separated into two types, mechanical heart valves (MHVs), and bioprosthetic heart valves (BHVs)5. MHVs consist of \u201cthree major components: an occluder, an occluder restraint, and a sewing ring\u201d while BHVs consist of \u201cthree xenograft tissue leaflets\u201d that are either porcine or bovine21. Similar to MHVs, some BHVs can also be composed of a cloth/metallic support 5 structure when surgically implanted21. There are several factors that determine whether a patient receives a MHV or a BHV. Some include age, where older patients might receive a BHV due to their tolerance of anticoagulant medication and younger patients might receive a MHV since there is an inverse correlation of tissue deterioration with age21. Another critical consideration is the use of a surgical or transcatheter valve replacement method21. On the other hand, IABPs are \u201cinternal counterpulsation devices that are placed inside the descending aorta\u201d and are a ubiquitous mechanical support system for failing hearts12. IABPs function through inflation of a balloon during ventricular diastole in order to increase pressure, blood flow, and perfusion12. After this, the balloon is deflated in order to \u201cinduce reduced cardiac afterload and enhanced cardiac output\u201d12. The timing of the inflation and deflation of the balloon is crucial to the function of this device, and as a result, controllers have been added to IABPs so that settings can be modified from one patient to another12. Overall, these CADs possess great hemodynamic performance without considerable side effects. However, the side-effects are unavoidable and some, such as hemolysis and thrombosis, can cause CAD malfunction, injury, and potentially even death36. Consequently, intense testing, verification, and validation of CAD reliability, safety, and efficiency is necessary in order for these devices to be approved by administrations such as the FDA in the US36. A few of these tests can include in vivo experiments on animals or patients, but these are typically cost-intensive especially when", "great hemodynamic performance without considerable side effects. However, the side-effects are unavoidable and some, such as hemolysis and thrombosis, can cause CAD malfunction, injury, and potentially even death36. Consequently, intense testing, verification, and validation of CAD reliability, safety, and efficiency is necessary in order for these devices to be approved by administrations such as the FDA in the US36. A few of these tests can include in vivo experiments on animals or patients, but these are typically cost-intensive especially when being used to iterate the device being tested36. Instead, the use of in vitro experiments that rely on the use of a mock circulatory loop (MCL) system is a cost-effective approach for the early stage of development and optimization of CAD products36. There are three categories of MCLs: mechanical, numerical, and hybrid36. 6 Figure 1. Mechanical Mock Circulatory Loop Diagram (M-MCL)20 According to current literature, a majority of research is performed using mechanical MCLs (M-MCL)36. Furthermore, utilizing M-MCLs can not only be directly applied to the testing of different CADs, but it can also provide the validation of numerical MCLs (N-MCL)36. M-MCLs are capable of replicating pulmonary circulation, coronary circulation, renal circulation, as well as human physiological parameters such as cardiac output (CO), heart rate, blood pressure, compliance, and resistance36. CO is known as the quantity of blood pumped by the heart in one minute and is the product of heart rate as well as stroke volume, which is the volume of blood ejected per heartbeat35,14. CO is measured in units of liters per minute and heart rate is measured as the number of beats per minute35. Furthermore, compliance is defined as \u201ca measure of arterial distensibility\u201d, meaning how stiff vasculature is and this can be measured through a change in volume over a change in pressure10. On the other hand, resistance, in terms of pulmonary vasculature, is a product of vascular length and viscosity over the radius of the 7 vessel10. As a result, knowledge of these physiological parameters is crucial when utilizing MCLs for CAD research. It is even possible for a combination of CADs to be used in the M-MCL to be tested under varying physiological conditions36. As a result, M-MCLs are a great tool suitable for the testing of CADs, which in turn can help target specific patients as a result of its multi-functionality36. The M-MCL in the CardioLab at SJSU requires adjustments to the loop", "7 vessel10. As a result, knowledge of these physiological parameters is crucial when utilizing MCLs for CAD research. It is even possible for a combination of CADs to be used in the M-MCL to be tested under varying physiological conditions36. As a result, M-MCLs are a great tool suitable for the testing of CADs, which in turn can help target specific patients as a result of its multi-functionality36. The M-MCL in the CardioLab at SJSU requires adjustments to the loop in order for it to replicate physiological conditions. As a result, this project will deal with diagnosing areas of improvement by characterizing flow within the loop, pressure in the line/across the valve, flow rate in the valve, as well as the overall M-MCL. To this effect, Vivitro, the manufacturer of the pump within the MCL, has provided a number of engineering hours that will be used throughout this project in order to get the MCL properly functioning. Verification of this functionality will be explained further in the materials and methods section. 2. Literature Review 2.1. Cardiovascular Function The heart is one of many critical organs present in the human body. It is a structure composed of muscle tissue that functions as the driving force for blood flow and it is designed to transport and circulate blood and essential nutrients throughout the circulatory system. This muscular tissue, also known as the myocardium, contracts to remove blood within the heart and out through the arteries and then relaxes to allow for new blood to flow back in from the veins21. The relaxed state of the heart is known as diastole, where blood is free to flow from the left atrium and fill into the left ventricle21. On the other hand, systole is the state when additional blood is pushed into the heart, eventually causing an increase in pressure and contraction of the heart to push out the blood into the arteries21. 8 In order to ensure that blood flow remains in one direction, a series of mechanisms, or heart valves, are present in each of the chambers of the heart and function by opening and closing due to changes in pressure experienced during myocardial contraction and relaxation21. Specifically, the heart contains four of these heart valves and each are known as the aortic valve, mitral valve, pulmonic valve, and tricuspid valve21. Because of their anatomical location, the tricuspid and mitral heart valves", "In order to ensure that blood flow remains in one direction, a series of mechanisms, or heart valves, are present in each of the chambers of the heart and function by opening and closing due to changes in pressure experienced during myocardial contraction and relaxation21. Specifically, the heart contains four of these heart valves and each are known as the aortic valve, mitral valve, pulmonic valve, and tricuspid valve21. Because of their anatomical location, the tricuspid and mitral heart valves are grouped together as atrioventricular valves whereas the pulmonic and aortic valves are grouped together as the semilunar valves21. Furthermore, the atrioventricular valves are responsible for the prevention of retrograde blood flow into the atria during ventricular systole while the semilunar valves are responsible for the prevention of retrograde blood flow into the ventricles during diastole21. 2.2. Compliance Of the key parameters to consider for utilizing an MCL, compliance is one factor that has a higher criticality for MCLs. As mentioned in Wei-Xu et al.\u2019s article, compliance is one physiological parameter that MCLs can replicate and is defined as the stiffness of vasculature, and is related to distensibility36,10. In a clinical setting, compliance can also facilitate the sorting of patients depending on their condition10. Generally, pulmonary circulation is simplified as a system with high flow and low pressure, meaning it can respond well to large volume changes but possesses a \u201climited contractile reserve\u201d to match the growth of afterload10. The use of right heart catheterization enables the calculation of compliance based on the obtained hemodynamic measurements10. From this calculation, Ghio et al. found that as the value of pulmonary compliance decreases, the pressure per stroke volume increases10. Shaun Gregory (2009) found a mock circulation loop that was based on the control of elastance, which is \u201cdefined as the instantaneous ventricular pressure versus ventricular 9 volume\u201d11. The value of compliance in this loop was manually adjusted with values of 1.37 mL/mmHg for arterial compliance and 6.74 mL/mmHg for pulmonary venous compliance11. The conclusions of the study that used this MCL were that with a simulated systemic circulation, the authors were able to control the contractility of the ventricle11. Furthermore, Gregory goes on to state that vein compliance is roughly 24 times greater than the compliance of arteries11. Another study in this article attained compliance by setting a value of \u201c10 mL/mmHg for the systemic venous chamber, 1 mL/mmHg for the pulmonary", "for arterial compliance and 6.74 mL/mmHg for pulmonary venous compliance11. The conclusions of the study that used this MCL were that with a simulated systemic circulation, the authors were able to control the contractility of the ventricle11. Furthermore, Gregory goes on to state that vein compliance is roughly 24 times greater than the compliance of arteries11. Another study in this article attained compliance by setting a value of \u201c10 mL/mmHg for the systemic venous chamber, 1 mL/mmHg for the pulmonary arterial chamber, and 5 mL/mmHg for the systemic arterial chamber\u201d by using a trapped volume air that was above the fluid for each chamber mentioned previously11. Elastin and collagen fibers play an important role in the way vasculature reacts mechanically to forces. Particularly, the wave-like and unorganized configuration of fibers when unpressurized results in a non-linear pressure-diameter curve as shown below3. Figure 2. Pressure-Diameter Curve of Elastin & Collagen3 As noted in the figure, collagen dominates at high pressures (above 120 mmHg) and diameters whereas elastin dominates at low pressures (below 80 mmHg), with the physiological range for pressure ranging between 80 and 120 mmHg3. Furthermore, the Young\u2019s modulus of elasticity 10 for elastin ranges from 0.3-1 MPa whereas the modulus for collagen is 100 MPa3. One thing to note is that blood vessels \u201care much stronger in the fiber direction than perpendicular to it\u201d3. Their viscoelastic properties with induced stress cause it to decrease over time, described as stress relaxation3. Despite this characterization of blood vessels, it is actually very difficult to replicate the vessel\u2019s mechanical properties in vessel substitutes3. However, the collection of this data will enable the development of substitutes that actually work. Compliance, specifically arterial compliance, is responsible for influencing cardiac load32. Decreases in arterial compliance are \u201cassociated with numerous physiological states and pathological processes\u201d including organ damage \u201cthrough excessive penetration of pulsatility\u201d32. In turn, an increase of pulsatile stress can contribute to endothelial dysfunction, coronary atherosclerosis, acute coronary heart syndrome, and atherosclerotic plaque ruptures32. As a result, it is important to be able to measure or even estimate the value of compliance through direct and indirect methods32. Direct methods are considered invasive due to the mechanical stress that is applied on the arteries in order to obtain their mechanical properties32. The indirect methods rely on blood pressure analysis, distension, and changes in flow32. Diving deeper into these, indirect methods can be split into 3 groups:", "atherosclerosis, acute coronary heart syndrome, and atherosclerotic plaque ruptures32. As a result, it is important to be able to measure or even estimate the value of compliance through direct and indirect methods32. Direct methods are considered invasive due to the mechanical stress that is applied on the arteries in order to obtain their mechanical properties32. The indirect methods rely on blood pressure analysis, distension, and changes in flow32. Diving deeper into these, indirect methods can be split into 3 groups: local, systemic, and surrogate indexes32. The difference between local and systemic methods is that local focuses on a selected part of the artery whereas systemic focuses on the entire arterial tree and consists of analytical and numerical sub methods32. On the other hand, surrogate indexes are representative of parameters that are influenced by compliance32. 2.3. Resistance There is an inverse relationship between compliance and resistance10. When arteries are able to expand during systole and return to their original configuration during diastole, this 11 means that the arteries are able to accumulate a blood volume and subsequently release it, generating a continuous blood flow throughout the cardiac cycle10. On top of that, pressure oscillations are reduced such that the diastolic pressure in the pulmonary artery decreases drastically less than in the right ventricle10. To expand upon the concept of resistance, resistance is the \u201camount of force exerted on circulating blood by the vasculature of the body\u201d34. The three factors that determine resistance are blood vessel length, vessel diameter, and blood viscosity34. It is difficult to obtain accurate measurements of resistance, which is why resistance is instead estimated with blood pressure and cardiac output34. Specifically, the blood pressure can be used to calculate the mean arterial pressure (MAP) which can then be plugged into the equation where MAP equals cardiac output times total peripheral resistance (TPR)34. Damaged vasculature can contribute to an increase in resistance which ends up damaging the vasculature even further or potentially preventing the flow of blood34. Resistance is also a useful clinical parameter when dealing with patients whose blood pressure is significantly high or low34. 2.4. Hypotension & Hypertension In the sections for compliance and resistance, it was mentioned that increased and decreased pressure can lead to numerous physiological states. Two of these in particular are hypotension and hypertension. Hypertension is the highest risk factor for morbidity, mortality, and is associated with a higher risk of CVD25. Hypertension", "further or potentially preventing the flow of blood34. Resistance is also a useful clinical parameter when dealing with patients whose blood pressure is significantly high or low34. 2.4. Hypotension & Hypertension In the sections for compliance and resistance, it was mentioned that increased and decreased pressure can lead to numerous physiological states. Two of these in particular are hypotension and hypertension. Hypertension is the highest risk factor for morbidity, mortality, and is associated with a higher risk of CVD25. Hypertension is defined as possessing persistent high blood pressure in the arteries25. In terms of mmHg units, hypertension occurs when systolic blood pressure, diastolic blood pressure, or both exceed 140 mmHG systolic and 90 mmHG diastolic25. Despite that, risk of CVD increases as soon as blood pressure reads 115/75 mmHg, which is considered part of the 12 normotensive range25. More often than not, individuals with hypertension do not realize they have the condition and consequently do not receive adequate or any treatment at all for it25. Hypertension is affected and influenced by environmental and pathophysiological factors and so evaluation of patients through their blood pressure measurements, evidence of organ damage, presence of comorbidities, and their predicted risk of atherosclerotic CVD is important when considering courses of treatment25. These treatments can include dietary changes and increased exercise or physical activity in order to lower blood pressure and prevent hypertension25. Consequently, hypertension is an easily preventable risk factor for CVD25. However, 3.5 billion adults possess \u201cnon-optimal\u201d systolic blood pressure levels that are greater than 110-115 mmHg and 874 million have systolic blood pressure levels that are greater than 140 mmHg25. Some causes of hypertension include: renal parenchymal disease, renal artery stenosis, primary aldosteronism, pheochromocytoma, and cushing syndrome25. If dietary or lifestyle changes do not help with lowering blood pressure, the use of low-dose pharmacological therapy can also be effective in lowering blood pressure and preventing hypertension25. Hypotension on the other hand, is the other end of pressure values. In this case, a pressure reading of 90/60 mmHg would indicate hypotension, where there is not enough force of blood against the vessel wall5. A few common causes of low blood pressure include allergic reactions, dehydration, standing up too quickly, or a side effect from a prescription medication5. Hypotension, as opposed to hypertension, is a relatively benign condition due to its asymptomatic nature25. However, if left untreated in severe cases, there is a potential risk", "other end of pressure values. In this case, a pressure reading of 90/60 mmHg would indicate hypotension, where there is not enough force of blood against the vessel wall5. A few common causes of low blood pressure include allergic reactions, dehydration, standing up too quickly, or a side effect from a prescription medication5. Hypotension, as opposed to hypertension, is a relatively benign condition due to its asymptomatic nature25. However, if left untreated in severe cases, there is a potential risk of death or multi-organ failure25. 13 2.5. Mock Circulatory Loop Research Liu et al. constructed a mock circulatory loop intended for performance testing of LVADs19. The versatility of the MCL in its simulation of various physiological conditions allowed Liu et al. to try and establish an MCL for bench testing of continuous flow (CF) LVADs19. In particular, their MCL design incorporated several components in the loop alongside the LVAD as seen in the figure below19. Figure 3. Liu et al.\u2019s MCL Design for CF-LVAD Performance Testing19 One aspect to note was that Liu et al. employed the use of three airtight tanks whose purpose was to simulate pulmonary, systemic venous, and systemic arterial compliances19. The values for compliance were found using the below equation. Figure 4. Equation for Compliance19 14 As explained in the article, C is the compliance of the tank, Vair is the volume of air within the tank, Pair is the absolute pressure of the tank with air, Vtank is the volume of the tank, Atank is the cross-sectional area of the tank, Pfluid is the absolute pressure of the fluid in the tank, hfluid is the height of the fluid in the tank, p is the fluid density, and g is acceleration due to gravity19. The protocol for testing included the following five scenarios for pathological and activity level: healthy/rest, healthy/sleep, healthy/exercise, congestive heart failure (CHF)/rest, and partially recovered CHF/rest19. From these five testing scenarios, Liu et al. found that in comparison to the operating parameters for the MCL and clinically obtained data, the output of the MCL matches values for CO and left-ventricular end diastolic pressure (LVED), particularly for the CHF/rest and partially recovered CHF/rest conditions19. There was only a difference of -1.8 mmHg between left ventricular pressure and systolic aortic pressure for the healthy/rest condition19. One area of improvement for the MCL used in this study is developing a method to replicate the Starling", "et al. found that in comparison to the operating parameters for the MCL and clinically obtained data, the output of the MCL matches values for CO and left-ventricular end diastolic pressure (LVED), particularly for the CHF/rest and partially recovered CHF/rest conditions19. There was only a difference of -1.8 mmHg between left ventricular pressure and systolic aortic pressure for the healthy/rest condition19. One area of improvement for the MCL used in this study is developing a method to replicate the Starling response since there was a finite volume variation in the diaphragm19. Farahmand et al. developed a non-clinical MCL database in order to dynamically characterize CO monitoring systems that are pressure-based7. Because there is a need to continuously monitor parameters such as CO and stroke volume and there are no consensus standards for the assessment of CO monitoring systems, the resulting database can be used as a tool to evaluate the dynamic attributes of said CO monitoring systems7. The MCL utilized for this study, shown in Figure 5, was designed to mimic three categories of hemodynamic values7. These included hyperdynamic, normovolemic, and cardiogenic shock7. 15 Figure 5. Schematic diagram of Farahmand et al\u2019s MCL7 Farahmand et al. assumed peripheral resistance was negligible for the intended use of the MCL since muscular arteries in the upper limb are not expected to be affected by the 3 categories of hemodynamic values7. Once the setup of the physical MCL components was completed, a central pressure to peripheral pressure transfer function based on each hemodynamic state was determined through sinusoidal frequency sweeps that ranged from 0.5 to 4.5 hertz7. The nine datasets collected using the MCL included CO resolution, CO response time to a known rapid CO change, as well as stroke volume resolution7. Furthermore, a simulation of the different levels of stroke volume variation was carried out by setting two respiration rates of 12 and 20 cycles/minute7. At these settings, Farahmand et al. was able to calculate the stroke volume variation by, using the MCL flow data, looking at the delta of stroke volume values at each respiration cycle, and taking the mean of stroke volume variations across all respiration cycles in a given step7. A few limitations of this study were that in its initial design, the MCL built by Farahmand et al. utilized a piston pump as the driving force for flow generation but this did not allow for the MCL to", "able to calculate the stroke volume variation by, using the MCL flow data, looking at the delta of stroke volume values at each respiration cycle, and taking the mean of stroke volume variations across all respiration cycles in a given step7. A few limitations of this study were that in its initial design, the MCL built by Farahmand et al. utilized a piston pump as the driving force for flow generation but this did not allow for the MCL to replicate changes in cardiac output and stroke volume variation7. As a 16 result, the piston pump was replaced with a positive displacement gear pump but despite this, the MCL is still unable to capture the nonlinear relationship between compliance and pressure7 . In addition to this, the resulting database that was created demonstrates inadequacies due to inherent variation from one individual to another, specifically relating to arterial property variation7. Petrou et al. designed a hybrid mock circulation loop (H-MCL) that iterates upon a previous design by increasing the number of reservoirs to four and adapting the numerical model accordingly in order to test active and passive CADs, such as BiVADs and TAHs26. The H-MCL discussed by Petrou et al., in addition to its four reservoirs, also employs the use of ultrasound flow probes, pressure transducers, reflux pumps, a vacuum pump, vacuum chamber, solenoid valve, an inlet valve, and outlet valves as shown in the schematic below26. Figure 6. H-MCL Design Schematic by Petrou et al. Definitions: PR, pressure reservoir; RP, reflux pump; FP, flow probe; PT, pressure transducer; VC, vacuum chamber; VP, vacuum pump26 17 Petrou et al. set up three test cases for their H-MCL. Each test case utilized a different numerical model, however, at its base, these models were separated into four components: a left heart, systemic circulation, pulmonary circulation, and right heart section26. The first test case simulated the circulation of an adult patient, with reduced right and left ventricle ejection fractions at thirty and twenty percent respectively as well as an HVAD each to support each ventricle26. The second test case utilized the same setup as the first, only without the simulation of decreased right and left ventricle ejection fractions26. Instead, the HVADs were set up to function as a TAH, with pulmonary vascular resistance increased to 0.5 mmHg*s/mL26. Lastly, the third test case focused on total cavopulmonary connection (TCPC), with clinical conditions simulating two exercise", "and left ventricle ejection fractions at thirty and twenty percent respectively as well as an HVAD each to support each ventricle26. The second test case utilized the same setup as the first, only without the simulation of decreased right and left ventricle ejection fractions26. Instead, the HVADs were set up to function as a TAH, with pulmonary vascular resistance increased to 0.5 mmHg*s/mL26. Lastly, the third test case focused on total cavopulmonary connection (TCPC), with clinical conditions simulating two exercise levels, one at rest and another of three metabolic equivalents of tasks (METs)26. Although targeted towards BiVADs and TAHs, the H-MCL developed could also be adapted to research for heart valves, both mechanical and artificial26. Ultimately, Petrou et al.\u2019s research demonstrated that a H-MCL possesses great versatility in testing, especially when trying to simulate various clinical conditions26. However, Petrou et al. acknowledged that there is a lack of validation for numerical models in correlation to clinical situations, as additional validations are required with each modification made to the model26. 2.6. Sylgard 184 and Photo Image Velocimetry (PIV) The use of imaging measurement methods for in-vitro studies, such as PIV, facilitate the investigation of flow dynamics, particularly fluid velocity fields, in vasculature2,8. PIV involves the use of lasers to detect the displacement of particles in fluid along with a high-speed camera8. PIV-related studies often involve the use of silicone polydimethylsiloxane (PDMS) Sylgard 184, with polyvinyl alcohol being used to dissolve 3D printed mold cores from the silicone8. Sylgard 18 184 in particular is highly versatile during its curing process, meaning various shapes can be achieved, allowing researchers to create specific test models1. The success of PIV studies is heavily reliant on accurately obtaining the displacement of the particles illuminated by the laser through the high-speed camera. As a result, one critical factor that ensures accuracy of measurements is the refractive index (RI) of not only the fluid, but also of the material used to house said fluid1. It is a requirement that the fluid should be as optically clear as possible and possess the same RI as the material of the test section/model because for any difference in RI, no matter how small, a certain degree of distortion can occur and invalidate any measurements collected1. In addition to this, the fluid itself should mimic the fluid properties of blood1. In most cases, a water-glycerol blood analog with a 60/40 ratio can", "material used to house said fluid1. It is a requirement that the fluid should be as optically clear as possible and possess the same RI as the material of the test section/model because for any difference in RI, no matter how small, a certain degree of distortion can occur and invalidate any measurements collected1. In addition to this, the fluid itself should mimic the fluid properties of blood1. In most cases, a water-glycerol blood analog with a 60/40 ratio can achieve a RI of 1.39, which is lower than the RI of Sylgard, 1.4141. In order to try and raise the RI of the water-glycerol mixture, sodium iodide (NaI) can be added since it does not change the viscosity of the fluid1. Brindise et al. tested six alternative Newtonian fluid formulations that consisted of additives such as urea, xylitol, sodium iodide, sodium chloride, glycerol, as well as xanthan gum1. The refractive indices for all samples were obtained using a refractometer1. In addition to this, the viscosity and density for each of the two- and three- component Newtonian fluids were determined1. Ultimately, Brindise et al. concluded that urea is an effective, inexpensive additive capable of adjusting the RI of blood analog mixtures with little effect to density and viscosity of the fluid, making it a great selection for PIV studies1. 3. Objective The objectives of this project can be separated into two iterative topics. Overall, the MCL is intended to replicate cardiovascular flow. In doing so, by simulating various physiological conditions, the MCL can be utilized to gather pressure and flow profiles and also evaluate the 19 effectiveness of mechanical heart valves, amongst the various types of cardiovascular assist devices. By ensuring the data collected at various settings agrees with the physiological data, the MCL can be utilized as a tool to test and validate these CADs as well as investigate diseases and conditions specific to a patient. The first objective is to revise and validate an experimental benchtop system, in this case the MCL, in order to replicate cardiac blood flow and perform testing of mechanical heart valves. There are several key components within the MCL that can be modified, re-designed, or eliminated in order to achieve physiological blood pressures and flow. From the initial scope for the project, this could include designing a new compliance chamber that is able to operate at various pressures and does not crack due", "is to revise and validate an experimental benchtop system, in this case the MCL, in order to replicate cardiac blood flow and perform testing of mechanical heart valves. There are several key components within the MCL that can be modified, re-designed, or eliminated in order to achieve physiological blood pressures and flow. From the initial scope for the project, this could include designing a new compliance chamber that is able to operate at various pressures and does not crack due to stress at the inlet and outlet. The intent is to make the chamber modular so that the volume of air can be controlled rather than increasing the volume of blood analog in the MCL. In addition to this, the use of engineering hours provided by Vivitro has resulted in a need for the test section to be re-designed so that the pressure readings are collected 1 diameter length upstream and 3 diameter lengths downstream. The tubing in the MCL can also be shortened in an effort to reduce the elevated pressure that persisted in the system and resulted in saturation of the pressure readings. Upon completion of this objective, the second objective of the project is to re-design the test section using Sylgard 184 as the build material so that the MCL can later be utilized for PIV measurements. In particular, this new test section will possess geometries reflecting those of the aortic sinus so that when the heart valve is tested in the MCL, a flow similar to the one across an implanted replacement valve can be achieved. Overall, the goal of this project is to obtain flow and pressure profiles from the MCL that are in agreement within 10-15% of the physiological profiles for various conditions such as rest, 20 exercise, and hypertension. In general, a healthy heart rate falls between 60 to 100 beats per minute, whereas blood pressure values fall between 90/60 mmHg and 120/80 mmHg, with a stroke volume between 50 to 100 mL per heart beat,30,2,5. 4. Materials & Methods 4.1. Materials The work performed towards this project will occur in Room 233J in the Charles W. Davidson College of Engineering building at San Jose State University. For safety requirements and security of the equipment in the lab, one individual cannot work on the MCL alone in the room but must be accompanied by a second individual. The MCL consists of various components", "mmHg, with a stroke volume between 50 to 100 mL per heart beat,30,2,5. 4. Materials & Methods 4.1. Materials The work performed towards this project will occur in Room 233J in the Charles W. Davidson College of Engineering building at San Jose State University. For safety requirements and security of the equipment in the lab, one individual cannot work on the MCL alone in the room but must be accompanied by a second individual. The MCL consists of various components that are critical to the function or data analysis portions of the project. One such component is the test section. The test section is a housing made out of acrylic, as seen in Figure 7, or, in the future, from Sylgard 184 that will contain the bi-leaflet heart valves intended to be tested within the MCL. The acrylic version of the test section possesses a basic inner cylindrical geometry whereas the Sylgard 184 test section will possess geometry reflecting that of the aortic sinus. As mentioned in the literature review, because the MCL can be used to perform PIV measurements, it is important that the refractive indices of the blood analog and the test section are equal to each other so that there is no refraction of the laser when taking velocity measurements. Acrylic possesses a refractive index of 1.49 whereas Sylgard 184 has a refractive index of 1.4141. 21 Figure 7. Acrylic test section from the SJSU CardioLab Mechanical heart valves are also another critical component of the MCL. For the purposes of this project, a bi-leaflet valve, Figure 8, from St. Jude will be tested in the MCL. These valves will be placed in the test section, and for the Sylgard test section, the mechanical valves will be located in the anatomical position of their physiological counterparts. As such, the function of the mechanical valves will also be to control the flow of the blood analog in one direction and prevent retrograde flow. Figure 8. Bi-leaflet Mechanical Heart Valve from St. Jude Another critical component of the MCL is the blood analog. A water-glycerol mixture with a 60/40 ratio is used for MCL experiments, this project will utilize distilled water as the blood analog during preliminary testing as a cost-savings alternative as well as to minimize exposure of the MCL and its components from the glycerol\u2019s stickiness. Compared to the water-glycerol mixture, distilled water does possess a RI", "direction and prevent retrograde flow. Figure 8. Bi-leaflet Mechanical Heart Valve from St. Jude Another critical component of the MCL is the blood analog. A water-glycerol mixture with a 60/40 ratio is used for MCL experiments, this project will utilize distilled water as the blood analog during preliminary testing as a cost-savings alternative as well as to minimize exposure of the MCL and its components from the glycerol\u2019s stickiness. Compared to the water-glycerol mixture, distilled water does possess a RI of 1.43 versus a RI of 1.39 for 60/40 water-glycerol mixtures1. Once the MCL is completed and confirmation runs determine the system is functional, the DI water will be switched out for the water-glycerol mixture. In addition to this, Tygon PVC tubing is connected to the beginning and end of the MCL and contains the volume of fluid media used for the system. The purpose of this tubing is to mimic the vasculature around the heart and as such, must not possess any sharp turns in order to 22 not create disruptions in the flow of distilled water across the loop. As mentioned previously, the length of this tubing across the MCL will be shortened in order to reduce elevated pressure. In order to transport the blood analog across the MCL from start to finish, the Vivitro SuperPump, as seen in Figure 9, manufactured by VivitroLabs will be used. This pump is responsible for generating the pulsatile flow in the MCL just as the heart would across the cardiovascular system. The Vivitro pump is capable of achieving cycle rates of 3 to 200 beats per minute (BPM), as well as displacement volumes from 0 to 180 mL. By controlling the cycle rates, or heart rate, as well as the displacement volumes, or stroke volume, the cardiac output of the pump can be calculated through the equation where cardiac output equals heart rate times stroke volume. Figure 9. VivitroLabs SuperPump A compliance chamber, or reservoir, is another component of the MCL. The purpose of this chamber is to replicate the elasticity of vasculature through a trapped volume of air. Two designs for this chamber will be considered, one made of acrylic and another 3D printed from resin. Since compliance is equal to the change in volume over the change in pressure of the fluid within the MCL, the volume of air in the chamber, or the height of blood analog in", "compliance chamber, or reservoir, is another component of the MCL. The purpose of this chamber is to replicate the elasticity of vasculature through a trapped volume of air. Two designs for this chamber will be considered, one made of acrylic and another 3D printed from resin. Since compliance is equal to the change in volume over the change in pressure of the fluid within the MCL, the volume of air in the chamber, or the height of blood analog in the chamber, can be adjusted accordingly to obtain the desired values specified in the methods section. Aside from the compliance chamber, a resistance valve is also incorporated into the MCL. The purpose of this valve is to regulate the pressure in the MCL, allowing various blood pressure conditions to be simulated. Although it is difficult to accurately measure the value of 23 resistance, it possesses an inverse relationship with compliance, therefore any changes to the compliance of the MCL as a result of the increase or decrease in resistance will reflect the respective resistance changes. In order to facilitate the data collection in the MCL, four additional components were integrated into the system. The first of these components was a flow sensor, which is responsible for outputting the flow rate within the MCL. Following this, three pressure sensors consisting of two gauge pressure sensors, one placed before and a second placed after the mechanical heart valve, and a third differential pressure sensor was added to the MCL. These pressure sensors are responsible for reading the pressures before and after the valve location and the differential pressure across the test section respectively. In the model of the test section, the gauge pressure sensor located upstream of the heart valve will measure the left ventricular pressure whereas the gauge pressure sensor placed downstream of the valve will measure the aortic pressure. As mentioned in the objectives, the redesign of the test section to measure pressures one diameter in length upstream and three diameters in length downstream from the valve will directly impact the placement of the sensor probes for the differential pressure sensor. The location of the sensors were based on positions within the MCL that would facilitate more accurate pressure drop readings before and after the mechanical heart valve. Overall, these three sensors will allow for the construction of a flow and pressure profiles which can then be compared to physiological", "to measure pressures one diameter in length upstream and three diameters in length downstream from the valve will directly impact the placement of the sensor probes for the differential pressure sensor. The location of the sensors were based on positions within the MCL that would facilitate more accurate pressure drop readings before and after the mechanical heart valve. Overall, these three sensors will allow for the construction of a flow and pressure profiles which can then be compared to physiological profiles. To assist with the construction of these profiles, a computer with LabView software, paired with a National Instruments data acquisition device, and LabChart software will interpret the analog and electrical current signals from the sensors, transforming them into digital signals. 24 4.2. Methods In order to satisfy the first objective of the project, several key components of the MCL will be modified, added, or removed from the MCL setup that possesses the existing acrylic test section. Some of these components included the tygon tubing, the resistance valves, or even the compliance chambers. At the initial stage of the project, it was known that the MCL was unable to reproduce human physiological pressures. As such, several iterations of the MCL setup were tested, spanning over a time period of several months, to determine if the changes being made would generate physiological pressures within the MCL. Each time a certain iteration of the MCL was tested, data was collected from the pressure sensors through the National Instruments data acquisition device paired with LabView, or through the flow sensor through LabChart. Once physiological pressures were achieved with the existing MCL, the design of the Sylgard test section was made using Solidworks in order to start addressing the second objective of the project. By leveraging the technique for creating silicone models from Falk et al., a corresponding mold based on the dimensions of the Sylgard test section was also made in Solidworks8. Important design configurations for this test section included being able to fit the mechanical heart valve from St. Jude as well as possessing two access ports from which to take pressure measurements from. Figure 7 displays the resulting mold models as well as the model for the Sylgard test section. 25 Figure 7. Solidworks Assemblies for a). Sylgard Test Section b). Right-half Mold and c). Left-half Mold In order to connect the new Sylgard test section into the existing MCL setup,", "design configurations for this test section included being able to fit the mechanical heart valve from St. Jude as well as possessing two access ports from which to take pressure measurements from. Figure 7 displays the resulting mold models as well as the model for the Sylgard test section. 25 Figure 7. Solidworks Assemblies for a). Sylgard Test Section b). Right-half Mold and c). Left-half Mold In order to connect the new Sylgard test section into the existing MCL setup, a pair of custom barbed tube fittings were designed to house quarter-inch threaded rods that will compress the Sylgard test section together with the use of threaded nuts. These barbed tube fittings will be resin printed and will allow for simple installation into the MCL. The Sylgard molds will then be 3D printed using Polyvinyl Butyral (PVB) filament. Printing the molds with PVB allows for the molds to be dissolved in Isopropyl Alcohol (IPA), making it easy to remove the cured Sylgard test sections that are held inside the molds. Once the PVB mold is finished printing, the Sylgard 184 silicone would then be mixed and prepared. The purchased Sylgard 184 kits come with two containers, part A and part B, which need to be mixed at a 10:1 ratio. The amount of Sylgard needed to pour the molds was determined by identifying the mass properties of the molds made in Solidworks after changing the density of the model to match that of Sylgard 184, 1.03 g/cm3. Once identified, parts A and B of the Sylgard kit are to be mixed together for ten minutes, and then left to dry for approximately three days at room temperature. Once cured, the Sylgard PVB molds will then be placed into a sealed container that contains IPA as seen in Figure 8. The container also has an inlet and outlet that 26 connects it to a fluid pump which cycles the IPA around the container to speed up the dissolution process of the PVB. Figure 8. IPA Dissolving Tank with connected water pump In order to obtain the pressure and flow profiles from the MCL with the Sylgard test section that will be compared to physiological data, a three factor, three level DOE was created. The three factors in question are the following: stroke volume, heart rate, and resistance. Each of these factors will be varied at three different levels: low, mid, and", "the container to speed up the dissolution process of the PVB. Figure 8. IPA Dissolving Tank with connected water pump In order to obtain the pressure and flow profiles from the MCL with the Sylgard test section that will be compared to physiological data, a three factor, three level DOE was created. The three factors in question are the following: stroke volume, heart rate, and resistance. Each of these factors will be varied at three different levels: low, mid, and high. The setup can be found in the table below. Table 2. DOE Setup 27 In total, a three factor, three level DOE will result in 27 unique trials with each trial duration set to 10 seconds. Upon completion of the 27 trials, common statistical parameters such as mean, standard deviation, and minimum/maximum values will be calculated. Furthermore, the peak pressure and flow rates will be compared to human physiological data. When setting up the MCL, the specific test section, acrylic or Sylgard, will be connected using the existing tygon tubing and worm clamps. Once in place, the loop will be filled with fluid media through the reservoir tank and bubbles within the MCL loop will be evacuated by lifting/lowering portions of the MCL in order to guide the bubbles into the reservoir tank. A luer lock syringe can be connected to the Vivitro superpump attachment head in order to remove any bubbles that may be stuck in the pump itself. Furthermore, the acrylic and Sylgard test sections possess two pressure ports where the three pressure sensors will connect to, and any air bubbles that may have been introduced as part of the installation into the MCL will need to be expelled from there as well. The overall objective is to remove all the visible bubbles so that there is little to no interference when it is time to record the flow and pressure sensor readings. Once the filling of the MCL and evacuation of bubbles is complete, a flow sensor will be clamped to the 1\u201d tygon tube that is upstream of the installed test section. Then the appropriate heart rate waveform and stroke volume will be selected on the Vivitro Superpump\u2019s controller based on the experimental setup determined in Table 2. At this point, both pressure and flow data are collected concurrently for a time period of t=10 seconds per trial. The pump is then stopped after collecting", "the MCL and evacuation of bubbles is complete, a flow sensor will be clamped to the 1\u201d tygon tube that is upstream of the installed test section. Then the appropriate heart rate waveform and stroke volume will be selected on the Vivitro Superpump\u2019s controller based on the experimental setup determined in Table 2. At this point, both pressure and flow data are collected concurrently for a time period of t=10 seconds per trial. The pump is then stopped after collecting the three trials of data for the test condition, and is adjusted based on the parameters of the next test. 28 5. Results As mentioned in the methods section, several iterations to the MCL with the existing acrylic test section were developed, and at each iteration, data was collected to document the effect of the changes made to the MCL setup. From the beginning to the end of the project, the MCL in the CardioLab underwent a total of five changes in setup. The initial, or baseline, setup of the MCL that possessed issues with elevated pressures and saturation of the pressure sensors, is depicted in Figure 9. Figure 9. May 2024 MCL Setup This version of the MCL possessed one reservoir tank, a minimized overall loop length, a differential pressure sensor, a gauge pressure sensor upstream of the valve, and a brass needle valve whose adjustment knob was positioned upwards. From this version of the MCL, the MCL setup was then changed in June 2024. This second iteration of the MCL (Figure 10) possessed the following: two compliance chambers, one 29 reservoir, increased overall loop length, one differential pressure sensor, one gauge pressure sensor, and one resistance ball valve. Figure 10. June 2024 MCL Setup Expanding from the list above, of the two compliance chambers added to the MCL, one small chamber was positioned prior to the test section, and another large chamber was positioned after the test section. One reservoir still remained in this setup; however, the container was changed from a glass beaker to a plastic container. Other changes to the MCL were the removal of the needle valve and replacement with a ball valve located after the reservoir tank. In total, three test conditions were executed for this iteration of the MCL in June 2024. The experimental setup is outlined in Table 3 below, and primarily focuses on the resistance valve, with heart rate and stroke", "One reservoir still remained in this setup; however, the container was changed from a glass beaker to a plastic container. Other changes to the MCL were the removal of the needle valve and replacement with a ball valve located after the reservoir tank. In total, three test conditions were executed for this iteration of the MCL in June 2024. The experimental setup is outlined in Table 3 below, and primarily focuses on the resistance valve, with heart rate and stroke volume parameters set to 60 BPM and 70 mL/stroke respectively. For the resistance valve position, a value of 25 signifies that the ball valve was only 25% open, a value of 50 signifies that the ball valve was 50% open, and a value of 100 signifies that the valve was completely open. 30 Table 3. MCL June 2024 Experiment Test # HR SV Resistance Valve Position 1 60 70 25 2 50 3 100 Each test consisted of 3 individual trials that were then averaged together. From this experiment, the minimum, maximum, average, and standard deviation of the peak differential pressure values were collected. The peak differential pressure values ranged from an average of 4.98 mmHg at a resistance valve position of 25 to as high as 8.21 mmHg at a resistance valve position of 100. The peak left ventricular pressure values ranged from 109.72 mmHg at a resistance valve position of 25 to 105.47 mmHg at a resistance valve position of 100. The results are tabulated below in Table 4 and the peak average pressures for the differential and left ventricular pressures are shown graphically in Figure 11. The individual graphs for left ventricular pressure and differential pressure are located in Appendix A. Table 4. MCL June 2024 Pressure Readings Peak Differential Pressures (mmHg) Peak Left Ventricular Pressures (mmHg) Test # HR SV Resistance Valve Position Min Max Std. Dev. Avg. Min Max Std. Dev. Avg. 31 1 60 70 25 4.92 5.05 0.05 4.98 108.87 110.6 0.42 109.72 2 50 8.55 11.83 0.84 9.62 109.51 110.54 0.37 109.94 3 100 7.82 8.59 0.27 8.21 105.17 105.95 0.26 105.47 Figure 11. MCL June 2024 Peak Average Pressures The MCL underwent another change in September 2024. Now being the third iteration, this MCL shown in Figure 12, possessed the following components: three compliance chambers, one reservoir tank, another increase in loop length, two gauge pressure sensors, one differential pressure sensor,", "70 25 4.92 5.05 0.05 4.98 108.87 110.6 0.42 109.72 2 50 8.55 11.83 0.84 9.62 109.51 110.54 0.37 109.94 3 100 7.82 8.59 0.27 8.21 105.17 105.95 0.26 105.47 Figure 11. MCL June 2024 Peak Average Pressures The MCL underwent another change in September 2024. Now being the third iteration, this MCL shown in Figure 12, possessed the following components: three compliance chambers, one reservoir tank, another increase in loop length, two gauge pressure sensors, one differential pressure sensor, two ball valves, and a brass needle valve whose adjustment knob was positioned downwards. 32 Figure 12. September 2024 MCL Setup The most notable changes for this iteration of the MCL were the addition of the third compliance chamber, located after the second compliance chamber, and the addition of two more resistance valves, one ball valve and re-incorporating the needle valve. The second ball valve was positioned in between the second and third compliance chambers while the needle valve returned to its original position after the final compliance chamber. Lastly, the first compliance chamber was changed from a glass bulb to a glass cylinder with larger volume and the new gauge pressure sensor was placed after the valve to record the value of aortic pressure. In order to verify where this MCL iteration stood in relation to being able to achieve physiological pressures, a total of 8 tests were performed. This experimental setup, found in Table 5, focused on varying the resistance valve position of the needle valve as well as various stroke volume levels. Table 5. MCL September 2024 Experiment Test # HR SV Resistance Valve Position 33 1 70 50 50 2 100 3 70 60 50 4 100 5 70 70 50 6 100 7 70 80 50 8 100 As in the previous experiment, each test consisted of three trials and the trials were then averaged together. The minimum, maximum, average, and standard deviation of the peak pressures for the differential, left ventricular, and aortic pressures were calculated and tabulated in Table 6. The individual graphs for each test condition are located in Appendix B. Table 6. MCL September 2024 Pressure Readings Peak Differential Pressures (mmHg) Peak Aortic Pressures (mmHg) Peak Left Ventricular Pressures (mmHg) Test # HR SV Resistance Valve Position Min Max Std. Dev. Avg. Min Max Std. Dev. Avg. Min Max Std. Dev. Avg. 1 70 50 50 17.72 19.23 0.47 18.97 74.23", "peak pressures for the differential, left ventricular, and aortic pressures were calculated and tabulated in Table 6. The individual graphs for each test condition are located in Appendix B. Table 6. MCL September 2024 Pressure Readings Peak Differential Pressures (mmHg) Peak Aortic Pressures (mmHg) Peak Left Ventricular Pressures (mmHg) Test # HR SV Resistance Valve Position Min Max Std. Dev. Avg. Min Max Std. Dev. Avg. Min Max Std. Dev. Avg. 1 70 50 50 17.72 19.23 0.47 18.97 74.23 76.05 0.58 75.61 110.84 117.93 2.13 116.44 2 100 16.14 16.26 0.04 16.22 66.91 68.04 0.31 67.48 101.52 102.82 0.45 102.08 3 70 60 50 20.57 22.8 0.7 22.42 90.38 92.73 0.77 91.27 123.55 137.88 4.71 136.1 34 4 100 18.47 19.27 0.25 19.1 81.37 83.12 0.53 82 119.44 121.04 0.55 119.99 5 70 70 50 26.63 27.01 0.14 26.81 107.37 109.62 0.67 108.21 160.95 162.87 0.68 161.98 6 100 21.86 22.04 0.06 21.94 96.47 98.25 0.67 97.4 136.88 137.07 0.18 136.83 7 70 80 50 31.14 31.35 0.07 31.24 126.11 127.86 0.52 126.79 187.35 189.2 0.56 188.17 8 100 24.72 24.9 0.07 24.78 111.43 114.24 1.01 112.68 153.28 154.98 0.54 154.25 The peak average pressures for the differential, left ventricular, and aortic pressures were graphed and are shown in Figure 13. Figure 13. MCL September 2024 Peak Average Pressures In the early part of November 2024, there was no change made to the MCL, but a maintenance service was performed on the loop. In particular, the system was flushed out due to the presence of algae as well as tarnish coming off of the brass needle valve, and the tubes, 35 connectors, and fittings were cleaned. After the service maintenance there was one change made to the resistance valve , which was replaced with a 1/4\u201d nylon ball valve. In order to evaluate the effect of switching the brass needle valve out for the 1/4\u201d nylon ball valve, a total of 6 tests per resistance valve were performed. The experimental setups are shown in Table 7 and Table 8. Table 7. MCL November 2024 Experiment: Brass Needle Valve Table 8. MCL November 2024 Experiment: Nylon Ball Valve Test # HR SV Resistance Valve Position Valve Type 7 60 50 50 Nylon Ball 8 100 Nylon Ball Test # HR SV Resistance Valve Position Valve Type 1 60 50 50 Brass Needle 2 100 Brass Needle 3 60 50 Brass", "total of 6 tests per resistance valve were performed. The experimental setups are shown in Table 7 and Table 8. Table 7. MCL November 2024 Experiment: Brass Needle Valve Table 8. MCL November 2024 Experiment: Nylon Ball Valve Test # HR SV Resistance Valve Position Valve Type 7 60 50 50 Nylon Ball 8 100 Nylon Ball Test # HR SV Resistance Valve Position Valve Type 1 60 50 50 Brass Needle 2 100 Brass Needle 3 60 50 Brass Needle 4 100 Brass Needle 5 70 50 Brass Needle 6 100 Brass Needle 36 9 60 50 Nylon Ball 10 100 Nylon Ball 11 70 50 Nylon Ball 12 100 Nylon Ball After completion of the tests for each valve, the results were compiled and tabulated in Table 9 and Table 10. Each battery of tests consisted of three individual trials that were then averaged together. The peak average pressures for the left ventricular, differential, and aortic pressures graphed in Figure 14 and Figure 15. The individual graphs for each test are located in Appendix C. Table 9. MCL November 2024 Pressure Readings: Brass Needle Valve Peak Differential Pressures (mmHg) Peak Aortic Pressures (mmHg) Peak Left Ventricular Pressures (mmHg) Test # HR SV Resistance Valve Position Valve Type Min Max Std Dev. Avg. Min Max Std Dev. Avg. Min Max Std Dev. Avg. 1 60 50 50 Brass Needle 17.14 17.25 0.03 17.19 33.97 57.94 10.42 52.58 99.74 101.57 0.56 100.77 2 100 Brass Needle 16.1 16.84 0.25 16.6 56.11 57.06 0.35 56.59 55.38 98.65 18.72 88.92 3 60 50 Brass 20.61 20.98 0.11 20.85 67.2 67.84 0.19 67.4 119.44 121.64 0.77 120.21 37 Needle 4 100 Brass Needle 18.02 20.05 0.8 19.55 64.98 65.54 0.21 65.28 115.59 116.95 0.51 116.19 5 70 50 Brass Needle 22.09 24.97 1.19 24.23 65.19 76.9 4.94 74.13 140.12 142.13 0.61 141.09 6 100 Brass Needle 23.63 23.83 0.07 23.72 75.23 75.78 0.19 75.5 134.32 136.03 0.51 134.89 Table 10. MCL November 2024 Pressure Readings: Nylon Ball Valve Peak Differential Pressures (mmHg) Peak Aortic Pressures (mmHg) Peak Left Ventricular Pressures (mmHg) Test # HR SV Resistance Valve Position Valve Type Min Max Std Dev. Avg. Min Max Std Dev. Avg. Min Max Std Dev. Avg. 7 60 50 50 Nylon Ball 30.61 34.25 1.13 33.24 74.89 76.36 0.4 75.69 170.28 171.11 0.29 170.72 8 100 Nylon Ball 25.21 28.64 1.07 27.76 71.54 72.57", "75.5 134.32 136.03 0.51 134.89 Table 10. MCL November 2024 Pressure Readings: Nylon Ball Valve Peak Differential Pressures (mmHg) Peak Aortic Pressures (mmHg) Peak Left Ventricular Pressures (mmHg) Test # HR SV Resistance Valve Position Valve Type Min Max Std Dev. Avg. Min Max Std Dev. Avg. Min Max Std Dev. Avg. 7 60 50 50 Nylon Ball 30.61 34.25 1.13 33.24 74.89 76.36 0.4 75.69 170.28 171.11 0.29 170.72 8 100 Nylon Ball 25.21 28.64 1.07 27.76 71.54 72.57 0.3 71.99 141.92 142.32 0.16 142.12 9 60 50 Nylon Ball 36.03 39.13 0.97 37.26 91.09 92.37 0.41 91.63 198.43 201.08 1.08 199.55 10 100 Nylon 30.2 31.86 0.57 31.15 86.82 87.3 0.18 87 167.39 169.68 0.74 168.34 38 Ball 11 70 50 Nylon Ball 43.92 45.02 0.32 44.29 108.3 110.51 0.66 109.12 236.06 241.67 1.79 240.11 12 100 Nylon Ball 35.44 35.89 0.19 35.67 101.99 103.1 0.37 102.52 186.81 197.3 3.21 195.09 Figure 14. MCL November 2024 Peak Average Pressures: Brass Needle Valve 39 Figure 15. MCL November 2024 Peak Average Pressures: Nylon Ball Valve The fourth iteration of the MCL in January 2025 had one change, the brass needle valve was replaced with a 1\u201d nylon ball valve. Based on the reference data obtained from Jeong et al., the corresponding experiment setup seen in Table 11 was executed for the MCL13. Table 11. MCL January 2025 Experiment Test # HR SV Resistance Valve Position 1 60 70 25 2 50 3 100 4 90 70 25 5 50 40 6 100 7 75 60 25 8 50 9 100 10 75 75 25 11 50 12 100 Upon completion of the experiments, the results were compiled and tabulated below in Table 12, with minimum, maximum, average, and standard deviation taken for each of the pressure readings. As with previous testing, three trials were performed per test and then were averaged together. The individual graphs for each of the tests are located in Appendix D. Table 12. MCL January 2025 Pressure Readings Peak Differential Pressures (mmHg) Peak Aortic Pressures (mmHg) Peak Left Ventricular Pressures (mmHg) Test # HR SV Resistance Valve Position Min Max Std Dev. Avg. Min Max Std Dev. Avg. Min Max Std Dev. Avg. 1 60 70 25 36.98 37.15 0.05 37.08 137.09 138.26 0.41 137.65 219.68 220.82 0.34 220.31 2 50 20.24 20.37 0.04 20.29 95.62 96.44 0.29 95.95 127.4 128.03 0.21 127.67 3 100", "tests are located in Appendix D. Table 12. MCL January 2025 Pressure Readings Peak Differential Pressures (mmHg) Peak Aortic Pressures (mmHg) Peak Left Ventricular Pressures (mmHg) Test # HR SV Resistance Valve Position Min Max Std Dev. Avg. Min Max Std Dev. Avg. Min Max Std Dev. Avg. 1 60 70 25 36.98 37.15 0.05 37.08 137.09 138.26 0.41 137.65 219.68 220.82 0.34 220.31 2 50 20.24 20.37 0.04 20.29 95.62 96.44 0.29 95.95 127.4 128.03 0.21 127.67 3 100 18.43 18.54 0.04 18.48 98.08 98.77 0.24 98.4 117.9 118.29 0.14 118.08 4 90 70 25 54.91 56.04 0.37 55.5 196.6 199.57 1.04 198.38 281.77 281.77 0 281.77 41 5 50 38.67 39.11 0.13 38.95 203.26 205.74 0.65 204.4 242.21 244.48 0.68 243.56 6 100 34.7 34.91 0.07 34.83 193.8 195.69 0.56 194.98 218.32 219.1 0.24 218.72 7 75 60 25 36.45 38.01 0.54 37.18 154.61 160.17 1.72 157.17 221.07 231.94 3.66 226.54 8 50 25.82 26.11 0.09 25.94 160.32 161.68 0.47 161.01 124.94 126.24 0.45 125.62 9 100 23.82 23.97 0.05 23.86 149.47 149.98 0.18 149.71 122.36 123.4 0.38 122.7 10 75 75 25 53.8 54.73 0.3 54.26 215.18 217.83 0.83 216.05 281.77 281.77 0 281.77 11 50 28.99 29.23 0.07 29.11 147.47 149.09 0.62 148.27 181.02 182.37 0.44 181.91 12 100 280.09 28.28 0.06 28.17 144.81 147.09 0.81 145.89 176.28 177.89 0.49 177.26 The peak average pressures for the left ventricular, aortic, and differential pressures were also plotted and are shown in Figure 16 below. Figure 16. MCL January 2025 Peak Average Pressures 42 As mentioned in the methods section, the Solidworks models for the MCL test section molds were 3D printed with white PVB filament and the results are shown in Figure 17. Figure 17. 3D Printed Molds made of PVB a). Right-half Mold b). Left-half Mold Once the molds were filled with Sylgard 184 mixture, cured for 3 days, and de-molded with the IPA tank, the Sylgard test sections were then ready for a test fit into the MCL. However, the resulting Sylgard test section, as seen in Figure 18, did not possess the qualities needed for potentially performing PIV measurements in the future. Figure 18. Sylgard Test Section 43 The images in Figure 18 depict various bubbles present in the cured Sylgard as well as an opacity that is the opposite of what is needed for PIV measurements, transparency. Furthermore, some of the locating and", "Sylgard test sections were then ready for a test fit into the MCL. However, the resulting Sylgard test section, as seen in Figure 18, did not possess the qualities needed for potentially performing PIV measurements in the future. Figure 18. Sylgard Test Section 43 The images in Figure 18 depict various bubbles present in the cured Sylgard as well as an opacity that is the opposite of what is needed for PIV measurements, transparency. Furthermore, some of the locating and design features that would help maintain stability of the assembled test section or were meant to hold the mechanical heart valve broke off or were damaged during the de-molding process from the PVB molds. Since this Sylgard mold was unable to be used for PIV measurements or even to house the St. Jude mechanical heart valve, a second iteration of the Sylgard test section was modeled as seen in Figure 19. The portions of the model that are transparent correspond to the components made out of Sylgard and the components of the assembly in dark gray are the second iteration of barbed tube fittings printed from clear resin. The barbed tube fittings possess 1/4\u201d holes which will house 1/4\u201d threaded rods that will help compress the tube fittings onto the Sylgard test section components through the use of threaded nuts. As in the previous iteration, the important design considerations were for the test section to be able to house the St. Jude mechanical valve and possess a location from which pressure measurements could be taken. Figure 19. 2nd Iteration of SolidWorks Sylgard Test Section After modeling, the molds were then 3D printed with clear filament PVB as seen in Figure 20 and then filled with mixed Sylgard. For this second iteration, the Sylgard was mixed for 10 minutes by hand and then de-gassed in a vacuum chamber to remove the bubbles 44 generated from the mixing procedure. The Sylgard was left to cure for three days prior to submerging the molds in the IPA dissolution tank to expose the cured Sylgard test sections. Figure 20. 2nd Iteration PVB Molds The resulting Sylgard test section molds are depicted below in Figure 21. Here, a much more transparent surface and interior of Sylgard was achieved, with the physical features of the mold remaining intact and free of damage. Figure 21. 2nd Iteration Cured Sylgard Test Section Contrary to what was designed in", "left to cure for three days prior to submerging the molds in the IPA dissolution tank to expose the cured Sylgard test sections. Figure 20. 2nd Iteration PVB Molds The resulting Sylgard test section molds are depicted below in Figure 21. Here, a much more transparent surface and interior of Sylgard was achieved, with the physical features of the mold remaining intact and free of damage. Figure 21. 2nd Iteration Cured Sylgard Test Section Contrary to what was designed in the first iteration of the Sylgard test section, the pressure ports were drilled out using a 1/16\u201d drill bit and then a 2\u201d piece of clear 1/4\u201d acrylic tube was inserted into each hole to allow for the pressure sensor readings to be taken from the MCL. With the Sylgard test section now ready to use in the MCL, the acrylic test section was removed from the loop and the assembly of the Sylgard test section consisting of the St. Jude mechanical heart valve, resin tube fittings, and threaded rods were installed as seen in Figure 22. 45 Figure 22. Assembled Sylgard Test Section installed in MCL At the time the Sylgard test section was completed, the CardioLab had also been equipped to perform PIV experiments on the MCL. Given that the second Sylgard test section had better surface quality and opacity than the first test section, Dr. Alessandro Bellofiore assisted in setting up and aligning the laser and camera to conduct the PIV experiments. Unfortunately, after attempts to position the Sylgard test section at the appropriate height and angle, and adjusting laser/camera settings in order to begin collecting PIV data, the experiment was cancelled. The parameters for the PIV experiment using FlowMaster software are shown in Figure 23 and an image of the Sylard test section through the camera during a test is shown in Figure 24. 46 Figure 23. FlowMaster Software Parameters 47 Figure 24. Image of the Sylgard Test Section under the PIV camera As seen in Figure 24, there are concentrated areas of light throughout the Sylgard to the left of the St. Jude mechanical heart valve as well as vertical \u201clines\u201d that are darker in appearance. As a result of this, the only measurements taken with the Sylgard test section in the MCL were pressure and flow readings using the experimental setup from Table 11. Each test consisted of three individual trials that were then", "Sylgard Test Section under the PIV camera As seen in Figure 24, there are concentrated areas of light throughout the Sylgard to the left of the St. Jude mechanical heart valve as well as vertical \u201clines\u201d that are darker in appearance. As a result of this, the only measurements taken with the Sylgard test section in the MCL were pressure and flow readings using the experimental setup from Table 11. Each test consisted of three individual trials that were then averaged together. Once the experiments were completed, the results for pressure were tabulated in Table 13 and the results for flow rate were tabulated in Table 14. 48 Table 13. MCL May 2025 Pressure Readings Peak Differential Pressures (mmHg) Peak Aortic Pressures (mmHg) Peak Left Ventricular Pressures (mmHg) Test # HR SV Resistance Valve Position Min Max Std Dev. Avg. Min Max Std Dev. Avg. Min Max Std Dev. Avg. 1 60 70 25 42.03 43.44 0.44 42.74 194.96 196.51 0.46 195.54 187.79 189.98 0.88 188.87 2 50 30.13 30.31 0.07 30.25 131.36 133.16 0.58 132.49 135.39 136.74 0.44 136.1 3 100 27.38 27.55 0.05 27.48 114.76 116.89 0.66 115.99 123.37 124.47 0.34 123.89 4 90 70 25 56.24 56.24 0 56.24 281.63 281.63 0 281.63 277.59 279.32 0.52 278.75 5 50 48.16 48.46 0.09 48.35 241.77 245.96 1.14 244.18 216.08 217.68 0.52 217.1 6 100 44.42 44.76 0.09 44.55 218.5 222.12 1.04 220.32 198.47 200.3 0.45 199.08 7 75 60 25 46.73 47.7 0.28 47.27 204.57 207.81 1 206.45 202.65 204.4 0.56 203.41 8 50 33.49 33.72 0.07 33.62 165.91 168.33 0.67 167.08 149.45 151.08 0.47 150.21 9 100 29.19 30.85 0.45 29.78 149.18 151.42 0.7 150.15 131.25 134.52 1.12 132.61 10 75 75 25 55.89 56.24 0.12 56.16 263.16 265.52 0.9 264.36 262.08 263.8 0.5 263.14 11 50 43.46 43.6 0.05 43.52 208.27 210.28 0.72 209.38 196.23 197.96 0.57 197 12 100 39.47 39.77 0.09 39.66 189.11 191.63 1.02 190.25 178.7 180.14 0.55 179.52 49 Table 14. MCL May 2025 Flow Rate Readings Peak Flow Rate (L/min) Test # HR SV Resistance Valve Position Min Max Std Dev. Avg. 1 60 70 25 17.1 17.48 0.12 17.25 2 50 18.99 19.56 0.16 19.24 3 100 19.36 20.04 0.22 19.73 4 90 70 25 19.61 21.23 0.47 19.95 5 50 23.14 25.4 0.74 24.06 6 100 23.37 24.25 0.31 23.89 7 75 60 25 16.88 17.11 0.08 17 8", "1.02 190.25 178.7 180.14 0.55 179.52 49 Table 14. MCL May 2025 Flow Rate Readings Peak Flow Rate (L/min) Test # HR SV Resistance Valve Position Min Max Std Dev. Avg. 1 60 70 25 17.1 17.48 0.12 17.25 2 50 18.99 19.56 0.16 19.24 3 100 19.36 20.04 0.22 19.73 4 90 70 25 19.61 21.23 0.47 19.95 5 50 23.14 25.4 0.74 24.06 6 100 23.37 24.25 0.31 23.89 7 75 60 25 16.88 17.11 0.08 17 8 50 18.92 19.61 0.25 19.19 9 100 19.25 20.17 0.31 19.68 10 75 75 25 19.84 20.36 0.17 20.03 11 50 22.19 22.83 0.2 22.55 12 100 22.63 24.37 0.5 23.75 50 In addition, the peak average pressures and the peak average flow rates were plotted and are shown in Figure 25 and Figure 26. The individual test graphs are located in Appendix E for the pressure readings and Appendix F for the flow rate readings. Figure 25. MCL May 2025 Peak Average Pressures Figure 26. MCL May 2025 Peak Average Flow Rates 51 6. Discussion 6.1. MCL Data: Acrylic Test Section As mentioned previously, changes were made to the MCL setup because of the initial issue regarding elevated/saturated pressure readings from May 2024. To reiterate, the main changes made to the second iteration of the MCL in June 2024 were the addition of two compliance chambers and the removal of the brass needle valve, replacing it with a ball valve after the reservoir tank of the MCL. The initial suspicion for the elevated pressures was that the brass needle valve was not functioning properly due to its positioning, where the adjustment knob was oriented upward meaning it could be trapping any air bubbles that remained in the MCL and affecting the regulation of resistance in the loop. As a result, by removing it entirely from the loop, which can be assumed as having the brass valve entirely open, it would facilitate focusing on the pressure issue. A simple three test experiment was executed with the June 2024 MCL setup and from the results for each test along with the graphs found in Figure 11, the Left Ventricular average pressures were found within 5 mmHg or less of each other despite having different ball valve settings: 25, 50, and 100 percent open. The largest differential pressure occurred during the second test where the ball valve was 50 percent open. Based", "would facilitate focusing on the pressure issue. A simple three test experiment was executed with the June 2024 MCL setup and from the results for each test along with the graphs found in Figure 11, the Left Ventricular average pressures were found within 5 mmHg or less of each other despite having different ball valve settings: 25, 50, and 100 percent open. The largest differential pressure occurred during the second test where the ball valve was 50 percent open. Based on the behavior of subsequent tests on other MCL iterations, there is a high likelihood that there may have been an air bubble that was trapped in the tube leading to the differential pressure sensor, which may have caused it to read the differential pressure lower than it actually was. However, the most important factor to note was that the pressure readings and waveforms, which reached a peak of above 100 mmHg, were not saturated as in the previous May 2024 setup, meaning that the addition of the two 52 compliance chambers had a positive effect on the MCL\u2019s ability to emulate the elasticity of vasculature in the human body, particularly during systole. In an effort to continue keeping the emulation of vasculature within the MCL but also re-introduce the regulation of resistance in the loop, one additional compliance chamber and ball valve were added in the third iteration of the MCL in September 2024 along with re-installing the brass needle, this time with the adjustment knob facing downwards. With this positioning of the needle valve, the risk of trapping any air bubbles within the valve would be minimized and would allow for better control of the resistance within the MCL. The experiment setup that was run for this iteration of the MCL expanded the values of heart rate and stroke volume that were tested along with the percent aperture of the resistance valve. From the results in Table 6 and graphs seen in Figure 13, a general trend is observed for all three pressure sensor readings, which is that as the resistance valve opening increases, the pressure values decrease. Although resistance and compliance are not proportional or simply inversely related, as the resistance valve was opened in the MCL as a result of the test conditions, the compliance chambers, along with the flexible tygon tubing, were able to dampen the pressure fluctuations to a larger degree. This is because", "and graphs seen in Figure 13, a general trend is observed for all three pressure sensor readings, which is that as the resistance valve opening increases, the pressure values decrease. Although resistance and compliance are not proportional or simply inversely related, as the resistance valve was opened in the MCL as a result of the test conditions, the compliance chambers, along with the flexible tygon tubing, were able to dampen the pressure fluctuations to a larger degree. This is because when looking at the equations for compliance in Figure 4, by opening the resistance valve, the MCL\u2019s pressure drop diminishes and the volume of fluid passing through the MCL increases, which results in a larger compliance value across the MCL. It is important to note that the increase in the number of compliance chambers in the MCL also contribute to decreasing the overall pressure in the loop. This effect and general trend is also observed for the remaining tests performed on the later iterations of the MCL. For instance, in the November 2024 iteration, an alternate resistance valve, specifically a 1/4\u201d nylon ball valve, was tested as an alternative to the brass needle valve. 53 The main concern for the MCL during this time period was the tarnish that was coming off of the brass and staying in the MCL. Previously, the CardioLab team had found that the membrane of the Vivitrolabs pump had ruptured and so in order to preserve the integrity of the MCL equipment moving forward in a preventative manner, the nylon ball valve was evaluated against the brass needle valve following the experiment setups found in Table 7 and Table 8. The results of this comparison, graphed individually in Appendix C and the average pressures graphed in Figure 14 and Figure 15 for the brass valve and nylon valve respectively, demonstrated the same behavior where the peak pressures decreased when the percentage aperture of the resistance valve increased. However, for the brass needle valve, only the aortic average pressures for tests 1-2 and 5-6 exhibited the opposite phenomenon where the pressure increased even when the resistance valve was opened from 50 to 100 percent. This variation in behavior could have been attributed to measurement artifacts considering tests 3-4 did display a decrease in pressure after opening the valve. This is further supported by comparing the aortic average pressures of the nylon valve, which followed the observed", "valve increased. However, for the brass needle valve, only the aortic average pressures for tests 1-2 and 5-6 exhibited the opposite phenomenon where the pressure increased even when the resistance valve was opened from 50 to 100 percent. This variation in behavior could have been attributed to measurement artifacts considering tests 3-4 did display a decrease in pressure after opening the valve. This is further supported by comparing the aortic average pressures of the nylon valve, which followed the observed trend. Unfortunately, the nylon ball valve demonstrated peak pressure values for the left ventricle pressure sensor that were above values of hypertension when at higher stroke volumes. These elevated pressure values were attributed to the sudden decrease in diameter from the tygon tubing of the MCL which is 1\u201d in diameter to the nylon valve which is 1/4\u201d in diameter. This increase in pressure was also visually observed in the third compliance chamber that was located downstream of the nylon valve. Specifically, the compliance chamber began to inflate as a result of the pressure build up. Although the 1/4\u201d nylon valve did not produce the same results as the brass needle valve during testing, the added benefit of removing the presence of tarnish altogether led to another 54 MCL experiment in January 2025 which evaluated a larger 1\u201d nylon ball valve. Contrary to the previous experimental matrices, the matrix followed in January 2025 as seen in Table 11 is adapted from the experimental setup found in Jeong et al.\u2019s paper13. Jeong et al. developed a cardiovascular simulator that included an artificial aorta and by controlling the heart rate and stroke volume parameters, they were able to generate blood pressure values that replicated human physiological values with an error of less than 1 mmHg13. Since the final objective of this project was to achieve human physiological pressures, Jeong et al.\u2019s data proved to be a great reference point since the data was also collected on a mock circulation loop. In particular, Jeong et al. explored varying heart rate from 60 BPM to 100 BPM while keeping stroke volume fixed in one experiment, and varied stroke volume from 60 mL to 80 mL while keeping heart rate constant in a second experiment13. As seen in Figure 27, graphs a) and b) displayed the effect of controlling stroke volume and heart rate respectively, in Jeong et al.\u2019s MCL13. Figure 27. Graphs a) and", "was also collected on a mock circulation loop. In particular, Jeong et al. explored varying heart rate from 60 BPM to 100 BPM while keeping stroke volume fixed in one experiment, and varied stroke volume from 60 mL to 80 mL while keeping heart rate constant in a second experiment13. As seen in Figure 27, graphs a) and b) displayed the effect of controlling stroke volume and heart rate respectively, in Jeong et al.\u2019s MCL13. Figure 27. Graphs a) and b) from Figure 5 in Jeong et al.13 Since the Vivitrolabs superpump controller was only programmed with a 60 BPM, 75 BPM, and 90 BPM waveform, a select number of cases were adapted from their experiment 55 setup in order to work with the MCL in the Cardiolab. The data obtained from the tests and displayed in Table 12 again show the same trend between pressure and percentage aperture of the resistance valve. However, there was a change in the definition of the resistance valve settings at 25%, 50%, and 100%. During testing, it was found that until a certain point when closing the ball valve, there is little to no change in the pressure readings obtained from the pressure sensors. This is because ball valves are not intended to allow for fine tuning of flow, rather only a simple on and off setting. In the case of the 1\u201d nylon ball valve, it took 3.5 turns to completely close the valve and the pressure readings remained constant through the first 3 turns, meaning the increase in pressure occurred from the last half turn remaining. Taking this into account, the 100% resistance valve was set to 3 turns, 50% was set to .25 turns after , and 25% was set to .125 turns after. The positions were marked on the valve to minimize variation between trials and test conditions. As a result of this change in definition, there was also a correlated change in the peak pressures obtained by the pressure sensors during the experiment. At a value of 25% for each heart rate and stroke volume setting, the peak left ventricular average pressures reached values of above 200 mmHg, significantly outside physiological hypertension values. The largest pressure reading was obtained when the heart rate was set to 90 BPM and the stroke volume was 70 mL. Despite this, the peak left ventricular average pressure readings at 50% and 100%", "a correlated change in the peak pressures obtained by the pressure sensors during the experiment. At a value of 25% for each heart rate and stroke volume setting, the peak left ventricular average pressures reached values of above 200 mmHg, significantly outside physiological hypertension values. The largest pressure reading was obtained when the heart rate was set to 90 BPM and the stroke volume was 70 mL. Despite this, the peak left ventricular average pressure readings at 50% and 100% resistance valve settings were closer to physiological pressures, 120 mmHg. 6.2. Sylgard Molds & PIV The promising results that the Cardiolab could achieve physiological pressures from the data obtained in the January 2025 experiments prompted the need for the test section made from Sylgard 184 to be completed. As mentioned in the results section, the first iteration of the 56 Sylgard mold did not possess the transparency and clarity needed to perform PIV measurements, leaving several lessons learned with regards to the technique of mixing and pouring. In particular, when the Sylgard was mixed initially, the silicone mix was poured directly into the test section molds without removing the bubbles that were introduced from the mixing process. In the second iteration of the Sylgard molds, after the mixing of the Sylgard sub-components, the Sylgard was placed into a vacuum chamber in order to degas the mixture and remove a majority of the bubbles. After the vacuum chamber, a needle was used to pop the remaining bubbles and also remove any visible dust or particles that entered the mixture. Once this setup process was complete, the Sylgard was poured from the mixing container into the new PVB molds, ensuring no new bubbles were generated from the pouring technique. The new PVB molds were designed with the intent to reduce the amount of Sylgard used, bringing the total volume of Sylgard used down from 22 oz in the first iteration, to 11 oz in this second iteration. In addition, the new PVB molds were printed with a lower infill percentage of 10% and a setting of one layer per wall. This stemmed from the first iteration PVB mold where the PVB took a long time to dissolve in the IPA tank due to the large thickness of the PVB. Unfortunately, during the second degassing procedure after the Sylgard was poured into the molds, one of the center cores had some filament collapse", "oz in this second iteration. In addition, the new PVB molds were printed with a lower infill percentage of 10% and a setting of one layer per wall. This stemmed from the first iteration PVB mold where the PVB took a long time to dissolve in the IPA tank due to the large thickness of the PVB. Unfortunately, during the second degassing procedure after the Sylgard was poured into the molds, one of the center cores had some filament collapse due to the low pressure environment. For future iterations of the test section mold, the number of layers in the 3D print can be increased to two or a hollow core design could be used so that there isn\u2019t any air contained within the core that could break the filament. Ultimately, as in the first iteration, these new molds were left to cure at room temperature, however, for future iterations of the Sylgard test section, using an air-circulating oven could potentially reduce curing time from two to three days to a magnitude of hours. 57 Despite attempts to position the second iteration Sylgard test section underneath the high speed camera and laser in preparation for PIV measurements, the quality of the Sylgard test section, although improved, was still not adequate to run a PIV experiment. As mentioned previously, Figure 24 displays an image of the Sylgard test section through the view of the high speed camera and the laser pointed perpendicular to the camera through the center of the test section. The concentrated areas of light that were observed indicated that there were still some bubbles or particles that were not removed during the degassing procedure in the vacuum chamber. In addition to this, the vertical \u201clines\u201d that were shown in the image likely stemmed from two things: one was the quality of the PVB mold and the second was contamination. During the preparation of the PVB molds, one step is to spray isopropyl alcohol (IPA) on the surfaces of the mold that will come into contact with the Sylgard. This is to ensure the PVB is smoothened out as much as possible and the cured Sylgard won\u2019t take on the layer design that is seen on typical fused disposition modeling (FDM) 3D printed components, which is what occurred to the second iteration of the Sylgard test section. However, another contributing factor to these \u201clines\u201d is likely the presence of", "is to spray isopropyl alcohol (IPA) on the surfaces of the mold that will come into contact with the Sylgard. This is to ensure the PVB is smoothened out as much as possible and the cured Sylgard won\u2019t take on the layer design that is seen on typical fused disposition modeling (FDM) 3D printed components, which is what occurred to the second iteration of the Sylgard test section. However, another contributing factor to these \u201clines\u201d is likely the presence of dust/particles in the Sylgard which causes a shadow to be cast when the laser is pointed through the Sylgard. As such, for future iterations of the Sylgard test section, several actions can be taken to ensure the quality of the Sylgard is improved. The first would be to have the mixture prepared in a fume hood or laminar flow hood to minimize the chance of dust/particle contamination in the Sylgard mixture. Furthermore, the PVB molds should be coated in a larger layer of IPA to allow for the surfaces coming into contact with the Sylgard to be smoothened out. Lastly, as mentioned above, using an air-circulating oven while reducing cure time, it could also assist in making any residual bubbles rise through the Sylgard during curing. 58 6.3. MCL Data: Sylgard Test Section With the Sylgard test section installed in the MCL, the same experiment matrix from January 2025 based on Jeong et al.\u2019s data, was also used in May 2025 to evaluate the MCL\u2019s ability to produce the same pressures despite the change in test section. As seen in Table 13 and Figure 25, the data collected showed similar behaviors when comparing the peak average pressure. The most notable behavior was that the Sylgard test section yielded peak average pressures that followed the same trend as previous acrylic test section iterations of the MCL prior to January 2025, where the pressure decreased as percentage aperture of the resistance valve increased. However, the peak aortic average pressures from the acrylic test section MCL in January 2025 displayed different trends for tests 4-9 where the 50% valve setting test demonstrated higher peak aortic average pressures than the 25% valve setting tests. Overall, the peak left ventricular average pressures for the 25% valve settings on both the May 2025 Sylgard test section data and the January 2025 acrylic test section data were within 20 mmHg of each other. On the other hand,", "increased. However, the peak aortic average pressures from the acrylic test section MCL in January 2025 displayed different trends for tests 4-9 where the 50% valve setting test demonstrated higher peak aortic average pressures than the 25% valve setting tests. Overall, the peak left ventricular average pressures for the 25% valve settings on both the May 2025 Sylgard test section data and the January 2025 acrylic test section data were within 20 mmHg of each other. On the other hand, the peak aortic average pressures for the same valve setting varied significantly from within 50 mmHg up to 100 mmHg and the peak differential average pressures were within 10 mmHg. For the other resistance valve settings at 50%, the differences between the Sylgard and the acrylic test section left ventricular, aortic, and differential peak average pressures were also within the values mentioned, if not much closer in range to each other. For instance, for test number 12 where the heart rate was set to 75 BPM and stroke volume was 75 mL, the left ventricular peak average pressures were within 2.26 mmHg of each other. Another factor to note between the Sylgard and acrylic test section data was that the pressures collected from the Sylgard test section MCL were lower than the pressures collected 59 from the acrylic test section MCL. One potential reason for this phenomenon is that since the Sylgard is not rigid like acrylic, it is likely that similar to the tygon tubing, the Sylgard could be expanding slightly during systole and having a pressure dampening effect. Effectively, increasing resistance, coupled with heart rate and stroke volume settings, has produced drastic effects on pressure within the MCL regardless of test section material. One of the primary focuses of this project was to reduce the pressures observed in the MCL, and the second was to observe the changes in flow rate in response to changes in heart rate, stroke volume, and resistance valve settings. From the data collected in May 2025, the flow rates also followed the same trend as the peak average pressures where flow rate increased as the valve percentage aperture increased. This behavior confirmed what is expected to be seen when the flow is restricted by reducing a portion of the loop within the MCL. On top of increasing the pressure due to the decrease in diameter at the resistance valve, the amount of fluid", "stroke volume, and resistance valve settings. From the data collected in May 2025, the flow rates also followed the same trend as the peak average pressures where flow rate increased as the valve percentage aperture increased. This behavior confirmed what is expected to be seen when the flow is restricted by reducing a portion of the loop within the MCL. On top of increasing the pressure due to the decrease in diameter at the resistance valve, the amount of fluid that could pass through the upstream portion of the MCL was also reduced , as much as 2.48 L/min in the case of test 1 in Table 14. 6.4. MCL Data & Physiological Data As mentioned in the results, when the resistance valve setting was set to 25%, the pressures achieved within the MCL for the simulated left ventricle and aorta were not in the normal physiological range, which is around 120 mmHg. However, a 25% resistance valve setting can be correlated to having very stiff vasculature and in the human body, there is a mixture of more- and less-compliant vasculature which function together to ensure that blood flows efficiently to bodily tissue. It is not to state that extremely high blood pressures are not physiologically possible, but considering the heart rate and stroke volume settings used, it would be unlikely for the left ventricular pressures to reach levels as high as 200 mmHg at 60 BPM, for 60 example. By taking this into account, it can be said that both the acrylic and Sylgard test sections were able to achieve human physiological pressures when the resistance valve was set to 50% and 100%. By looking at Table 13, the peak left ventricular average pressures ranged from mainly from 123 mmHg up to 150mmHg when at the 50% and 100% resistance valve settings. One set of data outputted left ventricular pressures that would be considered as physiologically hypertensive, at 179 mmHg. By analyzing the pressure data from each sensor, the MCL was confirmed to also be able to produce pressure readings that corresponded to typical heart valve behavior. Specifically, the differential pressure is positive, which matches the mechanism where left ventricular pressure increases as a result of the heart muscles contracting, causing the heart valve to open and push blood through the aorta and into the rest of the body. In addition to this, the values for left ventricular pressure", "mmHg. By analyzing the pressure data from each sensor, the MCL was confirmed to also be able to produce pressure readings that corresponded to typical heart valve behavior. Specifically, the differential pressure is positive, which matches the mechanism where left ventricular pressure increases as a result of the heart muscles contracting, causing the heart valve to open and push blood through the aorta and into the rest of the body. In addition to this, the values for left ventricular pressure were also higher than the aortic pressure readings, confirming that the MCL was behaving more closely as a benchtop system that can simulate human physiological pressure. Despite this, when analyzing the pressure profiles from each of the sensors, some deficiencies were identified in the MCL. For instance, when comparing the profiles to the Wiggers diagram, both shown in Figure 29, the aortic pressure profiles from the MCL did not resemble the profile outlined in the Wiggers diagram. 61 Figure 28. Pressure profiles from Test #3 in Table 13 (left) versus Wiggers Diagram (right)31 By referencing the Wiggers diagram above, the aortic pressure profile from the MCL had a much sharper shape and the pressure values on the Wiggers diagram show the aortic pressure is around 90 mmHg prior to the start of systole31. In addition to this, the MCL graphs for aortic pressure had data that ranged from negative values up to approximately 100 mmHg. On the other hand, the left ventricular pressures were closest to resembling the corresponding profile in the Wiggers diagram, although there were still minor differences. These differences can be attributed to noise from the pressure sensor as well as any remnant air bubbles within the loop. The MCL\u2019s left ventricular pressure profiles also had negative pressure values like the aortic pressures, but the negative values could be explained through the concept of heart valve regurgitation. By considering this phenomenon, where blood flows in the reverse direction due to issues with the 62 heart valve, the MCL\u2019s negative pressure values can be attributed to suction from the pump or potentially fluttering of the mechanical heart valve. In order to not only rely on only the gauge pressure sensors that were placed before and after the mechanical heart valve in the test section, it was necessary to have a differential pressure sensor to validate the readings being collected from the MCL. When comparing the left ventricular pressure", "direction due to issues with the 62 heart valve, the MCL\u2019s negative pressure values can be attributed to suction from the pump or potentially fluttering of the mechanical heart valve. In order to not only rely on only the gauge pressure sensors that were placed before and after the mechanical heart valve in the test section, it was necessary to have a differential pressure sensor to validate the readings being collected from the MCL. When comparing the left ventricular pressure data to the aortic pressure data, the difference between the pressure values roughly match up with the values being collected from the differential pressure sensor. This result was expected but when looking at the individual trials for each test case, there are some instances in time when the difference in pressure calculated by subtracting the aortic pressure from the left ventricular pressure results in a different value than what is collected from the differential pressure sensor. These differences can be attributed to a couple of factors such as noise generated by the MCL due to not being able to fully remove all air bubbles in the loop, a difference in accuracy between the differential pressure sensor and the gauge pressure sensors, or even the response time of the sensors being inherently different. It is also possible that after long periods of time, the sensors can experience some degree of drift. If this is the case for the current state of the MCL, a routine calibration schedule can be generated to avoid such issues. The flow rate data collected from the MCL also indicated a positive result on the MCL\u2019s ability to emulate human physiological flow rates. The shape of the flow rate profiles reflect pulsatile flow in the MCL and similar to physiological flow in the heart. According to literature, the blood flow rate, or cardiac output, in humans can range from 5-6 L/min when at a resting condition, or as high as 35 L/min during exercise. From the data tabulated in Table 14, the flow rates in the MCL ranged from as low as 17 L/min to as high as 24.06 L/min, which falls into the physiological range14. Similar to the pressure profiles, the flow rate profiles collected exhibited 63 some negative values which can also be explained by the concept of regurgitant flow and/or noise in the sensors. 6.5. MCL Compliance Chamber Selection The mock circulatory loop (MCL)", "as high as 35 L/min during exercise. From the data tabulated in Table 14, the flow rates in the MCL ranged from as low as 17 L/min to as high as 24.06 L/min, which falls into the physiological range14. Similar to the pressure profiles, the flow rate profiles collected exhibited 63 some negative values which can also be explained by the concept of regurgitant flow and/or noise in the sensors. 6.5. MCL Compliance Chamber Selection The mock circulatory loop (MCL) went through multiple revisions during the testing and verification phases, each major revision was marked by the replacement and or addition and subtraction of different components. The goals of these changes were to iteratively improve the MCL pressure waveforms in order to better simulate physiological and pathological conditions of the human cardiovascular system. One area of focus in many of the changes was in the use of compliance chambers that included the number of compliance chambers, location, size and shape, position, volume, and level of fill. The mathematical definition of compliance is the change in volume divide by the change in pressure and was expressed in Figure 4 of section 2.5. Hemodynamics refers to the study of blood through the human circulatory system. Blood is a non-Newtonian fluid, its viscosity and other properties change dynamically in response to local flow conditions, including velocity, pressure, and shear stress. All experimental work and calculations were performed using either deionized water or a glycerol deionized water mixture. Compliance in the human cardiovascular system due to the pulsating flow or cardiac rhythm can be represented by an analog circuit model with capacitance in series with a resistor. If the flow of blood is represented by current flowing in this circuit and the cardiac rhythm as a pulse Qt then Cs and P can be expressed by the equation 64 Figure 29. Equation for aortic compliance17 The pressure associated with a given aortic compliance Cs(P) is P(t)17 Zo \u2013 represents characteristic impedance of the ascending aorta Rs \u2013 represents peripheral resistance a, b \u2013 are constant based on the physiological conditions Figure 30. Equation for flow through compliance branch17 The flow through a compliance branch can be expressed Figure 31. Equation for flow through branch in terms of compliance17 Q(t) is the measured aortic flow Figure 32. Equation for relationship between aortic pressure and flow17 Setting the equations equal, 65 Figure 33. Mathematical model based", "is P(t)17 Zo \u2013 represents characteristic impedance of the ascending aorta Rs \u2013 represents peripheral resistance a, b \u2013 are constant based on the physiological conditions Figure 30. Equation for flow through compliance branch17 The flow through a compliance branch can be expressed Figure 31. Equation for flow through branch in terms of compliance17 Q(t) is the measured aortic flow Figure 32. Equation for relationship between aortic pressure and flow17 Setting the equations equal, 65 Figure 33. Mathematical model based on sampling interval aortic flow17 \u0394T is the sampling interval where \u0394T = ti + 1 -ti ~ dt Figure 34. Representation of sampling interval as a function17 The aortic pressure can be obtained as, Figure 35. Equation for predicted aortic pressure17 66 Figure 36. Analog model of the arterial system incorporating compliance capacitive element17 Analog model of the arterial system incorporating a pressure-dependent compliance element Cs (P). Pa0 and Q are aortic pressure and flow, respectively. Zo is the aortic characteristic impedance. Rs is the peripheral resistance Figure 37. Compliance linear model17 The compliance for the linear model is r - is the aortic diastolic pressure decay time constant Arterial compliance is an important biomechanical feature or property in the cardiovascular system; arterial compliance is the ability of arterial vessels to distend and increase volumetric capacity with increasing blood pressure. Compliance provides capacitance to the system as modeled in the previous figure 36. During the systolic phase when blood is ejected from the left ventricle the arterial compliance allows a slower rise in systolic pressure for a given stroke volume and a lower wall stress and less oxygen consumption22 and left ventricle (LV) performance is maintained at lower energetic cost15. Reduction or impairment of this compliance increases the blood pressure and thereby increases the LV load, which increases both cardiovascular risk factors and mortality rates. Almost every health provider visits, the doctor or nurse takes your blood pressure measurement as part of their basic health screenings. Decreased arterial compliance through stiffness or 67 hardening increases pressure and the velocity of blood flow through the arterial branches of the cardiovascular system. In the fluid flow calculations for the MCL using average flow rates of 70mL per stroke, 60 beats per minute (BPM), 25 mm diameter tubing and an average flow velocity of 0.143 m/s a Reynolds number calculation of 3575 indicates a mild turbulent flow. Using peak flow rates measured", "as part of their basic health screenings. Decreased arterial compliance through stiffness or 67 hardening increases pressure and the velocity of blood flow through the arterial branches of the cardiovascular system. In the fluid flow calculations for the MCL using average flow rates of 70mL per stroke, 60 beats per minute (BPM), 25 mm diameter tubing and an average flow velocity of 0.143 m/s a Reynolds number calculation of 3575 indicates a mild turbulent flow. Using peak flow rates measured at 24L/min (Table 14. MCL May 2025 Flow Rate Readings) gives a peak flow velocity of 0.816 m/s and resultant Reynolds number of 20,408 or turbulent flow. Figure 38. Reynolds number calculation In the first or baseline setup for the MCL no compliance chambers were used in the configuration (Figure 9. May 2024 MCL Setup). Compliance calculations for this configuration 68 for both static and dynamic flow values are not useful and the measured systolic pressures exhibited saturation levels beyond the sensor\u2019s measurable limits e.g. outside of physiological limits. The second iteration or major revision of the MCL setup configuration (Figure 10. June 2024 MCL Setup). Two compliance chambers were added to the MCL, one small chamber was positioned prior to the test section, and another large chamber was positioned after the test section. The first smaller compliance chamber was positioned prior to the St. Jude mechanical heart valve test section and the second compliance chamber was placed midway between the test chamber and the reservoir tank. The first compliance chamber consisted of an inverted bulb style chamber with the bulb portion pointed upwardly. The sealed bulb chamber possessed no purge or fill valves to facilitate adding or removing air from the chamber. Pressurization in the first compliance chamber was assumed to be higher than room atmospheric pressure due to the water column height of the tube feeding to the reservoir pressurizing the line. Using a simple calculation of 13.6 mm of height in a water column is equivalent to 1mmHg, the pressurization in the first compliance chamber due to the water column height was estimated to be 200 mm above the height of the test section. Starting from the Ideal Gas law Isolating volume and differentiating with respect to pressure 69 Starting from the definition of compliance where the change in volume divided by the change in pressure is the compliance of a system Compliance of an ideal gas", "mm of height in a water column is equivalent to 1mmHg, the pressurization in the first compliance chamber due to the water column height was estimated to be 200 mm above the height of the test section. Starting from the Ideal Gas law Isolating volume and differentiating with respect to pressure 69 Starting from the definition of compliance where the change in volume divided by the change in pressure is the compliance of a system Compliance of an ideal gas in a closed vessel is then the starting volume divided by pressure In the second iteration of the MCL loop the second compliance chamber had roughly 200 mL of air under STP. Volume V = 200 mL = 0.0002m3 Pressure P = 1 bar or 1 STP = 1 x 105 Pa To correct for the assumed pressurization in the chamber of 14.71 mmHg The first compliance chamber had roughly 30 mL of air under STP Volume V = 30 mL = 0.00003m3 Pressure P = 1 bar or 1 STP = 1 x 105 Pa + 14.7 mmHg 70 In the third iteration of the MCL, a third compliance chamber was added after the ball valve to simulate the systemic arterial compliance of the human cardiovascular system. This compliance chamber was a plastic water container with approximately 2 Imp gallon capacity and was positioned midway between the valve and the reservoir tank. Due to seal issues on the orifices for the drain, the compliance chamber was only minimally filled with water and for compliance calculations will be assumed to be 2 gallons of air at STP. Volume V = 2 Imp Gallons = 0.007571m3 Pressure P = 1 bar or 1 STP = 1 x 105 Pa The second compliance chamber prior to the heart valve test section was switched from a bulb style Pyrex vessel to a straight cylindrical vessel. The volume of air and the pressurization is assumed to be similar to the prior compliance chamber. Compliance chambers in the fourth and fifth iterations of the MCL setup were not modified and compliance is assumed to be the same for both sets of test results. Air volume fill for the chambers varied and measurements were not recorded in the test runs. Pressure drops for the 1 inch PVC tubing used in the MCL were calculated using the Hazen-Williams equation. Pipe Length = 1 foot 71 Pipe Diameter", "assumed to be similar to the prior compliance chamber. Compliance chambers in the fourth and fifth iterations of the MCL setup were not modified and compliance is assumed to be the same for both sets of test results. Air volume fill for the chambers varied and measurements were not recorded in the test runs. Pressure drops for the 1 inch PVC tubing used in the MCL were calculated using the Hazen-Williams equation. Pipe Length = 1 foot 71 Pipe Diameter = 25 mm Figure 39. Pressure loss in a pipe with frictional head loss calculation The calculated results for the tubing was approximately 0.01 kPa or 0.1 mmHg per foot of tubing including frictional head loss using Copely Developments Plc calculator for pressure drop calculations. The added pressure drops were far less than the sensitivity of the pressure sensors used in reading and therefore were not included in the calculations as being insignificant. In total three compliance chambers were used in the MCL configuration. The placement of the compliance chambers were based on the gross estimation of the compliance in the LV, Aorta, and systemic arterial network. The smallest compliance chamber was placed prior to the aortic valve to represent the compliance of the left ventricle, the second larger chamber was placed post St Jude mechanical heart valve to represent the aortic compliance and the third was placed after the ball valve to represent systemic compliance of the circulatory system. Specific locations of the compliance chambers in relation to other components were shown to be trivial due to the rigid nature of the PVC tubing used in the MCL. Placement of the chamber(s) a few inches or more closer to one component or farther did not create measurable changes in the pressure profiles of the waveforms. Size and fill of the compliance chambers impacted the pressure waveforms but not in a linear fashion. The addition of the compliance chambers from the first iteration of the MCL where there was none to the final configuration of three compliance 72 chambers demonstrated the capacitive effects by their addition. Systolic peak pressure dropped and the pressure waveform exhibited the decay function of the capacitive effects of the energy stored in the chambers. Modifications to the compliance chambers to more controllable and measurable chambers would improve the repeatability and consistency of tests and further enhance the MCL usefulness in replicating varying psychological conditions and", "the first iteration of the MCL where there was none to the final configuration of three compliance 72 chambers demonstrated the capacitive effects by their addition. Systolic peak pressure dropped and the pressure waveform exhibited the decay function of the capacitive effects of the energy stored in the chambers. Modifications to the compliance chambers to more controllable and measurable chambers would improve the repeatability and consistency of tests and further enhance the MCL usefulness in replicating varying psychological conditions and states of the cardiovascular system. Specific choices used in the selection of size and volume of the compliance chamber were guided by equipment available on hand versus designed through thoughtful inspection and calculations. 7. Conclusion Despite not being able to perform PIV measurements, the Sylgard test section held up during testing and did not experience any leaks from the barbed tube fittings nor at the middle section where the two sylgard halves came together to house the St. Jude mechanical heart valve, even at higher heart rates and stroke volumes. This allowed for the collection of pressure and flow data to determine whether or not the MCL was capable of reproducing human physiological data. The data collected throughout the project demonstrated positive results, since the MCL was able to achieve values of pressure ranging mostly from 123 mmHg to 150 mmHg and with a trial that went up to levels of hypertension at 179 mmHg. It is important to note that for the test conditions where the resistance valve was set to 25% aperture, pressures around or above 200 mmHg would be unlikely given the set of parameters that were tested. The flow rate measured from the MCL also fell within human physiological ranges which are 5-6 L/min to 35 L.min. Specifically, the MCL was measured at 17-24.06 L/min, demonstrating that the MCL with the 73 Sylgard test section was about to meet the goal of obtaining pressure and flow data that is in agreement with physiological values at various conditions. One potential limitation of the data collected for this project was the profile of the aortic pressures. Based on the reference from the Wiggers diagram, the profiles collected using the MCL from the Cardiolab did not match in shape despite the values being similar. However, the left ventricular pressure profiles did exhibit similarities to the profile in the Wiggers diagram, despite the amount of noise and presence of regurgitation", "data that is in agreement with physiological values at various conditions. One potential limitation of the data collected for this project was the profile of the aortic pressures. Based on the reference from the Wiggers diagram, the profiles collected using the MCL from the Cardiolab did not match in shape despite the values being similar. However, the left ventricular pressure profiles did exhibit similarities to the profile in the Wiggers diagram, despite the amount of noise and presence of regurgitation in the MCL. For future work performed on the Cardiolab\u2019s MCL, it is prudent to dial in the compliance of the MCL and obtain a better method to control the resistance in the loop. In doing so, it will facilitate changes to the loop length in the event other types of cardiovascular assist devices need to be tested or to test mechanical heart valves by simulating a certain type of cardiovascular disease. Improvements can also be made to the Sylgard test section. Not only can the volume be reduced to utilize less Sylgard mixture, but also as mentioned previously, the pouring and curing technique is vital to producing Sylgard components that can be used to perform PIV experiments. 74 8. References 1. Brindise, M. C., M. M. Busse, and P. P. Vlachos. Density- and viscosity-matched Newtonian and non-Newtonian blood-analog solutions with PDMS refractive index. Exp Fluids 59:173, 2018. 2. Britannica, T. Editors of Encyclopaedia (2024, April 30). blood pressure. Encyclopedia Britannica. https://www.britannica.com/science/blood-pressure 3. Camas\u00e3o, D. B., and D. Mantovani. The mechanical characterization of blood vessels and their substitutes in the continuous quest for physiological-relevant performances. A critical review. Materials Today Bio 10:100106, 2021. 4. Centers for Disease Control and Prevention. Heart disease facts. , 2023.at https://www.cdc.gov/heartdisease/facts.htm 5. Dangas, G. D., J. I. Weitz, G. Giustino, R. Makkar, and R. Mehran. Prosthetic Heart Valve Thrombosis. Journal of the American College of Cardiology 68:2670\u20132689, 2016. 6. Edward Charbek, M. Normal vital signs. , 2024.at https://emedicine.medscape.com/article/2172054-overview?form=fpf 75 7. Farahmand, M., E. Bodwell, G. A. D\u2019Souza, L. H. Herbertson, and C. G. Scully. Mock circulatory loop generated database for dynamic characterization of pressure-based cardiac output monitoring systems. Computers in Biology and Medicine 160:106979, 2023. 8. Falk, K. L., R. Medero, and A. Rold\u00e1n-Alzate. Fabrication of Low-Cost Patient-Specific Vascular Models for Particle Image Velocimetry. Cardiovasc Eng Tech 10:500\u2013507, 2019. 9. Gaziano T, Reddy KS, Paccaud F, et al. Cardiovascular Disease. In: Jamison DT, Breman JG,", ", 2024.at https://emedicine.medscape.com/article/2172054-overview?form=fpf 75 7. Farahmand, M., E. Bodwell, G. A. D\u2019Souza, L. H. Herbertson, and C. G. Scully. Mock circulatory loop generated database for dynamic characterization of pressure-based cardiac output monitoring systems. Computers in Biology and Medicine 160:106979, 2023. 8. Falk, K. L., R. Medero, and A. Rold\u00e1n-Alzate. Fabrication of Low-Cost Patient-Specific Vascular Models for Particle Image Velocimetry. Cardiovasc Eng Tech 10:500\u2013507, 2019. 9. Gaziano T, Reddy KS, Paccaud F, et al. Cardiovascular Disease. In: Jamison DT, Breman JG, Measham AR, et al., editors. Disease Control Priorities in Developing Countries. 2nd edition. Washington (DC): The International Bank for Reconstruction and Development / The World Bank; 2006. Chapter 33. Co-published by Oxford University Press, New York. 10. Ghio, S., S. Schirinzi, and S. Pica. Pulmonary arterial compliance: How and why should we measure it? Global Cardiology Science and Practice 2015:58, 2015. 11. Gregory, S. D. Simulation and Development of a Mock Circulation Loop with Variable Compliance, 2009. 12. Han, J., and D. Trumble. Cardiac Assist Devices: Early Concepts, Current Technologies, and Future Innovations. Bioengineering 6:18, 2019. 13. Jeong, J.-H., B. Lee, J. Hong, T.-H. Yang, and Y.-H. Park. Reproduction of human blood pressure waveform using physiology-based cardiovascular simulator. Sci Rep 13:7856, 76 2023. 14. King J, Lowery DR. Physiology, Cardiac Output. [Updated 2023 Jul 17]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2024 Jan-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK470455/. 15. Kolh P, D'Orio V, Lambermont B, et al. Increased aortic compliance maintains left ventricular performance at lower energetic cost. Eur J Cardiothorac Surg. 2000;17:272\u2013278. doi: 10.1016/S1010-7940(00)00341-9 16. Lewis, A., A. Koukoura, G.-I. Tsianos, A. A. Gargavanis, A. A. Nielsen, and E. Vassiliadis. Organ donation in the US and Europe: The supply vs demand imbalance. Transplantation Reviews 35:100585, 2021. 17. Li JK, Zhu Y. Arterial compliance and its pressure dependence in hypertension and vasodilation. Angiology. 1994 Feb;45(2):113-7. doi: 10.1177/000331979404500205. PMID: 8129185. 18. Li , R. L., J. Russ, C. Paschalides, G. Ferrari, H. Waisman, J. W. Kysar, and D. Kalfa. Mechanical considerations for polymeric heart valve development: Biomechanics, materials, design and manufacturing. Biomaterials 225:119493, 2019. 77 19. Liu, Y., P. Allaire, H. Wood, D. Olsen. Design and Initial Testing of a Mock Human Circulatory Loop for Left Ventricular Assist Device Performance Testing. Artificial Organs 29:341\u2013345, 2005. 20. Maines, B. H., and C. E. Brennen. Lumped Parameter Model for Computing the Minimum Pressure During Mechanical Heart Valve Closure. Journal of Biomechanical Engineering 127:648\u2013655,", "H. Waisman, J. W. Kysar, and D. Kalfa. Mechanical considerations for polymeric heart valve development: Biomechanics, materials, design and manufacturing. Biomaterials 225:119493, 2019. 77 19. Liu, Y., P. Allaire, H. Wood, D. Olsen. Design and Initial Testing of a Mock Human Circulatory Loop for Left Ventricular Assist Device Performance Testing. Artificial Organs 29:341\u2013345, 2005. 20. Maines, B. H., and C. E. Brennen. Lumped Parameter Model for Computing the Minimum Pressure During Mechanical Heart Valve Closure. Journal of Biomechanical Engineering 127:648\u2013655, 2005. 21. Mathew P, Kanmanthareddy A. Prosthetic Heart Valve. [Updated 2023 Jul 24]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2024 Jan-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK536987/. 22. McVeigh, G. E., Bank, A. J., Cohn, J. N., Willerson, J. T., Wellens, H. J. J., & Holmes, D. R. (2007). Cardiovascular medicine. 23. National Heart, Lung, and Blood Institute. How does it work? , 2023 at https://www.nhlbi.nih.gov/health/total-artificial-heart/how-it-works 24. National Heart, Lung, and Blood Institute. What is total artificial heart? , 2023.at https://www.nhlbi.nih.gov/health/total-artificial-heartRyu 25. Oparil, S., M. C. Acelajado, G. L. Bakris, D. R. Berlowitz, R. C\u00edfkov\u00e1, A. F. Dominiczak, G. Grassi, J. Jordan, N. R. Poulter, A. Rodgers, and P. K. Whelton. Hypertension. Nat Rev Dis Primers 4:18014, 2018. 78 26. Petrou, A., M. Granegger, M. Meboldt, and M. Schmid Daners. A Versatile Hybrid Mock Circulation for Hydraulic Investigations of Active and Passive Cardiovascular Implants. ASAIO Journal 65:495\u2013502, 2019. 27. Rajput FA, Zeltser R. Aortic Valve Replacement. [Updated 2023 May 29]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2024 Jan-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK537136/ 28. Rees, M., and P. Kohli. Multivessel coronary artery disease: Symptoms, causes, and treatment. , 2022.at https://www.medicalnewstoday.com/articles/multivessel-coronary-artery-disease 29. Ryu, R., and R. Tran. DOACs in Mechanical and Bioprosthetic Heart Valves: A Narrative Review of Emerging Data and Future Directions. Clin Appl Thromb Hemost 28:107602962211035, 2022. 30. Sidebotham, D., and I. J. Le Grice. Physiology and Pathophysiology. In: Cardiothoracic Critical Care. Elsevier, 2007, pp. 3\u201327.doi:10.1016/B978-075067572-7.50004-7 31. 1.Silverthorn, D. U. Constructing the Wiggers diagram using core concepts: A classroom activity. Advances in Physiology Education 46:714\u2013723, 2022. 79 32. \u0160vec, D., and M. Javorka. Noninvasive arterial compliance estimation. Physiol Res S483\u2013S494, 2021.doi:10.33549/physiolres.934798 33. Tsao, C. W. et al. Heart Disease and Stroke Statistics\u20142023 Update: A Report From the American Heart Association. Circulation 147:, 2023. 34. Trammel JE, Sapra A. Physiology, Systemic Vascular Resistance. [Updated 2023 Jul 10]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2024 Jan-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK556075/. 35.", "Constructing the Wiggers diagram using core concepts: A classroom activity. Advances in Physiology Education 46:714\u2013723, 2022. 79 32. \u0160vec, D., and M. Javorka. Noninvasive arterial compliance estimation. Physiol Res S483\u2013S494, 2021.doi:10.33549/physiolres.934798 33. Tsao, C. W. et al. Heart Disease and Stroke Statistics\u20142023 Update: A Report From the American Heart Association. Circulation 147:, 2023. 34. Trammel JE, Sapra A. Physiology, Systemic Vascular Resistance. [Updated 2023 Jul 10]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2024 Jan-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK556075/. 35. Vincent, J.-L. Understanding cardiac output. Crit Care 12:174, 2008. 36. Xu, K.-W., Q. Gao, M. Wan, and K. Zhang. Mock circulatory loop applications for testing cardiovascular assist devices and in vitro studies. Front. Physiol. 14:1175919, 2023. 80 9. Appendix A. June 2024: Table 4 Left Ventricular and Differential Pressure Graphs Figure 40. Table 4, Test #1, Left Ventricular and Differential Pressures. HR = 60 BPM, Stroke Volume = 70, Valve = 25 Figure 41. Table 4, Test #2, Left Ventricular and Differential Pressures. HR = 60 BPM, Stroke Volume = 70, Valve = 50 Figure 42. Table 4, Test #3, Left Ventricular and Differential Pressures. HR = 60 BPM, Stroke Volume = 70, Valve = 100 81 B. September 2024: Table 6 Left Ventricular, Differential, and Aortic Pressure Graphs Figure 43. Table 6, Test #1, Left Ventricular, Differential, and Aortic PressuresHR = 70, Stroke Volume = 50, Valve = 50 82 Figure 44. Table 6, Test #2, Left Ventricular, Differential, and Aortic PressuresHR = 70, Stroke Volume = 50, Valve = 100 Figure 45. Table 6, Test #3, Left Ventricular, Differential, and Aortic PressuresHR = 70, Stroke Volume = 60, Valve = 50 83 Figure 46. Table 6, Test #4, Left Ventricular, Differential, and Aortic Pressures. HR = 70, Stroke Volume = 60, Valve = 100 Figure 47. Table 6, Test #5, Left Ventricular, Differential, and Aortic Pressures. HR = 70, Stroke Volume = 70, Valve = 50 84 Figure 48. Table 6, Test #6, Left Ventricular, Differential, and Aortic Pressures. HR = 70, Stroke Volume = 70, Valve = 100 Figure 49. Table 6, Test #7, Left Ventricular, Differential, and Aortic Pressures. HR = 70, Stroke Volume = 80, Valve = 50 85 Figure 50. Table 6, Test #8, Left Ventricular, Differential, and Aortic Pressures. HR = 70, Stroke Volume = 80, Valve = 100 86 C. November 2024: Table 9 & Table 10 Left Ventricular, Differential, and Aortic", "48. Table 6, Test #6, Left Ventricular, Differential, and Aortic Pressures. HR = 70, Stroke Volume = 70, Valve = 100 Figure 49. Table 6, Test #7, Left Ventricular, Differential, and Aortic Pressures. HR = 70, Stroke Volume = 80, Valve = 50 85 Figure 50. Table 6, Test #8, Left Ventricular, Differential, and Aortic Pressures. HR = 70, Stroke Volume = 80, Valve = 100 86 C. November 2024: Table 9 & Table 10 Left Ventricular, Differential, and Aortic Pressure Graphs Figure 51. Table 9, Test #1, Left Ventricular, Differential, and Aortic Pressures: Brass Needle Valve. HR = 60 BPM, Stroke Volume = 50 mL, Valve = 50 87 Figure 52. Table 9, Test #2, Left Ventricular, Differential, and Aortic Pressures: Brass Needle Valve. HR = 60 BPM, Stroke Voume = 50 mL, Valve = 100 Figure 53. Table 9, Test #3, Left Ventricular, Differential, and Aortic Pressures: Brass Needle Valve. HR = 60 BPM, Stroke Volume = 60 mL, Valve = 50 88 Figure 54. Table 9, Test #4, Left Ventricular, Differential, and Aortic Pressures: Brass Needle Valve. HR = 60 BPM, Stroke Volume = 60 mL, Valve = 100 89 Figure 55. Table 9, Test #5, Left Ventricular, Differential, and Aortic Pressures: Brass Needle Valve. HR = 60 BPM, Stroke Volume = 70 mL, Valve = 50 90 Figure 56. Table 9, Test #6, Left Ventricular, Differential, and Aortic Pressures: Brass Needle Valve. HR = 60 BPM, Stroke Volume = 70 mL, Valve = 100 Figure 57. Table 10, Test #7, Left Ventricular, Differential, and Aortic Pressures: Nyon Ball Valve. HR = 60 BPM, Stroke Volume = 50 mL, Valve = 50 91 Figure 58. Table 10, Test #8, Left Ventricular, Differential, and Aortic Pressures: Brass Needle Valve. HR = 60 BPM, Stroke Volume = 50 mL, Valve = 100 92 Figure 59. Table 10, Test #9, Left Ventricular, Differential, and Aortic Pressures: Brass Needle Valve. HR = 60 BPM, Stroke Volume = 60 mL, Valve = 50 Figure 60. Table 10, Test #10, Left Ventricular, Differential, and Aortic Pressures: Brass Needle Valve. HR = 60 BPM, Stroke Volume = 60 mL, Valve = 100 93 Figure 61. Table 10, Test #11, Left Ventricular, Differential, and Aortic Pressures: Brass Needle Valve. HR = 60 BPM, Stroke Volume = 70 mL, Valve = 50 94 Figure 62. Table 10, Test #12, Left Ventricular, Differential, and Aortic Pressures: Brass", "= 60 BPM, Stroke Volume = 60 mL, Valve = 50 Figure 60. Table 10, Test #10, Left Ventricular, Differential, and Aortic Pressures: Brass Needle Valve. HR = 60 BPM, Stroke Volume = 60 mL, Valve = 100 93 Figure 61. Table 10, Test #11, Left Ventricular, Differential, and Aortic Pressures: Brass Needle Valve. HR = 60 BPM, Stroke Volume = 70 mL, Valve = 50 94 Figure 62. Table 10, Test #12, Left Ventricular, Differential, and Aortic Pressures: Brass Needle Valve. HR = 60 BPM, Stroke Volume = 70 mL, Valve = 100 D. January 2025: Table 12 Left Ventricular, Differential, and Aortic Pressure Graphs Figure 63. Table 12, Test #1, Left Ventricular, Differential, and Aortic Pressures HR = 60 BPM, Stroke Volume = 70 mL, Valve = 25 95 Figure 64. Table 12, Test #2, Left Ventricular, Differential, and Aortic Pressures HR = 60 BPM, Stroke Volume = 70 mL, Valve = 50 96 Figure 65. Table 12, Test #3, Left Ventricular, Differential, and Aortic Pressures HR = 60 BPM, Stroke Volume = 70 mL, Valve = 100 Figure 66. Table 12, Test #4, Left Ventricular, Differential, and Aortic Pressures HR = 90 BPM, Stroke Volume = 70 mL, Valve = 25 97 Figure 67. Table 12, Test #5, Left Ventricular, Differential, and Aortic Pressures HR = 90 BPM, Stroke Volume = 70 mL, Valve = 50 Figure 68. Table 12, Test #6, Left Ventricular, Differential, and Aortic Pressures HR = 90 BPM, Stroke Volume = 70 mL, Valve = 100 98 Figure 69. Table 12, Test #7, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 60 mL, Valve = 25 Figure 70. Table 12, Test #8, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 60 mL, Valve = 50 99 Figure 71. Table 12, Test #9, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 60 mL, Valve = 100 Figure 72. Table 12, Test #10, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 75 mL, Valve = 25 100 Figure 73. Table 12, Test #11, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 75 mL, Valve = 50 Figure 74. Table 12, Test #12, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 75 mL, Valve = 100", "Volume = 60 mL, Valve = 100 Figure 72. Table 12, Test #10, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 75 mL, Valve = 25 100 Figure 73. Table 12, Test #11, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 75 mL, Valve = 50 Figure 74. Table 12, Test #12, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 75 mL, Valve = 100 101 E. May 2025: Table 13 Sylgard MCL Left Ventricular, Differential, and Aortic Pressure Graphs Figure 75. Table 13, Test #1, Left Ventricular, Differential, and Aortic Pressures HR = 60 BPM, Stroke Volume = 70 mL, Valve = 25 Figure 76. Table 13, Test #2, Left Ventricular, Differential, and Aortic Pressures HR = 60 BPM, Stroke Volume = 70 mL, Valve = 50 102 Figure 77. Table 13, Test #3, Left Ventricular, Differential, and Aortic Pressures HR = 60 BPM, Stroke Volume = 70 mL, Valve = 100 Figure 78. Table 13, Test #4, Left Ventricular, Differential, and Aortic Pressures HR = 90 BPM, Stroke Volume = 70 mL, Valve = 25 103 Figure 79. Table 13, Test #5, Left Ventricular, Differential, and Aortic Pressures HR = 90 BPM, Stroke Volume = 70 mL, Valve = 50 Figure 80. Table 13, Test #6, Left Ventricular, Differential, and Aortic Pressures HR = 90 BPM, Stroke Volume = 70 mL, Valve = 100 104 Figure 81. Table 13, Test #7, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 60 mL, Valve = 25 Figure 82. Table 13, Test #8, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 60 mL, Valve = 50 105 Figure 83. Table 13, Test #9, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 60 mL, Valve = 100 Figure 84. Table 13, Test #10, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 75 mL, Valve = 25 106 Figure 85. Table 13, Test #10, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 75 mL, Valve = 50 Figure 86. Table 13, Test #10, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 75 mL, Valve = 100 F. May 2025: Table 14 Sylgard MCL Flow Rate Graphs 107 Figure", "#10, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 75 mL, Valve = 25 106 Figure 85. Table 13, Test #10, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 75 mL, Valve = 50 Figure 86. Table 13, Test #10, Left Ventricular, Differential, and Aortic Pressures HR = 75 BPM, Stroke Volume = 75 mL, Valve = 100 F. May 2025: Table 14 Sylgard MCL Flow Rate Graphs 107 Figure 87. Table 14, Test #1, Flow Rates HR = 60 BPM, Stroke Volume = 70 mL, Valve = 25 Figure 88. Table 14, Test #2, Flow Rates HR = 60 BPM, Stroke Volume = 70 mL, Valve = 50 108 Figure 89. Table 14, Test #3, Flow Rates HR = 60 BPM, Stroke Volume = 70 mL, Valve = 100 Figure 90. Table 14, Test #4, Flow Rates HR = 90 BPM, Stroke Volume = 70 mL, Valve = 25 109 Figure 91. Table 14, Test #5, Flow Rates HR = 90 BPM, Stroke Volume = 70 mL, Valve = 50 Figure 92. Table 14, Test #6, Flow Rates HR = 90 BPM, Stroke Volume = 70 mL, Valve = 100 110 Figure 93. Table 14, Test #7, Flow Rates HR = 75 BPM, Stroke Volume = 60 mL, Valve = 25 Figure 94. Table 14, Test #8, Flow Rates HR = 75 BPM, Stroke Volume = 60 mL, Valve = 50 111 Figure 95. Table 14, Test #9, Flow Rates HR = 75 BPM, Stroke Volume = 60 mL, Valve = 100 Figure 96. Table 14, Test #10, Flow Rates HR = 75 BPM, Stroke Volume = 75 mL, Valve = 25 112 Figure 97. Table 14, Test #10, Flow Rates HR = 75 BPM, Stroke Volume = 75 mL, Valve = 50 Figure 98. Table 14, Test #10, Flow Rates HR = 75 BPM, Stroke Volume = 75 mL, Valve = 100", "Fabrication and Testing of Rigid Polymeric Heart Valves B.S. Biomedical Engineering Program Biomedical Engineering Department Charles W. Davidson College of Engineering Prepared by: Naomi Bolotin Nhu Tu Hannah Tulabut Advised by: Dr. , Ph.D. Alessandro BellofioreFaculty Advisor San Jose State University 1 Abstract Valvular heart disease (VHD) is a condition where one or multiple valves of the heart cannot properly function. Factors that induce this disease include infections, old age, or other cardiac-related conditions (Small et al., 2024). This project investigates the potential use of 3D printed polymeric heart valves made from polycarbonate (PC) as an alternative to commercial valves for utilization in rapid prototype testing. Prototypes were designed based on a SolidWorks model of the St. Jude Medical bileaflet mechanical heart valve. Flow and pressure will be measured from the prototype valve and evaluated for comparison to Saint Jude\u2019s bileaflet mechanical heart valve using a mock circulation loop (MCL) in the aortic position. The flow velocity of the polymeric valves closely matched the shape of the St. Jude valve; however, an audible noise was confirmed due to the result from friction at the hinges. The polymeric valves show promise as a viable option in rapid prototype testing. Future researchers can reference the information from our study to refine the valves for improved use. Keywords Mechanical heart valve, Rigid polymer, 3D printing, Thrombogenicity, Aortic valve, Polycarbonate (PC), Prototyping, Rapid prototyping 2 Table of Contents Abstract ........................................................................................................................................... 2 Introduction .................................................................................................................................... 4 Literature review ............................................................................................................................ 5 FDA Considerations ...................................................................................................................... 5 Biomedical Motivation ................................................................................................................... 5 Statement of Need .......................................................................................................................... 6 I. Mechanical Heart Valve Fabrication using 3D Printer and Assembly ................................... 7 II. MHV Testing: Hydrodynamic Performance ......................................................................... 7 III. MHV Testing: Pressure Data ............................................................................................... 8 IV. MHV Testing: Particle Image Velocimetry (PIV) ................................................................ 9 V. Pressure and Flow Data Analysis ........................................................................................ 10 Results ........................................................................................................................................... 14 A. Hydrodynamic Performance ............................................................................................... 14 B. Flow Performance ............................................................................................................... 21 Discussion ..................................................................................................................................... 22 I. Design & Maintenance Constraints ...................................................................................... 23 II. Fabrication Constraints ....................................................................................................... 23 III. Experimental Setup Constraints ......................................................................................... 24 Conclusions ................................................................................................................................... 24 Future Works ............................................................................................................................... 25 Safety ............................................................................................................................................. 25 I. Safety when 3D Printing Filament ....................................................................................... 25 II. Use of Laser ........................................................................................................................ 26 III. Use of Test Section & Mock Circulation Loop ................................................................. 26 Acknowledgements ...................................................................................................................... 26 Cost Analysis ................................................................................................................................ 26 I. Material Cost ........................................................................................................................ 26 II. Equipment Cost ................................................................................................................... 27 References ..................................................................................................................................... 28 Appendix ....................................................................................................................................... 31 Appendix A. BME 198A Proposal .......................................................................................... 31 Appendix", "Constraints ...................................................................................... 23 II. Fabrication Constraints ....................................................................................................... 23 III. Experimental Setup Constraints ......................................................................................... 24 Conclusions ................................................................................................................................... 24 Future Works ............................................................................................................................... 25 Safety ............................................................................................................................................. 25 I. Safety when 3D Printing Filament ....................................................................................... 25 II. Use of Laser ........................................................................................................................ 26 III. Use of Test Section & Mock Circulation Loop ................................................................. 26 Acknowledgements ...................................................................................................................... 26 Cost Analysis ................................................................................................................................ 26 I. Material Cost ........................................................................................................................ 26 II. Equipment Cost ................................................................................................................... 27 References ..................................................................................................................................... 28 Appendix ....................................................................................................................................... 31 Appendix A. BME 198A Proposal .......................................................................................... 31 Appendix B. 198 B Technical Memos ..................................................................................... 31 Appendix C. Code used to create our Flow Data ..................................................................... 31 3 Introduction Valvular heart disease is a common yet lesser-known disease that affects 2.5% of the US population (Azari et al., 2020) and over 74 million people worldwide (Aluru et al., 2022). As the population ages, this issue becomes increasingly prevalent. One typical treatment option for patients with aortic valve diseases is surgical aortic valve repair with a rigid mechanical heart valve (MHV). The current rigid heart valves are durable but require lifelong anticoagulation therapy, due to their non-physiological flow patterns, posing unacceptable risks for certain patient groups (Harris et. al, 2015). The broad direction of current research is focused on improving MHVs to mimic the flow patterns of natural heart valves more closely. Researchers aim to preserve the durability of MHVs while removing the need for long-term use of anticoagulants. Iterative design and evaluation through rapid prototyping is crucial in advancing treatment modalities for valvular heart disease. To isolate the effects of design changes on performance, later-stage prototypes typically use the same materials as commercial valves: pyrolytic carbon (PyC) for the leaflets, and PyC-coated graphite, titanium or stainless steel alloys for the housing. However, these materials are not feasible for early-stage prototypes due to high setup costs for tools and equipment, or time and costs required to coordinate with external manufacturers if fabrication is not done internally. Early prototypes are usually fabricated through 3D printing, CNC machining, and injection molding. Among these, 3D printing offers significant advantages in rapid prototyping, cost-effectiveness, and customizability. Despite these benefits, many of the 3D printing materials used by research teams do not have similar mechanical properties to commercial valve materials (Wang et al., 2022; Krenin et al., 2024), highlighting the need for better 3D printing materials that can more closely replicate the properties of pyrolytic carbon and other metals used in commercial valves. To address", "fabricated through 3D printing, CNC machining, and injection molding. Among these, 3D printing offers significant advantages in rapid prototyping, cost-effectiveness, and customizability. Despite these benefits, many of the 3D printing materials used by research teams do not have similar mechanical properties to commercial valve materials (Wang et al., 2022; Krenin et al., 2024), highlighting the need for better 3D printing materials that can more closely replicate the properties of pyrolytic carbon and other metals used in commercial valves. To address this, our team is investigating the suitability of PC for heart valve prototyping, focusing on their potential to improve prototyping effectiveness and material properties. By leveraging these materials, we aim to create prototypes that more closely mimic the performance of commercial valves while reducing the costs and complexities associated with traditional manufacturing methods. This project will initially evaluate the flow performance of these materials, with future prospects to assess thrombogenicity and other critical factors. To fabricate prototypes from PC, our team 3D printed components from a SolidWorks model of a St. Jude Medical bi-leaflet prosthesis. To test the hemodynamic performance of the prototypes, a pulse duplicator and a custom mock circulation loop (MCL) system were used to simulate the cardiac cycle in the aorta. The pressure gradient and flow velocity were measured across three prototype valves and one St. Jude Medical prosthesis reference valve. With this information, the effective orifice area (EOA) and total regurgitant fraction (RF) were calculated for each valve prototype, and they were compared to those given by ISO 5840-2 and to the metrics of the reference valve. This comparison will determine if the prototypes meet regulatory requirements and if their flow performance is at least equivalent to that of the standard valve. 4 Literature Review Mechanical heart valves are life saving solutions used to replace native valves for patients with valvular disease. Pyrolytic carbon (PyC), the current industry standard for MHVs, contains material properties with excellent durability and hemocompatibility, making the material highly effective for long-term use. Nevertheless, the high fabrication cost of PyC MHV, costs starting at $3,000, poses challenges and limitations in rapid prototyping (Azari, 2020). The fabrication process of MHVs from PyC uses CNC machining, a manufacturing method where pre-programmed computer software is given directions to precisely control a given tool, to cut out components with great precision. Vapor depositioning is then used to evaporate the surface of the device to create a", "durability and hemocompatibility, making the material highly effective for long-term use. Nevertheless, the high fabrication cost of PyC MHV, costs starting at $3,000, poses challenges and limitations in rapid prototyping (Azari, 2020). The fabrication process of MHVs from PyC uses CNC machining, a manufacturing method where pre-programmed computer software is given directions to precisely control a given tool, to cut out components with great precision. Vapor depositioning is then used to evaporate the surface of the device to create a smooth surface (Fraczek-Szczypta et. al, 2023). These tools and processes can cost thousands of dollars for a single valve depending on the manufacturer, making testing slow from a limited budget. By identifying potential alternative materials and fabrication methods for early prototyping, these costs will be greatly reduced. From a review of recent literature, polycarbonate was selected for investigation due to its biocompatibility and established credentials in medical applications. PC-ISO polycarbonate meets ISO 10993 and USP Class VI standards (US Food and Drug Administration, 2020). While polycarbonate alone hasn\u2019t been widely investigated for rigid heart valves, polycarbonate urethane materials have shown recent, promising results in polymeric heart valve applications. For example, Robinson et al. used a combination of Bionate\u00ae Thermoplastic Polycarbonate Polyurethane (PCU) and a hydrogel coating to develop a heart valve that met ISO 5840-2:2021 requirements, demonstrating an effective orifice area of 1.52 \u00b1 0.34 cm 2 and a regurgitation fraction of 9.6 \u00b1 1.8% (2024). Rezvova et al. also identified that the top two most promising materials for polymeric heart valves are POSS-PCU and Hastalex (FGO-PCU), describing that heart valves made from these materials have shown biocompatibility and similar performance to commercial valves through in vitro testing (2023). FDA Considerations Our proposed mechanical heart valve design is based on the FDA-approved St. Jude Mechanical heart valve design, a Class III device under premarket approval with the product code LWQ under 21 CFR 870.3925. Our device is intended to be used solely for non-clinical laboratory research and bench-top testing. The prototype valve shares similar structural and fundamental characteristics with the established St. Jude valve design. The use of 3D printed polycarbonate in the replacement of pyrolytic carbon led to design changes and material changes. The approach towards this device is to see if rapid prototyping can be a valuable, low-risk framework in early-stage development and mechanical testing. Biomedical Motivation Researching these rigid polymeric alternatives is beneficial for not only reducing", "used solely for non-clinical laboratory research and bench-top testing. The prototype valve shares similar structural and fundamental characteristics with the established St. Jude valve design. The use of 3D printed polycarbonate in the replacement of pyrolytic carbon led to design changes and material changes. The approach towards this device is to see if rapid prototyping can be a valuable, low-risk framework in early-stage development and mechanical testing. Biomedical Motivation Researching these rigid polymeric alternatives is beneficial for not only reducing thrombogenicity but also offering cost-effective alternatives and proving great potential in 5 customizability. These factors tie into creating a more affordable valve that offers a temporary solution for lower (SES) individuals who cannot afford valves made from costly materials such as graphite and pyrolytic carbon. Additionally, these findings can be further developed by future researchers who can blend composite or hybrid materials into the polymers to enhance the performance and biocompatibility to meet the standards required for clinical application. Disparities in SES is an extensively examined subject in the health scene. Patients in the lower SES bracket have been observed to have poor health literacy as a result of expensive medicines, treatment, and out-of-pocket expenditures from hospital travel fees which make them opt for lower quality remedies (Prinja et al., 2019). In the valve replacement field, lower SES patients have been observed to partake in surgical aortic valve replacement (SAVR) rather than transcatheter aortic valve replacement (TAVR) (Brlecic et al., 2023). TAVR is a newer and less invasive procedure that uses biological tissue to replace the valve. This type of procedure is more readily available in metropolitan areas, despite having a larger aortic valve replacement (AVR) population in rural areas (Nathan et al., 2022). If our PC and PEEK MHVs prove to perform better or on par as MHVs in the market, a cheaper valve that has comparable quality can be offered as an affordable alternative for these lower SES populations. Furthermore, 3D printing has progressed to become an advanced manufacturing technique that enables the precise fabrication of complex components in medical devices for parts that need accurate and intricate detail. By successfully demonstrating our valves can be 3D printed with the selected materials, the potential for limitless customization possibilities are unlocked. Custom-designed valves can be tailored to meet a patient\u2019s anatomical needs for a flesh fit while reducing complications such as infection or prosthetic failure (Bhandari et al., 2023).", "populations. Furthermore, 3D printing has progressed to become an advanced manufacturing technique that enables the precise fabrication of complex components in medical devices for parts that need accurate and intricate detail. By successfully demonstrating our valves can be 3D printed with the selected materials, the potential for limitless customization possibilities are unlocked. Custom-designed valves can be tailored to meet a patient\u2019s anatomical needs for a flesh fit while reducing complications such as infection or prosthetic failure (Bhandari et al., 2023). Additionally, surgeons can visualize the size and placement of the valve better. This eliminates the need for frequent hospital visits and increases patient safety. Statement of Need Valvular heart disease continues to be a major contributor to morbidity globally, yet the development cycle for MHVs is long, costly, and relies almost exclusively on expensive machinery and materials like pyrolytic carbon and graphite. Current manufacturing methods limit early-stage design iteration, making it difficult to quickly evaluate various geometries and flow profiles. There is an urgent need for a rapid-prototyping platform that can produce functional MHV prototypes at a lower cost and lead time. 3D printing polycarbonate offers a promising alternative since it is a durable and biocompatible material that can be printed in fine detail. Enabling the production of multiple valve designs, shortening the iteration cycle. 3D-printed polycarbonate valves for early-stage performance testing enable research teams to accelerate identifying the most effective geometries and materials, address key gaps in understanding hemodynamic performance, and reduce both development time and costs. Rapid, low-cost prototyping of polycarbonate MHVs will therefore be instrumental in advancing solutions for valvular heart disease and ultimately improving patient outcomes. 6 Materials & Methods I. Mechanical Heart Valve Fabrication using 3D Printer and Assembly A. Materials 1. Bambu X1C 3D printer 2. Polymaker PolyMax PC filament 3. Bambu textured PEI plate 4. Nano Polymer Adhesive 5. Filament dehydrator 6. Sand paper 7. Rubberbands B. Methods The valve components were 3D printed in-house using the Bambu X1C, utilizing the PolyMax PC filament from Polymaker (PC02001). The printer\u2019s settings used the \u201cGeneric PC\u201d preset for the polycarbonate components. The leaflets and supporting ring were printed on top of a Bambu Lab engineering build plate that can withstand the high temperatures needed for extruding the polymer. The plate was covered with Vision Miner Nano Polymer Adhesive to prevent the parts from warping. After the heart valve components are 3D printed, the leaflet and", "printed in-house using the Bambu X1C, utilizing the PolyMax PC filament from Polymaker (PC02001). The printer\u2019s settings used the \u201cGeneric PC\u201d preset for the polycarbonate components. The leaflets and supporting ring were printed on top of a Bambu Lab engineering build plate that can withstand the high temperatures needed for extruding the polymer. The plate was covered with Vision Miner Nano Polymer Adhesive to prevent the parts from warping. After the heart valve components are 3D printed, the leaflet and supporting ring are inspected for any defects, such as warping, uneven surfaces, and \u201cwet\u201d print layers. The prototype valve was assembled by gently pressing on the supporting ring and inserting the leaflets into the hinges. It was manually tested for free motion of the leaflets and motion defects in the components due to printing variability. If the leaflets did not move freely when the valve was rotated, the leaflets were removed from the support ring. The inner sides near the hinge and the sides of the leaflets were sanded down as needed, using increasing grits of sandpaper to create a smooth surface finish. The valve was reassembled and retested until the leaflets could move freely. Rubber bands were placed around a mimicking suture ring to secure it in place, if need to prevent movements II. MHV Testing: Hydrodynamic Performance A. Materials 1. Vivitro Labs SuperPump AR Series 2. CardioLab mock circulation loop 3. Transonic Tubing Flow Module 4. Transonic ME 20 PXL flow sensor 5. Spirit level 6. Blood analogue 7. LabChart software B. Methods 7 Figure 1. Polycarbonate MHV in the sylgard test section with the Transonic tubing flow sensor. The experimental setup closely followed the guidelines set by ISO 5840-2:2021 for evaluating the hydrodynamic performance of a heart valve. The Vivitro SuperPump AR series pulsatile pump was set to a pre-programmed physiologic waveform with a heart rate (HR) of 60 BPM. A mechanical heart valve was inserted into the MCL chamber, and the chamber was attached to the rest of the loop using CardioLab\u2019s standard operating procedure. The MCL was filled with blood analog and straightened out with a spirit level. After the system was powered, the dial on the pulse duplicator was slowly raised until it reached stroke volumes of 50 or 70 mL/stroke for each heart rate. The system was left run for at least 10 consecutive cardiac cycles before being terminated by turning the dial", "the MCL chamber, and the chamber was attached to the rest of the loop using CardioLab\u2019s standard operating procedure. The MCL was filled with blood analog and straightened out with a spirit level. After the system was powered, the dial on the pulse duplicator was slowly raised until it reached stroke volumes of 50 or 70 mL/stroke for each heart rate. The system was left run for at least 10 consecutive cardiac cycles before being terminated by turning the dial back down to 0 mL/stroke. To collect data, the MCL chamber contains sensors that measure the left ventricular pressure before the valve, aortic pressure after the valve, and the pressure difference across the valve. Additionally, a Transonic tubing flow sensor was attached to the loop to measure the volumetric flow rate. The pressures over time were collected in a text file using previous CardioLab code in LabView. The corresponding flow rate over time was collected in LabChart and exported as a separate text file. This process was repeated for each heart rate and stroke volume combination to cover a range of physiological states. Additionally, the experiment was repeated on one prototype valve and one St. Jude Medical reference valve. III. MHV Testing: Pressure Data A. Materials 1. Vivitro Labs SuperPump AR Series 2. CardioLab mock circulation loop 3. Sylguard Test Section 4. Spirit level 5. Blood analogue 8 6. ADInstruments PowerLab 4/26 data acquisition hardware 7. LabView software B. Methods The MHV prototype was positioned in the sylgard test section and straightened with the spirit level once inserted into the MCL. The SuperPump was run for several minutes to let the blood analogue fill into the test section and eliminate as many air bubbles as possible. The pump closely replicated the following conditions: \u25cf 60 BPM, 50 mL/stroke, 140 Peak pressure \u25cf 60 BPM, 70 mL/stroke, 140 Peak pressure \u25cf 60 BPM, 50 mL/stroke, 180 Peak pressure Data acquisition was performed with the ADInstruments PowerLab 4/26 data acquisition hardware system, with visualization and analysis carried out by LabView software. 3 runs for each condition were conducted for a duration of 10 cardiac cycles. IV. MHV Testing: Particle Image Velocimetry (PIV) A. Materials 1. Vivitro Labs SuperPump AR Series 2. CardioLab mock circulation loop 3. Sylguard Test Section 4. Spirit level 5. Blood analogue, using salt and glycerin 6. Silver coated seeding particles 7. Vision Research Phantom v2640 high-speed camera 8.", "performed with the ADInstruments PowerLab 4/26 data acquisition hardware system, with visualization and analysis carried out by LabView software. 3 runs for each condition were conducted for a duration of 10 cardiac cycles. IV. MHV Testing: Particle Image Velocimetry (PIV) A. Materials 1. Vivitro Labs SuperPump AR Series 2. CardioLab mock circulation loop 3. Sylguard Test Section 4. Spirit level 5. Blood analogue, using salt and glycerin 6. Silver coated seeding particles 7. Vision Research Phantom v2640 high-speed camera 8. DaVis software 9. PIV laser 10. Laser goggles B. Methods The same setup and conditions covered in the pressure data procedure were used. Silver coated seeding particles were placed in the MCL and thoroughly distributed throughout the whole system before starting the experiment.The Vision Research Phantom v2640 high-speed camera was positioned to clearly capture both the movement of the valve during opening and closing, as well as the central region of the test section where flow dynamics are visible (see Figure 2 ). A mask was applied using the DaVis software to capture clear and relevant areas of the test section. Before starting the PIV laser, participants were required to wear laser goggles for eye protection. Additional safety precautions for laser handling are listed under the Safety section of this paper. Our technical advisor operated the PIV laser system for each condition. A short recording of the valve in operation was captured. The start and end of 1 cardiac beat was identified and saved for data processing. The color 9 map, arrow length, and arrow thickness were adjusted as needed to give a clear representation of the flow\u2019s performance. Figure 2. The grey-scale image (left) pictures the masking of our area of interest taken during PIV. The second image (right) shows the output post-processing. V. Pressure and Flow Data Analysis A. Materials 1. LabChart software 2. Microsoft Excel software 3. MATLAB software B. Methods The goal of data analysis was to compare the performance of the prototype heart valve to that of a 27mm St. Jude Medical Regent MHV. To evaluate the valve\u2019s ability to allow for unimpeded blood flow during systole, the average effective orifice area (EOA) was calculated. To assess the valve\u2019s ability to prevent backflow during diastole, the regurgitant fraction was determined. The following equations were used. \ud835\udc38\ud835\udc42\ud835\udc34 ( \ud835\udc50 \ud835\udc5a 2 ) = \ud835\udc44 \ud835\udc45\ud835\udc40\ud835\udc46 ( \ud835\udc5a\ud835\udc3f / \ud835\udc60 ) 51 . 6 \u2206 \ud835\udc43 (", "to compare the performance of the prototype heart valve to that of a 27mm St. Jude Medical Regent MHV. To evaluate the valve\u2019s ability to allow for unimpeded blood flow during systole, the average effective orifice area (EOA) was calculated. To assess the valve\u2019s ability to prevent backflow during diastole, the regurgitant fraction was determined. The following equations were used. \ud835\udc38\ud835\udc42\ud835\udc34 ( \ud835\udc50 \ud835\udc5a 2 ) = \ud835\udc44 \ud835\udc45\ud835\udc40\ud835\udc46 ( \ud835\udc5a\ud835\udc3f / \ud835\udc60 ) 51 . 6 \u2206 \ud835\udc43 ( \ud835\udc5a\ud835\udc5a\ud835\udc3b\ud835\udc54 ) \u03c1 ( \ud835\udc54 / \ud835\udc50 \ud835\udc5a 3 ) Equation 1. Effective orifice area equation (Chakraborty et al., 2024). \ud835\udc45\ud835\udc39 = \ud835\udc49 \ud835\udc5f\ud835\udc52\ud835\udc61\ud835\udc5f\ud835\udc5c\ud835\udc54\ud835\udc5f\ud835\udc4e\ud835\udc51\ud835\udc52 ( \ud835\udc5a\ud835\udc3f / \ud835\udc60\ud835\udc61\ud835\udc5f\ud835\udc5c\ud835\udc58\ud835\udc52 ) \ud835\udc46\ud835\udc49 ( \ud835\udc5a\ud835\udc3f / \ud835\udc60\ud835\udc61\ud835\udc5f\ud835\udc5c\ud835\udc58\ud835\udc52 ) \u00d7 100%Equation 2. Regurgitant fraction equation (Chakraborty et al., 2024). From these equations, the the root mean square forward flow rate ( Q RMS ), average differential pressure ( \u0394 P ), and retrograde flow volume ( V retrograde ) were calculated based on pressure and flow data. From data collection, the files with volumetric flow rates over time were saved in LabChart and exported as text files. For each of the three flow conditions, and for both the prototype and control valves, the six total text files were copied over onto Excel as separate sheets to consolidate the data into one file. The 10 procedure for copying over the data to Excel was repeated for the measured pressures over time, in a second spreadsheet file. Using the organized data, six graphs of the volumetric flow rate over time were plotted in MATLAB, along with six graphs of the differential, aortic, and left ventricular pressure over time. For the graphs of the volumetric flow rate, the root mean square forward flow rate ( Q RMS ) was calculated since it is necessary for the effective orifice area (EOA), and the retrograde flow volume ( V retrograde ) was calculated for use in determining the regurgitant fraction (RF). The Q RMS was calculated by manually selecting the range where a positive flow rate was observed during systole, squaring every flow rate, taking the average of the data, and taking the square root of the average. Figure 3 demonstrates the range used for the first Q RMS calculation, using flow rates collected from a polycarbonate prototype when the mock circulation loop was set to create a waveform with a 60 BPM heart rate, 70 mL stroke volume,", "fraction (RF). The Q RMS was calculated by manually selecting the range where a positive flow rate was observed during systole, squaring every flow rate, taking the average of the data, and taking the square root of the average. Figure 3 demonstrates the range used for the first Q RMS calculation, using flow rates collected from a polycarbonate prototype when the mock circulation loop was set to create a waveform with a 60 BPM heart rate, 70 mL stroke volume, and 140 mmHg peak pressure. Figure 3. Range selection for the root mean square forward flow ( Q RMS ) . This process was repeated for each valve and flow condition. For the St. Jude Medical Regent mechanical heart valve, V retrograde was calculated by manually selecting the range where the observed flow rate was negative. For the polycarbonate prototype valves, the selected range started at the first negative flow rate that was observed, and it ended at the first local maximum flow rate. Figure 4 11 demonstrates an example of the range used for the first V retrograde calculation, demonstrating the difference between range selection for the commercial valves and the prototype valves. To calculate the retrograde flow volume, the area under the curve was estimated using the trapezoid method. Figure 4. Range selection for the retrograde flow volume ( V retrograde ), comparing the process for the control valve (left) and the prototype valve (right) . The process was repeated to calculate the Q RMS and V retrograde values for all 8 cardiac cycles of the control valve flow rate data. It was also repeated for all 10 cycles of the prototype valve flow data, though only the first 8 values were used. Next, the average differential pressure ( \u0394 P ) during ejection was calculated for each flow condition and heart valve for determining the EOA. Ejection is expected to start when the pressure in the left ventricle spikes and exceeds the aortic pressure since it causes the heart valve to open, and it\u2019s expected to end when the pressure in the left ventricle decreases to the point when it is below the aortic pressure. Based on this principle, corresponding time ranges were manually selected, and the average of the differential pressures was calculated. Figure 5 demonstrates the method below. This process was also repeated for 8 cardiac cycles of commercial valve data, and 10 cardiac cycles", "ventricle spikes and exceeds the aortic pressure since it causes the heart valve to open, and it\u2019s expected to end when the pressure in the left ventricle decreases to the point when it is below the aortic pressure. Based on this principle, corresponding time ranges were manually selected, and the average of the differential pressures was calculated. Figure 5 demonstrates the method below. This process was also repeated for 8 cardiac cycles of commercial valve data, and 10 cardiac cycles of prototype valve data. However, only the first 8 values were used to calculate averages. 12 Figure 5. Range selection for the average differential pressure ( \u0394 P ) . It was based on the left ventricular pressure (yellow), but values from the differential pressure (blue) were used in calculations. This process was repeated for each valve and flow condition. Using the calculated terms, the average effective orifice area and the regurgitant fraction were determined for the six permutations of flow conditions and mechanical heart valves, using values from the first 8 cardiac cycles. Along with each term, a density of 1.060 g/cm 3 was used for the blood analog, and the stroke volume was 50 or 70 mL/stroke depending on the flow condition. The average of each metric was then calculated along with its 95% confidence interval. Three double-sided, 2-sample T tests were performed to compare each prototype to the commercial valve. Results of the calculations are shown in the upcoming section. 13 Results A. Hydrodynamic Performance Figure 6. Pressure and volumetric flow rate data from a polycarbonate prototype valve running in the mock circulation loop at a heart rate of 60 BPM, stroke volume of 50 mL, and peak pressure of 140 mmHg. The first eight cardiac cycles are shown. 14 Figure 7. Pressure and volumetric flow rate data from a polycarbonate prototype valve running in the mock circulation loop at a heart rate of 60 BPM, stroke volume of 70 mL, and peak pressure of 140 mmHg. The first eight cardiac cycles are shown. 15 Figure 8. Pressure and volumetric flow rate data from a polycarbonate prototype valve running in the mock circulation loop at a heart rate of 60 BPM, stroke volume of 70 mL, and peak pressure of 180 mmHg. The first eight cardiac cycles are shown. 16 Figure 9. Pressure and volumetric flow rate data from a 27mm St. Jude Medical Regent valve running", "volume of 70 mL, and peak pressure of 140 mmHg. The first eight cardiac cycles are shown. 15 Figure 8. Pressure and volumetric flow rate data from a polycarbonate prototype valve running in the mock circulation loop at a heart rate of 60 BPM, stroke volume of 70 mL, and peak pressure of 180 mmHg. The first eight cardiac cycles are shown. 16 Figure 9. Pressure and volumetric flow rate data from a 27mm St. Jude Medical Regent valve running in the mock circulation loop at a heart rate of 60 BPM, stroke volume of 50 mL, and peak pressure of 140 mmHg. The first seven cardiac cycles are shown. 17 Figure 10. Pressure and volumetric flow rate data from a 27mm St. Jude Medical Regent valve running in the mock circulation loop at a heart rate of 60 BPM, stroke volume of 70 mL, and peak pressure of 140 mmHg. The first seven cardiac cycles are shown. 18 Figure 11. Pressure and volumetric flow rate data from a 27mm St. Jude Medical Regent valve running in the mock circulation loop at a heart rate of 60 BPM, stroke volume of 70 mL, and peak pressure of 180 mmHg. The first eight cardiac cycles are shown. Table 1. Average EOA for a polycarbonate prototype and a St. Jude Medical Regent Reference MHV under different flow conditions Flow Conditions EOA (PolyC) \u00b1 95% CI EOA (SJM Regent) \u00b1 95% CI 95% CI of the Difference P Value 19 60 BPM, 50 mL SV, 140 mmHg PP 0.8029 \u00b1 0.0099 0.7324 \u00b1 0.0076 < 0.001 60 BPM, 70 mL SV, 140 mmHg PP 1.1859 \u00b1 0.0924 1.0852 \u00b1 0.0089 0.037 60 BPM, 70 mL SV, 180 mmHg PP 0.9098 \u00b1 0.0024 0.8446 \u00b1 0.1114 < 0.001 Table 2. Average EOA for a polycarbonate prototype and a St. Jude Medical Regent Reference MHV under different flow conditions Flow Conditions RF (PolyC) \u00b1 95% CI RF (SJM Regent) \u00b1 95% CI 95% CI of the Difference P Value 60 BPM, 50 mL SV, 140 mmHg PP 10.5038 \u00b1 0.5434 8.6210 \u00b1 1.1453 0.006 60 BPM, 70 mL SV, 140 mmHg PP 9.1637 \u00b1 0.9363 8.2242 \u00b1 1.3878 0.210 60 BPM, 70 mL SV, 180 mmHg PP 10.7913 \u00b1 0.6270 7.3470 \u00b1 0.8527 < 0.001 20 B. Flow Performance To verify if our mechanical heart valve would perform similar to the standard, a flow performance", "CI RF (SJM Regent) \u00b1 95% CI 95% CI of the Difference P Value 60 BPM, 50 mL SV, 140 mmHg PP 10.5038 \u00b1 0.5434 8.6210 \u00b1 1.1453 0.006 60 BPM, 70 mL SV, 140 mmHg PP 9.1637 \u00b1 0.9363 8.2242 \u00b1 1.3878 0.210 60 BPM, 70 mL SV, 180 mmHg PP 10.7913 \u00b1 0.6270 7.3470 \u00b1 0.8527 < 0.001 20 B. Flow Performance To verify if our mechanical heart valve would perform similar to the standard, a flow performance can be viewed with the particle image velocimetry testing, as shown in Figures 12 and 13 . Doing so enabled us to observe the flow of the blood analong throughout the cardiac cycle like in Figure 12 where we can see the systolic pressure of the blood analog flowing through the valve. At the center, a light jet stream can be seen forming. Figure 13 shows us how the flow at diastolic pressure with the presence of regurgitant or back flow of analog going through the valve, appearing a light yellow on the far center right of the image. Figures 12 & 13. Flow velocity plot of PC MHV at physiologic conditions of 60 BPM, 50 mL stroke volume, and 140 mmHg peak pressure at systolic pressure (left) and diastolic pressure (right). Figure 15 & 16. Flow velocity plot of PC MHV at physiologic conditions of 60 BPM, 70 mL 21 stroke volume, and 140 mmHg peak pressure at systolic pressure (systolic) and diastolic pressure (right). Figure 17. St. Jude at physiologic conditions of 60 BPM, 70 mL stroke volume, and 140 mmHg peak pressure at systolic pressure. Figure 18 & 19. Flow velocity plot of PC heart valve at physiologic conditions of 60 BPM, 70 mL stroke volume, and 180 mmHg peak pressure at systolic pressure (left) and diastolic pressure (right). Discussion The results obtained from this experiment provide valuable insights into the usability and behavior of PC through various physiological conditions. There is not much information on 22 the use of PC for this application, so although we may not have gotten the results we wanted, our study contributes towards other experimentation ideas for future researchers. I. Design & Maintenance Constraints The process of designing and 3D printing our valve, with PC filament, allows us to create a print with dimensions similar to the St. Jude MHV. However, there were limitations and constraints due to 3D printing errors", "various physiological conditions. There is not much information on 22 the use of PC for this application, so although we may not have gotten the results we wanted, our study contributes towards other experimentation ideas for future researchers. I. Design & Maintenance Constraints The process of designing and 3D printing our valve, with PC filament, allows us to create a print with dimensions similar to the St. Jude MHV. However, there were limitations and constraints due to 3D printing errors and with our selected material itself. When not stored in a dry or low moisture environment, the filament would begin to absorb moisture. This absorption of moisture caused the filament to become wet and distorted while being printed, leading to a poor print quality. To prevent and remove excess moisture into the filament, our advisor purchased us a filament dryer to store the filament. After 12 hours of dehydrating, the majority of the moisture was removed and print quality improved. The printer continued to create inconsistencies in print quality. Due to the continuous inconsistencies, we would print 10 duplicates of the same valve for each run of print to compensate, with 2 or 3 valves being passable to assemble and use for testing. Before printing, modifications were made to the original Solidworks file due to differences in leaflet thickness and shallow hinges. When the valve design was first printed, we measured the leaflet thickness to be 1 mm instead of 0.8 mm, like the St. Jude MHV. Later, we observed that at least one leaflet would always dislodge from the hinge once the stroke volume reached a certain threshold. This issue was resolved by making the hinges deeper and extending the tips of the leaflet by 0.1 mm on both sides. With the changes, the MHV was able to stay as a single unit for all the physiological conditions measured. Our team decreased the resolution to 0.08mm to improve the resolution to achieve a smoother finish but manual adjustments, like sanding, needed to be made. We noticed improper maintenance such as not extruding old filament and cleaning the nozzle was not performed. The print quality would be reduced. II. Fabrication Constraints There are some factors that may contribute to the given performance of our PC MHV. During the early stages of testing, we noticed that the flow data for our prototype valve exhibited significantly more noise than the control valve.", "resolution to achieve a smoother finish but manual adjustments, like sanding, needed to be made. We noticed improper maintenance such as not extruding old filament and cleaning the nozzle was not performed. The print quality would be reduced. II. Fabrication Constraints There are some factors that may contribute to the given performance of our PC MHV. During the early stages of testing, we noticed that the flow data for our prototype valve exhibited significantly more noise than the control valve. To identify the cause, we were advised to test both valves while they were fully opened and closed to determine if the problem area was at the hinges or surface roughness. The test revealed no significant difference when the valve is fully opened and closed, indicating the primary source of noise was from friction at the hinges. 3D printing naturally forms small ridges throughout the design. The ridges located in the insides of the hinges could not be sanded down, while the outside surface of the hinge area may not have been sanded down enough. Additionally, some prints may be inconsistent to others, since we noticed defects around the hinges for some valve batches. 23 Another issue we observed was an imbalance in the leaflets. After the second leaflet is inserted, the first leaflet becomes looser regardless of insertion order. This causes a slight flutter that is visible in the PIV videos taken. Further experimentation such as creating a leaflet with increased mass on one side may help address this issue. Overall, the results of our experiment highlight the need for continued refinement in our valve design. III. Experimental Setup Constraints Accurate measurements are essential for analyzing the performance of our valve. While effort is made to optimize our experimental setup, potential for errors cannot be completely eliminated. The presence of air bubbles in the system may disrupt flow patterns, affect sensor readings, and scatter light during PIV data collection. The placement of the flow sensor influences the results as well. Placing it too far up or downstream affects the accuracy of data. In addition, placing the sensor on the curve of a tube generates a source of error. When the system is running, vibration from the pump or external sources may introduce noise to the data. There are errors in the experimental setup for PIV too. The sylgard test section does not appear perfectly clear and has vertical lines", "The placement of the flow sensor influences the results as well. Placing it too far up or downstream affects the accuracy of data. In addition, placing the sensor on the curve of a tube generates a source of error. When the system is running, vibration from the pump or external sources may introduce noise to the data. There are errors in the experimental setup for PIV too. The sylgard test section does not appear perfectly clear and has vertical lines showing as seen in Figure 1 . These vertical lines can interfere with the particle tracking algorithm, producing an inaccurate flow path. Furthermore, shadows and glare are cast in some areas of the images as a result of the lab\u2019s structure. One ceiling light is always left on and has no switch to turn it off. The light interferes with the laser\u2019s path causing these irregularities. By acknowledging these limitations, we can interpret our results better and refine our setup techniques. Conclusions Using rigid polymers to fabricate MHVs via 3D printing is a relatively new research area. While 3D printing technology is gaining momentum in recent years, little is known about its application in the clinical space. Our research in analyzing and evaluating the performance of a polycarbonate bi-leaflet heart valve against an industry standard MHV shows potential, but much work is needed. Results indicate that the effective orifice area was greater for the prototype compared to the commercial valve. However, both were significantly below the minimum threshold specified in ISO 5840. No significant difference in regurgitant fraction was found between the prototype and commercial valves. Both remained within the limit set by the standard as well. Based on this comparison, 3D printed polycarbonate prototypes demonstrate similar flow performance to valves fabricated with industry standard materials and processes. 24 Future Works While the development of our polycarbonate MHV represents a significant step towards analyzing the usability of polycarbonate in rapid prototyping, further research is necessary to optimize or enhance its physiological performance. A critical area for future CardioLab researchers to continue from our project is evaluating the precise measurement of leaflet opening and closing angles in the MHV. Assessing these angles is essential for evaluating the valve\u2019s range of motion and identifying potential design limitations. Such measurements could yield deeper insights into the valve\u2019s mechanical behavior and functionality, potentially guiding the development of solutions to reduce leaflet imbalance in future", "prototyping, further research is necessary to optimize or enhance its physiological performance. A critical area for future CardioLab researchers to continue from our project is evaluating the precise measurement of leaflet opening and closing angles in the MHV. Assessing these angles is essential for evaluating the valve\u2019s range of motion and identifying potential design limitations. Such measurements could yield deeper insights into the valve\u2019s mechanical behavior and functionality, potentially guiding the development of solutions to reduce leaflet imbalance in future MHV designs. Another promising direction involves experimenting with material additives to enhance the stiffness of the polycarbonate used in the MHV. Although polycarbonate is known for its durability, it remains more flexible compared to industry standard materials. This flexibility can result in distortion as stroke volume increases, leading to the formation of a gap near the center of the valve. Incorporating additives such as carbon fiber reinforcement may help mitigate warping and improve overall flow performance. Additionally, the rapid prototyping capabilities of 3D printing offer opportunities to explore a wider range of materials. By leveraging the speed and flexibility of 3D printing, researchers can systematically investigate alternative materials using the experimental procedures established in this study as a guide. These continued efforts will be crucial for advancing MHV design and ensuring optimal performance in future applications. Safety I. Safety when 3D Printing Filament When operating the Bambuu X1C printer, several safety precautions were followed to minimize the risk of injury when using the printer. The printer was set up in a clean, leveled and stable benchtop away from any hazards, such as water, open wires, and chemicals. Due to the risk of release of hazardous fumes, the printer was placed in a well ventilated area. Prior to starting the print, all surrounding individuals were informed of the printer being used, in the case of an unexpected emergency. Strict protocols were followed while the printer was in operation and after finish of print. All personnel were not allowed to open the printer immediately after use and had to wait at least 5 minutes for the print plate to cool before opening the printer, preventing the risk of burns when touching. Because of the small, thin design of our printed parts and the strong adhesive used, we used a scraper to remove our parts from the print plate to prevent any cuts and leaflet lodging into our skin. Gloves were always worn", "of print. All personnel were not allowed to open the printer immediately after use and had to wait at least 5 minutes for the print plate to cool before opening the printer, preventing the risk of burns when touching. Because of the small, thin design of our printed parts and the strong adhesive used, we used a scraper to remove our parts from the print plate to prevent any cuts and leaflet lodging into our skin. Gloves were always worn when washing the printing plate after use to prevent any adhesives from sticking onto our skin. When necessary, the printer was turned off and unplugged from the outlet. The nozzle was completely cool prior to performing maintenance, such as cleaning the inside of the printer and replacing the printer nozzle. Gloves were worn throughout maintenance in order to prevent any risk of leftover filament scraps from being left on the skin. 25 II. Use of Laser A Class 3 Laser was used when collecting PIV images. With our advisor being the only authorized personnel to use the laser, he was present throughout our testing process. Prior to turning on the laser, a blackout curtain was opened to separate us from the laser. UV goggles were worn prior to turning on the laser, and we stayed behind the curtain throughout the duration the laser was on. III. Use of Test Section & Mock Circulation Loop The Mock Circulation Loop is a large set up with multiple tubes, multiple gallons of blood analog solution, and equipment. To ensure safety when moving equipment around and preventing spilling of blood analog, at least 2 individuals were required to be present in the lab at all times. Appropriate PPE including nitrile gloves, and lab coats were worn when handling the blood analog due to the high concentrations of salt and silver coated seeding particles. Acknowledgements Our team wants to acknowledge our technical advisor, Dr. Alessandro Bellofiore, for supporting us throughout our project. Even with his incredibly busy schedule, he inspired us to begin our project, provided feedback on our work, and assisted us with our data collection. We also want to thank Natalia Briseno for creating the test section and blood analog for our setup and data collection. She not only helped us in data collection but also in data processing and teaching us how to use the Da Vinci program. Cost Analysis Reviewing our", "Bellofiore, for supporting us throughout our project. Even with his incredibly busy schedule, he inspired us to begin our project, provided feedback on our work, and assisted us with our data collection. We also want to thank Natalia Briseno for creating the test section and blood analog for our setup and data collection. She not only helped us in data collection but also in data processing and teaching us how to use the Da Vinci program. Cost Analysis Reviewing our budget for this project, when including the cost of equipment and materials required for this project, our total cost is $198.01, listed below in Table 3 . Table 3: Estimated Total Cost Total Cost Overview Estimated Cost Materials $ 89.01 Equipment $109.00 Estimated Overall Cost $198.01 I. Material Cost The overall cost for the project, including shipping cost, is estimated to be approximately $ 89.01. From the listed materials on Table 4 ,used for testing, the cost of each material was based on the given quantity purchased. Any equipment noted with an asterisk (*) are equipment provided by the lab, such as equipment used for printing the filament and performing tests. Thus, costs have been excluded. 26 Table 4. Material Cost Materials Quantity Total Cost Polycarbonate Filament 1 kg $34.11 0.4 mm Hardened Steel Bambu Hotend X1C 1 nozzle $35.99 Nano Polymer Adhesive 50 mL $19.00 Blood Analog with Seeding* n/a n/a Estimated Total Cost $ 89.01 II. Equipment Cost The estimated total cost of materials is estimated to be around $109.00. Table 5 outlines the overall equipment costs. Any equipment noted with an asterisk (*) are equipment provided by the lab, such as equipment used for printing the filament and performing tests. Thus, costs have been excluded. Table 5. Equipment Cost Equipment Cost Filament Dehydrator 109.00 Bambuu X1C Printer* n/a Custom Mock Circulation Loop (MCL)* n/a Total Cost $109.00 27 References 1. Small, A. M., Yutzey, K. E., Binstadt, B. A., Key, K. V., Nabila Bouatia-Naji, Milan, D., Aikawa, E., Otto, C. M., & Hilaire, C. S. (2024). Unraveling the Mechanisms of Valvular Heart Disease to Identify Medical Therapy Targets: A Scientific Statement From the American Heart Association. Circulation, 150(6). https://doi.org/10.1161/cir.0000000000001254 2. Aluru, J. S., Barsouk, A., Saginala, K., Rawla, P., & Barsouk, A. (2022). Valvular Heart Disease Epidemiology. Medical Sciences , 10 (2), 32. https://doi.org/10.3390/medsci10020032 3. Azari, S., Rezapour, A., Omidi, N., Alipour, V., Tajdini, M., Sadeghian, S.,", "B. A., Key, K. V., Nabila Bouatia-Naji, Milan, D., Aikawa, E., Otto, C. M., & Hilaire, C. S. (2024). Unraveling the Mechanisms of Valvular Heart Disease to Identify Medical Therapy Targets: A Scientific Statement From the American Heart Association. Circulation, 150(6). https://doi.org/10.1161/cir.0000000000001254 2. Aluru, J. S., Barsouk, A., Saginala, K., Rawla, P., & Barsouk, A. (2022). Valvular Heart Disease Epidemiology. Medical Sciences , 10 (2), 32. https://doi.org/10.3390/medsci10020032 3. Azari, S., Rezapour, A., Omidi, N., Alipour, V., Tajdini, M., Sadeghian, S., & Bragazzi, N. L. (2020). A systematic review of the cost-effectiveness of heart valve replacement with a mechanical versus biological prosthesis in patients with heart valvular disease. Heart failure reviews , 25 (3), 495\u2013503. https://doi.org/10.1007/s10741-019-09897-9 4. Harris, C., Croce, B., & Cao, C. (2015). Tissue and mechanical heart valves. Annals of cardiothoracic surgery , 4 (4), 399. https://doi.org/10.3978/j.issn.2225-319X.2015.07.01 5. Kreinin, Y., Talmon, Y., Levi, M., Khoury, M., Or, I., Raad, M., Bolotin, G., Sznitman, J., & Korin, N. (2024). A Fibrin-Thrombin Based In Vitro Perfusion System to Study Flow-Related Prosthetic Heart Valves Thrombosis. Annals of Biomedical Engineering , 52 (6), 1665\u20131677. https://doi.org/10.1007/s10439-024-03480-6 6. Chakraborty, S., Simon, M. G., & Bellofiore, A. (2025). Rethinking mechanical heart valves in the aortic position: new paradigms in design and testing. Frontiers in Cardiovascular Medicine, 11, 1458809. https://doi.org/10.3389/fcvm.2024.1458809 7. International Organization for Standardization. (2015). Cardiovascular implants\u2014Cardiac valve prostheses\u2014Part 2: Surgically implanted heart valve substitutes (Standard No. ISO 5840-2:2015). International Organization for Standardization. https://www.iso.org 8. Wang, Z., Dasi, L.P. & Hatoum, H. Controlling the Flow Separation in Heart Valves Using Vortex Generators. Ann Biomed Eng 50, 914\u2013928 (2022). https://doi.org/10.1007/s10439-022-02966-5 9. International Organization for Standardization. (2015). Cardiovascular implants\u2014Cardiac valve prostheses\u2014Part 1: General requirements. (Standard No. ISO 5840-1:2015). International Organization for Standardization. https://www.iso.org 10. Fraczek-Szczypta, A., Kondracka, N., Zambrzycki, M., Gubernat, M., Czaja, P., Pawlyta, M., Jelen, P., Wielowski, R., & Jantas, D. (2023). Exploring CVD Method for 28 Synthesizing Carbon-Carbon Composites as Materials to Contact with Nerve Tissue. Journal of functional biomaterials, 14(9), 443. https://doi.org/10.3390/jfb14090443 11. US Food and Drug Administration. (2020). Use of International Standard ISO 10993-1,\" Biological evaluation of medical devices-Part 1: Evaluation and testing within a risk management process. US Dep Heal Hum Serv Food Drug Adm , 1-68. 12. Robinson, A., Nkansah, A., Bhat, S., Karnik, S., Jones, S., Fairley, A., Leung, J., Wancura, M., Sacks, M., Dasi, L., & Cosgriff-Hernandez, E. (2023). Hydrogel-polyurethane fiber composites with enhanced microarchitectural control for heart valve", "Tissue. Journal of functional biomaterials, 14(9), 443. https://doi.org/10.3390/jfb14090443 11. US Food and Drug Administration. (2020). Use of International Standard ISO 10993-1,\" Biological evaluation of medical devices-Part 1: Evaluation and testing within a risk management process. US Dep Heal Hum Serv Food Drug Adm , 1-68. 12. Robinson, A., Nkansah, A., Bhat, S., Karnik, S., Jones, S., Fairley, A., Leung, J., Wancura, M., Sacks, M., Dasi, L., & Cosgriff-Hernandez, E. (2023). Hydrogel-polyurethane fiber composites with enhanced microarchitectural control for heart valve replacement . https://doi.org/10.1101/2023.09.29.560202 13. Rezvova, M. A., Klyshnikov, K. Y., Gritskevich, A. A., & Ovcharenko, E. A. (2023). Polymeric Heart Valves Will Displace Mechanical and Tissue Heart Valves: A New Era for the Medical Devices. International Journal of Molecular Sciences , 24 (4), 3963. https://doi.org/10.3390/ijms24043963 14. Prinja, S., Sharma, Y., Dixit, J., Thingnam, S. K. S., & Kumar, R. (2019). Cost of Treatment of Valvular Heart Disease at a Tertiary Hospital in North India: Policy Implications. PharmacoEconomics - Open, 3(3), 391\u2013402. https://doi.org/10.1007/s41669-019-0123-6 15. Brlecic, P. E., Hogan, K., Treffalls, J. A., Sylvester, C. B., Coselli, J. S., Moon, M. R., Rosengart, T. K., Chatterjee, S., & Ghanta, R. K. (2023). Socioeconomic disparities in procedural choice and outcomes after aortic valve replacement. JTCVS Open, 16, 139\u2013157. https://doi.org/10.1016/j.xjon.2023.10.002 16. Nathan, A. S., Yang, L., Yang, N., Eberly, L. A., Khatana, S. A. M., Dayoub, E. J., Vemulapalli, S., Julien, H., Cohen, D. J., Nallamothu, B. K., Baron, S. J., Desai, N. D., Szeto, W. Y., Herrmann, H. C., Groeneveld, P. W., Giri, J., & Fanaroff, A. C. (2022). Racial, Ethnic, and Socioeconomic Disparities in Access to Transcatheter Aortic Valve Replacement Within Major Metropolitan Areas. JAMA Cardiology, 7(2), 150. https://doi.org/10.1001/jamacardio.2021.4641 17. Bhandari, S., Yadav, V., Ishaq, A., Sailakshmn Sanipini, Chukwuyem Ekhator, Rafeef Khleif, Alee Beheshtaein, Jhajj, L. K., Aimen Waqar Khan, Khalifa, A. A., Muhammad Arsal Naseem, Bellegarde, S. B., & Muhammad Ather Nadeem. (2023). Trends and Challenges in the Development of 3D-Printed Heart Valves and Other Cardiac Implants: A Review of Current Advances. Cureus. https://doi.org/10.7759/cureus.43204 29 Appendix Appendix A. BME 198A Proposal 198A Technical MemosAppendix B. 198 B Technical Memos BME 198 Technical MemosAppendix C. Code used to create our Flow Data 4/25 30", "Heart Valves and Other Cardiac Implants: A Review of Current Advances. Cureus. https://doi.org/10.7759/cureus.43204 29 Appendix Appendix A. BME 198A Proposal 198A Technical MemosAppendix B. 198 B Technical Memos BME 198 Technical MemosAppendix C. Code used to create our Flow Data 4/25 30", "Monoleaflet Heart Valve Performance B.S. Biomedical Engineering Program Biomedical Engineering Department Charles W. Davidson College of Engineering BME 198 - Spring 2025 Aaheli Das Sarai Gallardo Samantha Yeo Advised by: Alessandro Bellofiore, Ph.D. Faculty Advisor San Jose State University 0 Table of Contents Table of Contents 1 Authorization 2 Figure List 3 Executive Summary 5 Key Words 5 Literature Review 5 Biomedical Motivation 7 I. Physiology 7 II. Background 8 III. Gaps in Knowledge 9 Statement of Need 10 Materials & Methods 10 I. Design Criteria 10 II. 3D Printed Valve 11 III. Experimental Testing 13 Results 14 I. Pressure Data 14 II. Flow during Systole phases for 50 mL/stroke at 180 PP 16 III. Flow during Systole phases for 70 mL/stroke at 180 PP 17 Discussion 19 I. Hemodynamic Performance Comparison 19 II. Flow Characteristics During Systole Phases 20 III. Calculations involving: EOA and Regurgitation Fraction (RF) 21 IV. Design Considerations and Mechanical Limitations 22 V. Limitations of the In-Vitro Testing Setup 22 Conclusion 23 Future Works 24 Safety 24 I. Safety Issues with Laser 24 Cost Analysis 24 Acknowledgements 25 Appendix 25 References 26 1 Authorization Researcher 1 Researcher 2 Researcher 3 Sarai Gallardo Aaheli Das Samantha Yeo SID: 014928355 SID: 015743611 SID: 014898377 Contact: sarai.gallardo@sjsu.edu Contact: aaheli.das@sjsu.edu Contact: samantha.yeo@sjsu.edu Signature: Signature: Signature: Date: 5/16/25 Date: 5/16/25 Date: 5/16/25 Advisor: Alessandro Bellofiore Contact: alessandro.bellofiore@sjsu.edu 2 Figure List Figure 1 Healthy vs. Diseased Aortic Valve Comparison Figure 2 Ball in a Cage Mechanical Heart Valve Figure 3 Tilted Disk Mechanical Heart Valve Top View Figure 4 Tilted Disk Mechanical Heart Valve Side View Figure 5 Leaflet base with the butterfly hinge design Figure 6 Final leaflet design with the circular shaped rod Figure 7 Valve design 3D printed in PLA filament Figure 8 Valve design printed in PVB Figure 9 Experimental Setup of the mock circulatory loop (MCL) Figure 10 Pressure vs. Time (1st 2 beats) at 50 BPM and 140 PP Figure 11 Pressure vs. Time (1st 2 beats) at 50 BPM and 180 PP Figure 12 Pressure vs. Time (1st 2 beats) at 70 BPM and 180 PP Figure 13 Early Systole for flow during Systole phases for 50 mL/stroke at 180 PP Figure 14 Early Systole Streamline for flow during Systole phases for 50 mL/stroke at 180 PP Figure 15 Late Systole for flow during Systole phases for 50 mL/stroke at 180 PP Figure 16 Late", "and 140 PP Figure 11 Pressure vs. Time (1st 2 beats) at 50 BPM and 180 PP Figure 12 Pressure vs. Time (1st 2 beats) at 70 BPM and 180 PP Figure 13 Early Systole for flow during Systole phases for 50 mL/stroke at 180 PP Figure 14 Early Systole Streamline for flow during Systole phases for 50 mL/stroke at 180 PP Figure 15 Late Systole for flow during Systole phases for 50 mL/stroke at 180 PP Figure 16 Late Systole Streamline for flow during Systole phases for 50 mL/stroke at 180 PP Figure 17 Early Systole for flow during Systole phases for 70 mL/stroke at 180 PP Figure 18 Early Systole Streamline for flow during Systole phases for 70 mL/stroke at 180 PP Figure 19 Late Systole for flow during Systole phases for 70 mL/stroke at 180 PP Figure 20 Late Systole Streamline for flow during Systole phases for 70 mL/stroke at 180 PP 3 Executive Summary Valvular heart disease is a condition where one or more of the heart\u2019s four valves do not function properly, disrupting normal blood flow through the heart and to the rest of the body. According to the Centers for Disease Control and Prevention, approximately 5 million people in the U.S. are diagnosed with heart valve disease each year, resulting in nearly 25,000 deaths annually 4 . Currently, mechanical heart valves are engineered to mimic the physiology of human heart valves. However, these devices are prone to complications such as thrombosis and excessive regurgitation. Our lab aims to improve upon current designs by developing a new mechanical heart valve that minimizes thrombosis and regurgitation through improved hemodynamic performance. We tested this hypothesis by combining computational modeling in SolidWorks, 3D prototyping, and in-lab testing. To simulate heart beats and pressure flow, we tested in the lab by placing the 3D-printed monoleaflet prototype in the mock circulatory loop test (MCL). A particle image velocimetry (PIV) was also used to analyze the pressure gradients across the valve. While these approaches provide valuable insight, our current findings are limited by their in-vitro nature. Since the MCL testing is conducted outside the body, they may not fully replicate the in-vivo environment. Future work will focus on further validating our design under more physiologically accurate conditions. Key Words Hemodynamic performance, thrombosis, regurgitation, mock circulatory loop, steady state, anticoagulation, monoleaflet, systole, diastole Literature Review The cardiac cycle is the sequence of", "also used to analyze the pressure gradients across the valve. While these approaches provide valuable insight, our current findings are limited by their in-vitro nature. Since the MCL testing is conducted outside the body, they may not fully replicate the in-vivo environment. Future work will focus on further validating our design under more physiologically accurate conditions. Key Words Hemodynamic performance, thrombosis, regurgitation, mock circulatory loop, steady state, anticoagulation, monoleaflet, systole, diastole Literature Review The cardiac cycle is the sequence of events that occurs to complete one heartbeat increment. These events occur back and forward between the contraction phase (systole), and relaxation phase (diastole) in both the atria and ventricles. The cycle begins with the atrial systole being triggered to contract because of depolarization, resulting in pushing about 20-30% of blood through the tricuspid and mitral valves in the relax ventricles(Anonymous). The valve 4 continues to fill with blood as the ventricular systole is triggered commencing the isovolumic contraction. As the ventricles begin to contract, the pressure rises, which triggers the closure of the tricuspid and mitral valves and produces the first heart sound. Once the chamber of the heart reaches max pressure, the pressure forces the pulmonary and aortic valves to open, allowing the blood to be injected into the major arteries. Next the ventricular diastole begins triggering the ventricles to relax and the polymuray and aortic valves to close causing a second heartbeat sound. Finally, the cycle reaches the last phase which is the late ventricular diastole, here is when the tricuspid and mitral valves reopen as the atrial pressure exceeds the ventricular pressure. As a result, the subtle amount of blood is allowed through the ventricles as the cycle times to an end. The cardiac cycle can be affected by a condition known as valvular heart disease (VHD) in which one or more of the heart's valves do not function properly resulting in a unidirectional flow of the blood during the cardiac cycle. There are many medical conditions that contribute to the deterioration of the valves such as Rheumatic disease, endocarditis, and congenital heart valve disease. Once the valve is compromised, the valve can not close and open properly leading to stenosis or regurgitation, both of which impair the blood circulation. In many cases, damaged heart valves cannot be repaired naturally or through medication alone. As a result, mechanical heart valves are often used as a long-term solution", "blood during the cardiac cycle. There are many medical conditions that contribute to the deterioration of the valves such as Rheumatic disease, endocarditis, and congenital heart valve disease. Once the valve is compromised, the valve can not close and open properly leading to stenosis or regurgitation, both of which impair the blood circulation. In many cases, damaged heart valves cannot be repaired naturally or through medication alone. As a result, mechanical heart valves are often used as a long-term solution to restore proper valve function and maintain efficient blood flow. Mechanical heart valves have been engineered to mimic the physiological function of a human valve so it can serve as a replacement. Currently, there are several mechanical heart valve options on the market such as trileaflet, bileaflet, and tilting disk designs. Each of these valves operates slightly differently than one another to simulate the open and close mechanism during the cardiac cycle. However, the St. Judes bileaflet is the most commonly used one because of its favorable flow characteristics and durability. Despite their advantages, mechanical heart valves still face performance challenges, particularly with a condition known as thrombosis. Thrombosis is the formation of blood clots and it often occurs on the hinges of the mechanical heart valves. At the hinges, there is a lot of turbulence which damages the blood cells and therefore triggers the blood coagulation process causing blood clots to build at the hinges. 5 Mechanical stress also contributes to damaging the cells. Therefore, our project was to design a monoleaflet and try to reduce the amount of stress at the hinges and improve hydrodynamics. Biomedical Motivation I. Physiology To get started it is necessary to understand why there is a need for a mechanical heart valve in the first place. In order to do this, it is crucial to define what it is functioned to do in accordance within the human body. The valve that is being focused on for this project is the aortic heart valve that lies between the left ventricle and the aorta of the heart. In an ideal condition, the blood would be pumped from the left ventricle into the aorta and the aortic valve would only allow for that flow to be one way. However, due to the fact that there are many people suffering from aortic valve disease, that ideal situation might not happen all the time. This disease is a", "focused on for this project is the aortic heart valve that lies between the left ventricle and the aorta of the heart. In an ideal condition, the blood would be pumped from the left ventricle into the aorta and the aortic valve would only allow for that flow to be one way. However, due to the fact that there are many people suffering from aortic valve disease, that ideal situation might not happen all the time. This disease is a very risky condition to have since it allows backflow from the aorta into the valve causing many side effects from shortness of breath all the way to painting. To account for this, two types of artificial heart valves are in the making. One of them is a typical bioprosthetic valve that utilizes another animal's body part and the other is making a mechanical one that doesn\u2019t involve any parts from a living organism. Due to the fact that using heart valves from other animals has its own set of complications, this paper will focus on the motivation behind creating a mechanical heart valve. 6 Figure 1 : Healthy vs. Diseased Aortic Valve Comparison II. Background Knowing that the development of heart valves first started around the 1950s with the ball caged valve, it was evident that researchers and scientists had been working on a solution for quite some time now. In addition to the ball-caged valve, there was also a tilted disk version that was released to the market but had a few complications regarding thrombosis and hemodynamics. Since then, there have been significant advancements with the bi-leaflet heart valves being the most used today, but there has not been much improvements toward it since there are still a lot of problems with the devices that are on the market. Current research focuses on improving blood flow dynamics, reducing thrombogenicity and enhancing hemocompatibility, therefore this new research would help find ways on how to improve on them. 7 Figure 2: Ball in a Cage Figure 3: Tilted Disk Top View Figure 4: Slanted Disk Side View III. Gaps in Knowledge To address current limitations in this area of cardiovascular medicine, it is important to understand where the gaps in knowledge lie.One of the primary challenges is finding a way to prevent thrombosis without the need for anticoagulation medication. Finding a way to achieve this would eliminate the use of blood", "ways on how to improve on them. 7 Figure 2: Ball in a Cage Figure 3: Tilted Disk Top View Figure 4: Slanted Disk Side View III. Gaps in Knowledge To address current limitations in this area of cardiovascular medicine, it is important to understand where the gaps in knowledge lie.One of the primary challenges is finding a way to prevent thrombosis without the need for anticoagulation medication. Finding a way to achieve this would eliminate the use of blood thinners, which then will significantly reduce associated risks and complications that come with it. Another necessary improvement is to improve the hemodynamics for the mechanical heart valve. This is closely linked to thrombogenicity as the 8 greater the thrombus formation then the greater the flow disturbances in results.While these issues are complex, it is achievable if both the design and material of the design is optimized. Statement of Need Valvular heart disease affects millions of patients worldwide and often would require valve replacement to restore normal cardiac function. Current mechanical heart valves are often associated with risks such as thrombosis and lifelong anticoagulation therapy. While bioprosthetic valves offer better biocompatibility, it is not as durable and is not ideal for younger patients. These limitations show a critical need for a mechanical heart valve with reliable performance that reduces the risks of thrombosis and regurgitation. By developing and testing a monoleaflet valve design, we aim to optimize the hemodynamic performance and minimize regurgitant flow. Materials & Methods I. Design Criteria When designing the mechanical heart valve, the main consideration was that the valve needed to open to the maximum of 90\u00b0 for maximum flow rate. This was implemented to the design by creating the hinge to be a butterfly hinge so that it can control the angle of how much the leaflet opens as shown in figure 1. Another important design consideration for the leaflet was to make sure that it was not too loose or tight when connected to the base. This was done by offsetting the valve and the base by a few millimeters. Friction was important to consider as well so that the leaflet would be able to open and close smoothly. The rod with the leaflet started out as a rectangular shape but produced too much friction so it was later changed to a circular shaped rod to help reduce that friction as shown in Figure 6", "was not too loose or tight when connected to the base. This was done by offsetting the valve and the base by a few millimeters. Friction was important to consider as well so that the leaflet would be able to open and close smoothly. The rod with the leaflet started out as a rectangular shape but produced too much friction so it was later changed to a circular shaped rod to help reduce that friction as shown in Figure 6 . All design iterations of the monoleaflet mechanical heart valve were created using SolidWorks 2023, a professional-grade computer-aided design (CAD) software. This platform enabled precise modeling of the valve geometry, including the leaflet, hinge mechanism, and housing structure. One key feature used throughout the design process was the Configurations tool in SolidWorks. This feature allows for there to be multiple variations of the design within a single document. By adjusting individual 9 parameters such as leaflet angle, hinge dimensions, and housing geometry, we were able to explore and evaluate different design approaches efficiently, without affecting the overall model structure. Figure 5: Leaflet base with the butterfly hinge design Figure 6: Final leaflet design with the circular shaped rod 10 II. 3D Printed Valve To validate the CAD designs, each monoleaflet heart valve iteration was fabricated using 3D printing technology. Initial prototypes were printed using PLA (Polylactic Acid), a biodegradable material known for its low melting point and ability to produce smooth, high-quality surface finishes. However, consistent issues arose with the precision of the hinge portion of the valve. To address this, we used a narrower nozzle to enhance detail and accuracy. Despite this adjustment, the PLA filament retained too much moisture, which led to imprecise cuts and dimensional inconsistencies in the hinge area as shown in Figure 7 . As a result, we transitioned to PVB (Polyvinyl Butyral), a drier filament that provided cleaner and more accurate prints as shown in Figure 8 . This change significantly improved the precision of critical features, especially around the leaflet rod and hinge interface. Figure 7: Valve design 3D printed in PLA filament 11 Figure 8: Final valve design printed in PVB III. Experimental Testing To evaluate the hemodynamic performance of the monoleaflet mechanical heart valve,in-vitro testing was conducted using a Mock Circulatory Loop (MCL) system integrated with Particle Image Velocimetry (PIV) analysis. The Mock Circulatory Loop (MCL) was designed to replicate physiological", "in Figure 8 . This change significantly improved the precision of critical features, especially around the leaflet rod and hinge interface. Figure 7: Valve design 3D printed in PLA filament 11 Figure 8: Final valve design printed in PVB III. Experimental Testing To evaluate the hemodynamic performance of the monoleaflet mechanical heart valve,in-vitro testing was conducted using a Mock Circulatory Loop (MCL) system integrated with Particle Image Velocimetry (PIV) analysis. The Mock Circulatory Loop (MCL) was designed to replicate physiological flow and pressure conditions observed in the human heart. The system included a pulsatile pump to simulate cardiac output, compliant tubing to mimic vascular elasticity, and a fluid reservoir as shown in Figure 9 . The valve prototype was positioned within a transparent chamber that allowed direct visualization of flow behavior. A mixture of distilled water and silver-coated glass particles was used to replicate the viscosity of blood, facilitate accurate flow simulation, and enhance optical clarity for recording. A pulsatile pump was strategically placed within the MCL system to simulate rhythmic heartbeats and generate physiologically relevant pressure and flow conditions. A high-resolution camera was positioned above the valve to capture velocity gradients across the flow field. To enable precise visualization, a laser was used to illuminate the seeded fluid mixture, causing the reflective silver-coated particles to scatter light. This setup allowed for clear identification of flow patterns and turbulence occurring during both the systolic and diastolic phases of the cardiac cycle. 12 Particle Image Velocimetry (PIV) was then used to capture video recordings of the velocity gradients generated with each pulse, allowing for detailed analysis of fluid motion throughout the cardiac cycle. Figure 9: Experimental Setup of the mock circulatory loop (MCL) 13 Results I. Pressure Data Figure 10: Pressure vs. Time (1st 2 beats) at 50 BPM and 140 PP Figure 11: Pressure vs. Time (1st 2 beats) at 50 BPM and 180 PP 14 Figure 12: Pressure vs. Time (1st 2 beats) at 70 BPM and 180 PP Figure 10-12 describes the three different pressures (differential, aortic and left ventricular) versus time. The graphs show the pressure of the first two beats. The yellow line represents the differential pressure, orange line represents the aortic pressure and the pink/red line represents the left ventricular pressure as the pump was going 50 and 70 BPM. The use of differential pressure is to assess the flow pressure across the valve and", "Time (1st 2 beats) at 70 BPM and 180 PP Figure 10-12 describes the three different pressures (differential, aortic and left ventricular) versus time. The graphs show the pressure of the first two beats. The yellow line represents the differential pressure, orange line represents the aortic pressure and the pink/red line represents the left ventricular pressure as the pump was going 50 and 70 BPM. The use of differential pressure is to assess the flow pressure across the valve and the resistance of the valve opening and closing. As shown on the graphs, our data for it is significantly higher due to the high resistance of the valve fully opening. The aortic pressure is the pressure in the aorta when the left ventricular ejects the fluid. The left ventricular pressure shows that it is low when in diastole and high when in systole. 15 II. Flow during Systole phases for 50 mL/stroke at 180 PP Figure 13: Early Systole Figure 14: Early Systole Streamline Figure 15: Late Systole Figure 16: Late Systole Streamline 16 Figures 13 -16 show the velocity gradients and streamlines of the flow during 50 BPM at 180 PP. Early systole shows a large amount of forward flow when the valve is opening, specifically a bit at the top and bottom of the leaflet when it opens. But as the arrows show, there is some backflow hitting the back of the leaflet. Streamlines show a lot of circular vortexes as the flow goes forward. During late systole, there is a large amount of forward flow on the top of the leaflet opening meaning much of the flow is going forward at the top opening rather than the bottom. Streamlines also showed lots of backflow and vortexes happening at the bottom of the leaflet opening. The colors correspond to the velocity of the flow rate, blue meaning a low velocity to red meaning a high velocity or flow as the fluid goes through the valve. III. Flow during Systole phases for 70 mL/stroke at 180 PP Figure 17: Early Systole Figure 18: Early Systole Streamline 17 Figure 19: Late Systole Figure 20: Late Systole Streamline Figures 17 - 20 show the velocity gradients and streamline of the flow during 70 BPM at 180 PP. During early systole, while it does show a large amount of forward flow when the leaflet opens, there is also a large amount of", "as the fluid goes through the valve. III. Flow during Systole phases for 70 mL/stroke at 180 PP Figure 17: Early Systole Figure 18: Early Systole Streamline 17 Figure 19: Late Systole Figure 20: Late Systole Streamline Figures 17 - 20 show the velocity gradients and streamline of the flow during 70 BPM at 180 PP. During early systole, while it does show a large amount of forward flow when the leaflet opens, there is also a large amount of regurgitation in the middle of the valve. The streamline also shows less vortexes and large amounts of back flow hitting the back of the leaflet. During late systole, there is a large area of velocity on the top of the valve. While there still is a large area of back flow, it is much less compared to the early systole stage. In the streamline, there is a large area of vortexes in between the forward and backflow velocities because of the low pressure in the area of the flow pushing forward. Discussion I. Hemodynamic Performance Comparison Taking into account the pressure and flow data, it was difficult to conceptualize what would be considered good data to have as there is no other monoleaflet data to compare it to. In order to make this comparison, it was essential to first know what each graph represents. The charts displaying pressure data in Figures 10 - 12 has it split into three different plots 18 overlapping one another to get a better visual of a side by side comparison. The red curve represents the pressure in the left ventricle and the orange curve represents the pressure in the aortic valve. These two plots are later used to calculate and generate the yellow curve that shows the difference in pressure between the two cardiac structures. The prototype was tested under the following conditions: one at 50 BPM and 140 PP, the second at 50 BPM and 180 PP, and the last at 70 BPM and 180 PP. BPM stands for beats per minute whereas PP stands for peak to peak pressure. The reason why it was analyzed at two different BPMs was so that it can be put to the test at a resting heart rate, 50 BPM, and a normal physiological heart rate at 70 BPM. These were later tested at two different peak to peak pressures to examine how they would perform", "BPM and 180 PP, and the last at 70 BPM and 180 PP. BPM stands for beats per minute whereas PP stands for peak to peak pressure. The reason why it was analyzed at two different BPMs was so that it can be put to the test at a resting heart rate, 50 BPM, and a normal physiological heart rate at 70 BPM. These were later tested at two different peak to peak pressures to examine how they would perform under different cardiac conditions. For our case, the valve showed more resistance to opening and closing therefore the pressures were increased. Even though normal aortic systolic pressure is around 120, the valve was tested first at 140 due to that resistance. The valve was then tested at 180 to simulate a high stress environment such as severe hypertension or stenosis. The results show the three sets of pressures at all combinations of beats per minute and peak pressures but one which was a combination of 70 BPM at 140 PP. This was due to a lack of time so the valve was tested in the rest of the pairings. After conducting this experiment, it was realized that the difference in pressure between the left ventricle and aorta was substantially higher at a combination of 50 BPM and 140 PP. This may have been because the valve was having trouble opening all the way since the breathing rate was a lot lower which resulted in a much less intense flow compared to 70 BPM. II. Flow Characteristics During Systole Phases Figures 17 - 20 show the valve under 50 BMP and 70 BPM through the two main stages of systole at a constant 180 PP. This test was done at the same time as the pressure and flow data but to get the results shown above there was a slightly different process used. To get a better look at the flow characteristics during the systole phases, the camera was placed directly on the MCL tube. In addition to that the laser was also used to shine over the camera to properly see where the most intensity lied in the flow by analysing how the particles in the fluid were moving 19 around. The color scale shown on the right side of the graph represents the intensity of the flow where blue is the lowest and red corresponds to the highest intensity.", "characteristics during the systole phases, the camera was placed directly on the MCL tube. In addition to that the laser was also used to shine over the camera to properly see where the most intensity lied in the flow by analysing how the particles in the fluid were moving 19 around. The color scale shown on the right side of the graph represents the intensity of the flow where blue is the lowest and red corresponds to the highest intensity. The graphs that were collected at the different stages of systole were paired up with an identical graph that had streamlines helping us get a better look at the direction of the flow. First we tested it at 50 BPM and it showed an increase in the intensity of the flow as you go from early systole to late systole. However, the results from the 70 BPM had a higher intensity but it accounted for a lot of backflow which was not ideal. This may have been because the valve was able to open and close a lot better than before but since it had no stopper to make it only flow one way, there was a lot more regurgitation than estimated. III. Calculations: EOA and Regurgitation Fraction (RF) Using Figures 10 - 12 , it will help find the EOA and regurgitation fraction (RF) of the valve in 50 beats per minute (BPM) and 70 BPM at 140 peak pressure (PP) and 180 PP. EOA stands for effective orifice area which is the measurement of the opening size of the heart valve when fluid flows through during systole. This valve reflects how well the valve allows forward flow, so the larger the EOA, the less resistance and the better forward flow of it. Using the following equation, it can help us calculated EOA: . Q is the mean \ud835\udc38\ud835\udc42\ud835\udc34 ( \ud835\udc50\ud835\udc5a ) 2 = \ud835\udc44 \ud835\udc45\ud835\udc40\ud835\udc46 51 . 6 \u2206 \ud835\udc43 \u03c1 systolic flow rate (mL/s), \u0394 P is the mean pressure across the valve during ejection of the fluid, and \ud835\udc91 is the density of the fluid. To find Q RMS , find the time intervals from 0 velocity from start to end. Each flow rate was then squared, with the average taken, then was square rooted. \u0394 P was estimated from the pressure data graph in Figures 10 - 12 . Using the two pressure points calculated,", "\u2206 \ud835\udc43 \u03c1 systolic flow rate (mL/s), \u0394 P is the mean pressure across the valve during ejection of the fluid, and \ud835\udc91 is the density of the fluid. To find Q RMS , find the time intervals from 0 velocity from start to end. Each flow rate was then squared, with the average taken, then was square rooted. \u0394 P was estimated from the pressure data graph in Figures 10 - 12 . Using the two pressure points calculated, we were able to find the differential pressure of them. \ud835\udc91 is the density of the fluid in which the valve is being tested at, which is supposedly similar to the density of blood at 1.06. The EOA was then calculated for each cycle, for a total of 10 cycles. The average then resulted in 0.337 \u00b1 0.117 cm 2 for 50 BPM at 180 PP, 0.234 \u00b1 0.116 cm 2 for 50 BPM at 140 PP and 0.419 \u00b1 0.119 cm 2 for 70 BPM at 180 PP. The EOA calculated is much lower than the St. Judes Bileaflet valve when compared as its EOA target is \u22651.7 cm 2 . The main reason why the EOA is 20 much lower for our valve is the fluid pressure of the valve not opening fully to 90\u00b0 as it was designed to be. Regurgitatant fraction (RF) is the percentage of fluid that flows backwards through the valve. The following equation was used to calculate RF: . V retrograde \ud835\udc45\ud835\udc39 = \ud835\udc49 \ud835\udc5f\ud835\udc52\ud835\udc61\ud835\udc5f\ud835\udc5c\ud835\udc54\ud835\udc5f\ud835\udc4e\ud835\udc51\ud835\udc52 \ud835\udc46\ud835\udc49 \ud835\udc65 100% (mL/stroke) is the integral of the regurgitant flow part of the flow graph. SV (mL/stroke) is the stroke volume set for the testing, which is either 50 or 70 mL. The results showed that the RF for 50 BPM at 140 PP was 35.49 \u00b1 0.15%, for 50 BPM at 180 PP it was 30.17 \u00b1 0.19%, and for 70 BPM at 180 PP it was 19.70 \u00b1 0.08%. The RF calculated for the Bileaflet valve was shown to be 15% or less meaning our valve has a severe regurgitation rate, but it did decrease as the BPM increased. IV. Design Considerations and Mechanical Limitations The results that were obtained show how there might have been a couple discrepancies in all the data that was collected. This is due to multiple reasons, the first one being that the 3D printed model was not to par", "was 19.70 \u00b1 0.08%. The RF calculated for the Bileaflet valve was shown to be 15% or less meaning our valve has a severe regurgitation rate, but it did decrease as the BPM increased. IV. Design Considerations and Mechanical Limitations The results that were obtained show how there might have been a couple discrepancies in all the data that was collected. This is due to multiple reasons, the first one being that the 3D printed model was not to par in many ways. The machine that was used to print the prototype was not able to accurately cut out hinges according to the design that was provided because of the nozzle tip being too big and the hinge being too small. Since the tolerance was pretty high, the nozzle would often mess up some parts of the design by making the filament a little too thick or thin at certain points. To make up for this, the valves were printed multiple times to reduce the variability overall. Another idea that could be incorporated in the future is making the base of the valve shorter in length so that it doesn\u2019t make as much noise when the leaflets are brushing against the interior of the base. Not only will following through with this idea reduce the noise but it will also be beneficial in other areas as well. For example, it will lessen the amount of time the base needs to print since there is less material to print. Lastly, decreasing the length of the base will also allow for a better visualization when it comes to collecting data since it was quite difficult to see when and how much the leaflets were moving during testing with the MCL. Researching on other materials to print the prototype is another possibility to help find one with 21 good texture and smoothness so that it won't have to be sandpapered as often and it won\u2019t tear the MCL tube while being placed inside. V. Limitations of the In-Vitro Testing Setup Even though the simulation was set up pretty close to the physiology of the heart, there were still several limitations to the in-vitro testing assembly. For instance, substituting the water for blood is not allowed since it would interfere with the interior of the pump as it is not equipped to deal with real blood. In addition to that, there were several moments", "it won\u2019t tear the MCL tube while being placed inside. V. Limitations of the In-Vitro Testing Setup Even though the simulation was set up pretty close to the physiology of the heart, there were still several limitations to the in-vitro testing assembly. For instance, substituting the water for blood is not allowed since it would interfere with the interior of the pump as it is not equipped to deal with real blood. In addition to that, there were several moments where a lot of fluid was spilled or had the chance of spilling when setting up the valve in the test section, as well as taking it out. To avoid this all together, it would be very helpful if there was more space to work with since all the equipment was set up on the main table and there were many other obstacles like other tables and chairs that were blocking access to the mock circulatory loop set up. Making adjustments to the camera and laser would also help in making more space because it hangs right over the MCL tube making it difficult to take it out and put the liquid inside. Conclusion In conclusion, the monoleaflet valve is still in need of redesigning and continuous testing before it could be brought out into the medical field. Given that the ones that are out on the market right now are bi-leaflet and tri-leaflet, this innovation could potentially be the next one to be released after making a couple more adjustments and refinements to the prototype. To see if this model was a fair competitor, the team had to first design, test, and build an artificial monoleaflet heart valve that abides by a series of constraints in order for it to closely mimic an actual heart valve that would be inside a human body. The prototype was initially designed on SolidWorks and then 3D printed. Afterwards, it was tested in the mock circulatory loop to mimic how it would actually function in a human heart. When testing on pressure, it was observed that the prototype showed that the pressure was too high as the leaflet showed resistance, making it unable to fully open at 90\u00b0 angle as it was intended to. The main issues with the valve were that it wouldn\u2019t fully open or close and it accounted for a lot of regurgitation since the valve would function as a", "the mock circulatory loop to mimic how it would actually function in a human heart. When testing on pressure, it was observed that the prototype showed that the pressure was too high as the leaflet showed resistance, making it unable to fully open at 90\u00b0 angle as it was intended to. The main issues with the valve were that it wouldn\u2019t fully open or close and it accounted for a lot of regurgitation since the valve would function as a two way valve instead of one. This made our flow data display a lot of backflow 22 which is not ideal. For the future, we hope to make changes to the material so that it can open and close a lot more smoothly facing less friction overall. This project is something worth looking into since it's on a new kind of heart valve design that is not researched out in the public. Future Works As stated before there can be many adjustments made to this device that can help improve the structure, quality, and efficiency of the design. One major change that comes to mind is utilizing a more high-resolution 3D printer than the one that was provided. This significant hardware enhancement will help the model achieve finer details and smoother leaflet edges resulting in improved flow performance. In addition to that, substituting the current printing materials with more biocompatible or mechanically superior alternatives could be better to use so that it can more accuratley mimic physiological conditions. Other than that, optimizing the hinge design and reducing friction would ultimately enhance leaflet mobility and endure long-term durability. Safety I. Safety Issues with Laser During the testing of the valve, we utilized the laser to capture flow velocity fields in experimental fluid dynamics. With use of the laser, it can illuminate tracer particles that follow the flow and therefore making it visible for the camera\u2019s images. When the laser is in use, it is supervised under the surveillance of Professor Bellofiore. The laser is a class four meaning that extra attention should be given when the laser is operating. Improper care and procedures can lead to a risk of eye injuries such as blindness as well as skin burns if come in contact with it. Therefore, we prioritize safety before starting the testing procedures. We wore appropriate personal protective equipment (PPE), including a lab coat, a special type of safety goggles", "use, it is supervised under the surveillance of Professor Bellofiore. The laser is a class four meaning that extra attention should be given when the laser is operating. Improper care and procedures can lead to a risk of eye injuries such as blindness as well as skin burns if come in contact with it. Therefore, we prioritize safety before starting the testing procedures. We wore appropriate personal protective equipment (PPE), including a lab coat, a special type of safety goggles during laser operation and gloves. There was also a large black curtain separating us from the laser when it was operating. 23 Cost Analysis The budget covers various aspects, including software licenses, prototyping costs, and testing materials. During the research and development of the monoleaflet design, the following materials are needed: SolidWorks license, filament, and 3D printer. According to the Solidworks plans and pricing, the cost of owning a solidworks license per person is about $2000 a year for the basic package. Therefore, for the entire team to have access to the program will cost about $6000 for a year. The prototyping portion of this project will consist of using a Bambu Lab X1 Carbon and the cost of this device is about $1200. The filaments used were PLA and PVB and the cost for one kilogram was about $19 each. However this material was shared across all the teams who were part of the cardio lab so the cost was split up and it was difficult to keep track of how much each team was using. Because most, if not all, of these items and licences were already provided, we can say that the overall cost was $0.00 since nothing was paid out of pocket for our team. Acknowledgements Sarai Gallardo Executive Summary, Literature Review, Materials and Methods, Discussion Aaheli Das Biomedical Motivation, Results, Discussion, Conclusion, Future Works, Cost Analysis Samantha Yeo Statement of Need, Results, Discussion, Safety Combined Contributions Authorization, Figure Table, Appendix, References 24 Appendix 1. Compilation of the technical memorandums: https://docs.google.com/document/d/1hJMTPqdLL7JFJeYoI6QWyE4nzeittxtJiQYCVpla-24/edit?usp=sharing 2. Reference to original BME 198A proposal: https://docs.google.com/document/d/1Q3rbB4pIYkrD6uTe9HP5Z2ajvTPDtFzntUMB8UmYjPU/edit?tab=t.0 3. Reference to original Final Proposal Presentation: https://docs.google.com/presentation/d/1Zoui2FucgHbPv2qgjlV2VdTFZxGFMmJnee2LTVGcMZw/edit?slide=id.g31c45062dbc_2_23#slide=id.g31c45062dbc_2_23 25 References 1. Anatomy and physiology II. Cardiac Cycle | Anatomy and Physiology II. (n.d.). https://courses.lumenlearning.com/suny-ap2/chapter/cardiac-cycle/ 2. Bambu Lab X1C 3D printer. Bambu Lab US. (n.d.). https://us.store.bambulab.com/products/x1-carbon?variant=42698346037384&country=US&currency=USD&utm_medium=product_sync&utm_source=google&utm_content=sag_organic&utm_campaign=sag_organic&gad_source=1&gclid=Cj0KCQiAgdC6BhCgARIsAPWNWH0z1ciFw_cBx2GzujkuBmNxcD2JNWksZbD_3vi6j1ADo30puqBf04oaAvRnEALw_wcB 3. Brodin, L. A., Ekestr\u00f6m, S., Forssell, G., Lindblom, D., & Pehrsson, S. K. (1988). Strut-fracture of a Bj\u00f6rk-Shiley tilting disc valve", "Contributions Authorization, Figure Table, Appendix, References 24 Appendix 1. Compilation of the technical memorandums: https://docs.google.com/document/d/1hJMTPqdLL7JFJeYoI6QWyE4nzeittxtJiQYCVpla-24/edit?usp=sharing 2. Reference to original BME 198A proposal: https://docs.google.com/document/d/1Q3rbB4pIYkrD6uTe9HP5Z2ajvTPDtFzntUMB8UmYjPU/edit?tab=t.0 3. Reference to original Final Proposal Presentation: https://docs.google.com/presentation/d/1Zoui2FucgHbPv2qgjlV2VdTFZxGFMmJnee2LTVGcMZw/edit?slide=id.g31c45062dbc_2_23#slide=id.g31c45062dbc_2_23 25 References 1. Anatomy and physiology II. Cardiac Cycle | Anatomy and Physiology II. (n.d.). https://courses.lumenlearning.com/suny-ap2/chapter/cardiac-cycle/ 2. Bambu Lab X1C 3D printer. Bambu Lab US. (n.d.). https://us.store.bambulab.com/products/x1-carbon?variant=42698346037384&country=US&currency=USD&utm_medium=product_sync&utm_source=google&utm_content=sag_organic&utm_campaign=sag_organic&gad_source=1&gclid=Cj0KCQiAgdC6BhCgARIsAPWNWH0z1ciFw_cBx2GzujkuBmNxcD2JNWksZbD_3vi6j1ADo30puqBf04oaAvRnEALw_wcB 3. Brodin, L. A., Ekestr\u00f6m, S., Forssell, G., Lindblom, D., & Pehrsson, S. K. (1988). Strut-fracture of a Bj\u00f6rk-Shiley tilting disc valve diagnosed by echocardiography--a case report. European heart journal, 9(2), 191\u2013193. https://doi.org/10.1093/oxfordjournals.eurheartj.a062474 4. Centers for Disease Control and Prevention. (2023, February 2). About heart valve disease . U.S. Department of Health & Human Services. https://www.cdc.gov/heart-disease/about/heart-valve-disease.html 5. COMSOL 5+2 Server Educational Price List - COMSOL MULTIPHYSICS \u00ae Software Price List all licenses. Studocu. (n.d.). https://www.studocu.com/pt-br/document/universidade-de-sao-paulo/geofisica-aplicada/comsol-52-server-educational-price-list/71719882 6. Djadoudi, H. (2024, March 21). Comsol Multiphysics Expert Review, pricing and alternatives - 2024. WorQuick. https://www.worquick.com/post/comsol-multiphysics-review#:~:text=COMSOL%20uses%20a%20tiered%20pricing,test%20it%20out%20before%20purchasing. 26 7. Dasi, L. P., Simon, H. A., Sucosky, P., & Yoganathan, A. P. (2009, February). Fluid Mechanics of artificial heart valves . Clinical and experimental pharmacology & physiology. https://pmc.ncbi.nlm.nih.gov/articles/PMC2752693/ 8. Ellis, J. T., & Yoganathan, A. P. (2000). A comparison of the hinge and near-hinge flow fields of the St Jude Medical Hemodynamic Plus and regent bileaflet mechanical heart valves. The Journal of Thoracic and Cardiovascular Surgery, 119(1), 83\u201393. https://doi.org/10.1016/s0022-5223(00)70221-2 9. Lindblom, D., Bjork, V., & Semb, B. (n.d.). Mechanical failure of the bjork-shiley valve. https://www.jtcvs.org/article/S0022-5223(19)35850-7/pdf 10.Lumen Learning. (n.d.). Cardiac cycle . In Anatomy and Physiology II . https://courses.lumenlearning.com/suny-ap2/chapter/cardiac-cycle/ 11. MacIsaac, S., Jaffer, I. H., Belley-C\u00f4t\u00e9, E. P., McClure, G. R., Eikelboom, J. W., & Whitlock, R. P. (2019). How did we get here?: A historical review and critical analysis of anticoagulation therapy following mechanical valve replacement. Circulation, 140(23), 1933\u20131942. https://doi.org/10.1161/circulationaha.119.041105 12. Mayo Foundation for Medical Education and Research. (2023, September 27). Aortic Valve Disease . Mayo Clinic. https://www.mayoclinic.org/diseases-conditions/aortic-valve-disease/symptoms-causes/syc-20355117#dialogId1043509 13. MediLexicon International. (n.d.). Diastole vs. Systole: What is the difference?. Medical News Today. https://www.medicalnewstoday.com/articles/321447#:~:text=Systole%20is%20defined%20by%20the,and%20tissues%20of%20the%20body. 27 14. Kumar, T., Singh, A., Thakre, S., Acharya, S., Shukla, S., & Kumar, S. (2023, July 19). Scientific Evolution of Artificial Heart Valves: A narrative review. Cureus. https://pmc.ncbi.nlm.nih.gov/articles/PMC10438674/ 15. PETG HF. Bambu Lab US. (n.d.-b). https://us.store.bambulab.com/products/petg-hf?variant=42881154973832&country=US&currency=USD&utm_medium=product_sync&utm_source=google&utm_content=sag_organic&utm_campaign=sag_organic&srsltid=AfmBOopNxOMC3ICcnsa_9m6-SJ5-B8vHeFbcS8Q7RC1C39fqAXaf84UsQyc&gQT=1 16. SOLIDWORKS, (2024, November 5). SOLIDWORKS Plans and Pricing . www.solidworks.com/how-to-buy/solidworks-plans-pricing?utm_source=google&utm_medium=cpc&utm_campaign=202411_nam_sw_googleStoreCost_en_lab_brand_us&utm_adgroup=License&utm_term=solidwork+license&gad_source=1&gclid=Cj0KCQiAgdC6BhCgARIsAPWNWH0lIqz8838TtRk_H1uL7Pkz1UHSATSMCbp3qtwmU1XbeOcIkTXvtSUaAjSlEALw_wcB. 17. Steyerberg, E. W., Kallewaard, M., Van Der Graaf, Y., Van Herwerden, L. A., & Habbema, J. D. F. (2000). Decision analyses for prophylactic replacement of the", "difference?. Medical News Today. https://www.medicalnewstoday.com/articles/321447#:~:text=Systole%20is%20defined%20by%20the,and%20tissues%20of%20the%20body. 27 14. Kumar, T., Singh, A., Thakre, S., Acharya, S., Shukla, S., & Kumar, S. (2023, July 19). Scientific Evolution of Artificial Heart Valves: A narrative review. Cureus. https://pmc.ncbi.nlm.nih.gov/articles/PMC10438674/ 15. PETG HF. Bambu Lab US. (n.d.-b). https://us.store.bambulab.com/products/petg-hf?variant=42881154973832&country=US&currency=USD&utm_medium=product_sync&utm_source=google&utm_content=sag_organic&utm_campaign=sag_organic&srsltid=AfmBOopNxOMC3ICcnsa_9m6-SJ5-B8vHeFbcS8Q7RC1C39fqAXaf84UsQyc&gQT=1 16. SOLIDWORKS, (2024, November 5). SOLIDWORKS Plans and Pricing . www.solidworks.com/how-to-buy/solidworks-plans-pricing?utm_source=google&utm_medium=cpc&utm_campaign=202411_nam_sw_googleStoreCost_en_lab_brand_us&utm_adgroup=License&utm_term=solidwork+license&gad_source=1&gclid=Cj0KCQiAgdC6BhCgARIsAPWNWH0lIqz8838TtRk_H1uL7Pkz1UHSATSMCbp3qtwmU1XbeOcIkTXvtSUaAjSlEALw_wcB. 17. Steyerberg, E. W., Kallewaard, M., Van Der Graaf, Y., Van Herwerden, L. A., & Habbema, J. D. F. (2000). Decision analyses for prophylactic replacement of the Bj\u00f6rk-Shiley Convexo-concave heart valve: Medical Decision Making, 20(1), 20\u201332. https://doi.org/10.1177/0272989x0002000103 18. Zabalgoitia, M. (2007). Echocardiographic recognition and quantitation of prosthetic valve dysfunction. In C. M. Otto (Ed.), The practice of clinical echocardiography (3rd ed., pp. 577\u2013604). W.B. Saunders. https://doi.org/10.1016/B978-1-4160-3640-1.50029-0 28", "Streamlining Jaffe Assay Design and Preserving Sensitivity with Atomized Picric Acid San Jos\u00e9 State University, Charles W. Davidson College of Engineering In fulfillment of course requirements for BME 298, Spring 2025. Reading Committee: Dr. Abdulmelik Mohammed and Dr. Alessandro Bellofiore Course Instructor: Ashkan Aryaei Adwait Pathak Introduction Prevalence of Chronic Kidney Disease According to the CDC, nephritis, nephrotic syndrome, and nephrosis accounted for 57,937 deaths in 2022, making it the 9th leading cause of death. Kidney disease is a major health issue in the United States, with approximately 37 million people, about 1 in 7 Americans, suffering from chronic kidney disease (CKD) [1] . The growing global burden of chronic kidney disease (CKD) underscores the need for accessible and affordable diagnostic tools, especially in regions with limited healthcare infrastructure. Chronic kidney disease affects all demographics worldwide, with certain populations, such as older adults and those with conditions like diabetes and hypertension, being at higher risk. The financial strain of CKD management on healthcare systems is substantial, as current diagnostics are costly and often inaccessible in low resource settings, limiting early detection and intervention opportunities. Traditional methods for monitoring kidney function, such as laboratory based blood tests, require specialized equipment and trained personnel, creating barriers to timely and widespread screening [1] . These methods are often unavailable or cost prohibitive in rural or low resource settings, and they rely heavily on healthcare infrastructure that may not be readily accessible to large portions of the affected population. This has led to a substantial unmet need for point of care diagnostics that can be deployed in remote or low resource settings. With an estimated 850 million people worldwide affected by kidney diseases, and CKD prevalence steadily rising due to increasing rates of diabetes, hypertension, and an aging population, there is significant demand for innovative solutions to lower costs and improve accessibility of testing. 1 Link to Creatinine Creatinine is a stable byproduct of muscle metabolism, produced at a consistent rate and minimally reabsorbed by the kidneys, making it a reliable indicator of kidney function. Elevated blood creatinine levels reflect impaired kidney filtration and are used to estimate the glomerular filtration rate (GFR), a key marker of kidney health [2] . According to NIH, a GFR of 60 or higher is considered normal, while a GFR below 60 may indicate kidney disease, and a GFR of 15 or lower may suggest kidney failure [2]", "muscle metabolism, produced at a consistent rate and minimally reabsorbed by the kidneys, making it a reliable indicator of kidney function. Elevated blood creatinine levels reflect impaired kidney filtration and are used to estimate the glomerular filtration rate (GFR), a key marker of kidney health [2] . According to NIH, a GFR of 60 or higher is considered normal, while a GFR below 60 may indicate kidney disease, and a GFR of 15 or lower may suggest kidney failure [2] . The relationship between creatinine concentration and GFR is determined through blood tests, with increased creatinine levels indicating a reduced GFR. Reduced GFR levels are more probable in patients with higher age, smoking habits, diabetes, and hypertension [3] . Measuring creatinine levels is a practical and effective method for assessing kidney function, and the development of paper based analytical devices (PADs) offers a promising approach to simplify and enhance this process, potentially reducing healthcare costs by enabling more individuals to monitor their kidney function in an affordable way. Paper Based Analytical Devices and Low Cost Disease Detection PADs are simple, cost effective tools for detecting biomolecules, including creatinine, using localized, specific reactions. PADs are typically made by patterning hydrophobic barriers (e.g., wax, ink, or photolithography) on cellulose paper to define fluidic pathways. The paper channels guide the flow of samples and reagents through capillary action without the need for external pumps [4] . Detection zones on the PAD are pre loaded with specific reagents, such as enzymes or chromogens, that react with the analyte (e.g., creatinine) to produce a visible color change. The resulting color intensity can be quantified using a smartphone camera or other simple optical 2 readers, making PADs an accessible and portable option for diagnostic testing in low resource settings. Low cost Diagnostic Tests Paper based analytical devices offer a versatile and scalable approach to meeting this demand. Their low production costs, ease of use, and ability to deliver rapid results make them particularly valuable in rural or underfunded healthcare systems, where early detection of kidney dysfunction can prevent disease progression and improve outcomes. The economic benefits of developing such low cost, portable diagnostics are substantial, as PADs have the potential to significantly reduce the financial burden of CKD monitoring on both patients and healthcare systems, while also lowering overall healthcare costs through earlier detection and more accessible screening options. Furthermore, the development of portable and user", "to deliver rapid results make them particularly valuable in rural or underfunded healthcare systems, where early detection of kidney dysfunction can prevent disease progression and improve outcomes. The economic benefits of developing such low cost, portable diagnostics are substantial, as PADs have the potential to significantly reduce the financial burden of CKD monitoring on both patients and healthcare systems, while also lowering overall healthcare costs through earlier detection and more accessible screening options. Furthermore, the development of portable and user friendly PADs aligns with the broader trend toward decentralized healthcare, empowering patients to monitor their health in real time and reducing the burden on overextended clinical laboratories. The global market for diagnostic devices is projected to reach $99.8 billion by 2026, driven by a combination of technological advances and the rising need for personalized, home based health management. The introduction of PADs for creatinine detection could revolutionize kidney health monitoring by offering a practical, scalable, and life saving tool to millions of patients at risk for CKD. 3 Literature Review Basis of Detection: Why Picric Acid? Creatinine is a cyclic compound with the formula C \u2084 H \u2087 N \u2083 O, produced from the breakdown of creatine and commonly measured in biological assays to evaluate kidney function. Picric acid, or 2,4,6 trinitrophenol (C \u2086 H \u2082 (NO \u2082 ) \u2083 OH), is an aromatic compound with acidic properties, widely used in biochemical reactions for its ability to form colored complexes with creatinine [4,5] . Figure 1: The Jaffe Reaction. Picric acid (red) combines with Creatinine (blue). The Janovsky complex (right) is an orange colored substance. Typically, endogenous creatinine levels are measured using the Jaffe reaction (Figure 1) or Jaffe kinetic assays, both of which rely on a colorimetric change to quantify creatinine. In the Jaffe reaction, creatinine combines with picric acid (originally yellow) to form an orange colored compound. In the past, researchers have relied on spectrometers or spectrophotometers to quantify color change and wavelength absorbance. However, today it is feasible to employ a compact smartphone or tablet\u2019s camera in order to quantify color change. Advancements in camera quality offers new opportunities for accurate quantification of analytes like creatinine. 4 Smartphone cameras are designed for general imaging, so they capture a limited portion of the visible spectrum (approximately 400 to 700 nm) with relatively low spectral resolution due to their use of RGB filters. In contrast, spectrophotometers are engineered for", "quantify color change and wavelength absorbance. However, today it is feasible to employ a compact smartphone or tablet\u2019s camera in order to quantify color change. Advancements in camera quality offers new opportunities for accurate quantification of analytes like creatinine. 4 Smartphone cameras are designed for general imaging, so they capture a limited portion of the visible spectrum (approximately 400 to 700 nm) with relatively low spectral resolution due to their use of RGB filters. In contrast, spectrophotometers are engineered for precise spectral analysis and can measure a broader range, including ultraviolet (UV) to near infrared (NIR) wavelengths, with high spectral resolution. While spectrophotometers can prove to be a useful validation tool, the feature extraction aspect of image analysis using ML/AI suggests that such a method could yield highly accurate results without the use of spectrophotometry. Reaction Conditions and Expected Result One study developed a device that enhances the detection of creatinine in blood by employing a colorimetric approach, demonstrating promising results for both range and accuracy. This device achieved a stable chromogenic signal across the temperature range 25- 37\u00baC and a reaction time of 5 minutes. Specifically, it was found that the signal output remains consistent at varying concentrations of creatinine when the temperature and reaction time are optimally adjusted, implying that deviations from these parameters can destabilize the results. Elevated temperatures, for example, were shown to disrupt the colorimetric response, while extending the reaction time did not markedly enhance the detection accuracy, but rather introduced potential for variability [5] . Therefore, maintaining precise control over these variables is crucial for reliable creatinine measurement, reflecting the device's capability to deliver accurate results under carefully regulated conditions [5] . Factors Affecting Device Perfomance The geometry of channels in paper based analytical devices also significantly impacts both the sensitivity and detection range of assays. Narrower, shorter channels often amplify sensitivity by concentrating the analyte, whereas broader, elongated channels accommodate a 5 greater range, but may dilute the sample, thus attenuating sensitivity [6,7] . Channel depth further modulates these effects; shallow channels enhance interactions at the cost of volume, while deeper channels support larger samples but can diminish sensitivity [6,7] . Tapered channels, which converge towards detection zones, intensify the analyte concentration but might restrict the range, whereas uniform channels provide consistent flow at the expense of sensitivity. Complex designs involving branching or merging channels can enhance both sensitivity and range but may", "range, but may dilute the sample, thus attenuating sensitivity [6,7] . Channel depth further modulates these effects; shallow channels enhance interactions at the cost of volume, while deeper channels support larger samples but can diminish sensitivity [6,7] . Tapered channels, which converge towards detection zones, intensify the analyte concentration but might restrict the range, whereas uniform channels provide consistent flow at the expense of sensitivity. Complex designs involving branching or merging channels can enhance both sensitivity and range but may introduce variability. Thus, optimizing channel geometry requires a delicate balance, achieved through iterative refinement, to meet the specific demands of the assay. Sensitivity is greatly affected by the type of paper used in the device. Different papers have varying capillary flow rates, porosities, and thicknesses, which are crucial factors in the operation of paper based microfluidic devices. The capillary flow rate, which is the speed of sample movement through the membrane, decreases exponentially over the membrane\u2019s length [4] . For this reason, capillary flow time provides an accurate representation of assay effectiveness. Porosity, or the volume of air within the paper, significantly affects how fluids move through the device, where higher porosity generally increases the speed of capillary action [4] . The thickness of the membrane also plays a key role, influencing not only the amount of sample absorbed but also the mechanical durability and the visibility of signals. Thicker membranes can absorb more sample but may change the distribution of reagents and reduce signal clarity, while thinner membranes may not hold up well in the manufacturing process [4,10] . These factors must be precisely controlled to ensure the accuracy and consistency of paper based microfluidic assays. Established Device Assembly Methods This device, created and improved upon by CardioLab students from previous years, already has established some functionality relying on existing fabrication methods. The 6 microfluidic device has a three layer structure, featuring a top filter paper layer, a middle plasma separation membrane, and a bottom backing layer. Thermal Transfer Printing In a three layer design, thermal printing offers a precise approach to patterning the detection zone on the filter paper layer. Figure 2: Device Design: Top Layer: Whatman Grade 4 filter paper with printed ink barriers for guiding fluid flow. Middle Layer: Plasma separation membrane designed to isolate plasma from whole blood. Bottom Layer: Plastic card backing to provide structural support. The top filter paper layer (see Figure 2)", "plasma separation membrane, and a bottom backing layer. Thermal Transfer Printing In a three layer design, thermal printing offers a precise approach to patterning the detection zone on the filter paper layer. Figure 2: Device Design: Top Layer: Whatman Grade 4 filter paper with printed ink barriers for guiding fluid flow. Middle Layer: Plasma separation membrane designed to isolate plasma from whole blood. Bottom Layer: Plastic card backing to provide structural support. The top filter paper layer (see Figure 2) receives the sample and contains the reaction site. By creating heat induced patterns on the filter paper, a precise zone of picric acid can be sequestered into specific regions, forming well defined detection zones for creatinine and sample loading zones [17] . Thermal ink printing also enhances consistency, sensitivity, and aids massively in large scale production. Once the detection zones are created, the filter paper layer is aligned with the plasma separation membrane. This middle layer selectively isolates plasma, allowing only plasma and target analytes, like creatinine, to reach the detection area while filtering out red blood cells and other components that could interfere with the reaction. The bottom plastic card 7 backing stabilizes the entire device, providing structural support that keeps all layers aligned [17] . This backing layer is rigid and non porous, ensuring that fluids remain within the active regions and flow consistently through the detection zone. Together, these three layers, in conjunction with thermal printing a hydrophobic ink on the filter paper, have the potential to create a stable, sensitive, and effective platform for creatinine detection. Existing literature [14] supports the feasibility of using thermal printing in fabricating paper based analytical devices (PADs), showing it to be a reliable, scalable, and precise method for reagent deposition. Studies have demonstrated that thermal printing can contain reagents with high spatial accuracy and consistency [15] , creating reaction zones that enhance sensitivity and reproducibility in colorimetric assays. For instance, research on PADs has shown that thermal printing effectively transfers reagents to paper substrates without the need for liquid dispensing, reducing both the risk of uneven distribution and reagent waste [15,16] . Additionally, thermal printing enables rapid fabrication of PADs at low cost, making it an ideal technique for developing disposable, single use diagnostic devices. This setup can enable reliable colorimetric analysis, but interference of extraneous substances still needs to be managed. Caveats and Potential for Improvement: Reaction Site and", "research on PADs has shown that thermal printing effectively transfers reagents to paper substrates without the need for liquid dispensing, reducing both the risk of uneven distribution and reagent waste [15,16] . Additionally, thermal printing enables rapid fabrication of PADs at low cost, making it an ideal technique for developing disposable, single use diagnostic devices. This setup can enable reliable colorimetric analysis, but interference of extraneous substances still needs to be managed. Caveats and Potential for Improvement: Reaction Site and Kinetics To maximize sensitivity in a paper based analytical device for creatinine detection, it is essential to carefully control the concentration and application form of picric acid in the reaction center [8] . For an effective reaction, the picric acid must be applied in a concentration high enough to ensure a visible color change with minimal creatinine present, yet not so concentrated that it overwhelms the reaction zone or leads to inconsistencies in signal intensity. When creating a stable reaction center, it is crucial to apply picric acid to the paper substrate in a way that ensures even distribution and complete adherence [8,18] . Using picric acid and applying a controlled, thin 8 film across the paper\u2019s detection zone, the reaction zone can be primed for reaction with aqueous creatinine. The reaction between picric acid and creatinine is not highly specific, making it less reliable for precise measurements. Picric acid reacts with other organic compounds, so contaminants can skew results. Colorimetric changes, especially at low creatinine concentrations, can be inconsistent. Early studies [22] demonstrate that the effect of picric acid concentration on colorimetric change plateaus around 3\u20134 \u00d7 10 \u207b \u00b3 M. In contrast, creatinine concentration significantly influences colorimetric change. Consequently, further increasing picric acid concentration beyond this range is not effective for accurately detecting low creatinine levels, highlighting the need for alternative approaches or adjustments in methodology. Techniques such as spray coating or dip coating, followed by a drying step under controlled humidity and temperature conditions, can be utilized to achieve a uniform layer of picric acid across the detection surface [9] . This method ensures the picric acid is readily accessible to incoming creatinine solutions, which is especially important when dealing with minute concentrations. The concentration of picric acid must also be optimized to balance sensitivity and signal clarity. Low concentrations may fail to generate a sufficiently strong chromogenic response, while excessively high concentrations can increase background noise", "step under controlled humidity and temperature conditions, can be utilized to achieve a uniform layer of picric acid across the detection surface [9] . This method ensures the picric acid is readily accessible to incoming creatinine solutions, which is especially important when dealing with minute concentrations. The concentration of picric acid must also be optimized to balance sensitivity and signal clarity. Low concentrations may fail to generate a sufficiently strong chromogenic response, while excessively high concentrations can increase background noise and lead to saturation effects, obscuring accurate readings. Typically, a concentration range of 0.01\u20130.05 M is optimal for detecting low levels of creatinine [18] . This range allows the reaction center to exhibit a clear and sharp color change, even in response to minimal amounts of creatinine, which is crucial for detecting early markers of kidney dysfunction or for monitoring creatinine levels in clinical settings. 9 Additional studies provide further insight into reaction conditions that influence the sensitivity and precision of the Jaff\u00e9 reaction. Experiments investigating the effects of temperature and reaction time on the G (green) + B (blue) intensity of the yellow orange complex reveal important considerations. As shown in the figure below, the reaction temperature significantly affects signal intensity. Creatinine samples with concentrations of 0.19 mg/dL, 3.82 mg/dL, and 7.64 mg/dL were analyzed across a temperature range of 25\u201350 \u00b0C, with a fixed reaction time of 8 minutes. The results indicate that the G (green) + B (blue) intensity increases as the reaction temperature rises to 37 \u00b0C but slightly decreases as the temperature is further increased to 50 \u00b0C. This suggests that the optimal reaction temperature for maximum signal intensity is 37 \u00b0C. Maintaining this temperature is crucial for ensuring a consistent and strong chromogenic response [5] . Figure 3: Optimal reaction temperature (37\u00baC) and time (5 minutes) for colorimetric change. Microfluidic paper based platform for whole blood creatinine detection, Tseng et al. (2018) [5] Reaction time is another critical parameter for optimizing creatinine detection. As illustrated by Tseng et al., intensity of creatinine control samples increases with reaction time for durations between 1\u20134 minutes, indicating that the reaction is incomplete within this interval 10 (Figure 3) [5] . For reaction times longer than 5 minutes, the intensity stabilizes, confirming that the reaction has reached completion. Overall, optimizing the conditions for picric acid application and the reaction parameters for the Jaff\u00e9 reaction is essential for maximizing", "time is another critical parameter for optimizing creatinine detection. As illustrated by Tseng et al., intensity of creatinine control samples increases with reaction time for durations between 1\u20134 minutes, indicating that the reaction is incomplete within this interval 10 (Figure 3) [5] . For reaction times longer than 5 minutes, the intensity stabilizes, confirming that the reaction has reached completion. Overall, optimizing the conditions for picric acid application and the reaction parameters for the Jaff\u00e9 reaction is essential for maximizing the sensitivity and reliability of paper based analytical devices for creatinine detection. Careful consideration of picric acid concentration, application methods, reaction temperature, and time ensures consistent and accurate readings. Additionally, understanding potential interferences in the assay is critical for maintaining the clinical relevance of creatinine measurements, especially in complex biochemical environments. These optimizations collectively enhance the capability of paper based devices in both diagnostic and monitoring applications. Detection Methods Paper based analytical devices (PADs) offer a promising and cost effective method for detecting creatinine. Specifically, integrating smartphone based detection methods with PADs has the potential to revolutionize point of care diagnostics. By leveraging advanced image analysis and machine learning algorithms, these devices can not only increase accuracy but also provide real time results accessible to both patients and healthcare providers [11] . To accurately measure creatinine levels, a sensor employing colorimetric detection displays color changes e.g., shift from yellow to orange) corresponding to varying creatinine concentrations. Such a process involves capturing images or optical data of the reaction, which undergoes feature extraction to assess color attributes like hue and intensity [11,12] . A machine learning model, typically a convolutional neural network (CNN), is trained on a dataset with known creatinine levels to recognize patterns in these color changes. This model, once validated and tested for accuracy, is deployed for real time analysis, predicting creatinine concentrations from new samples. 11 Calibration processes and ongoing model refinement ensure that the system adapts to environmental variations and maintains precision, leveraging AI to interpret subtle color shifts effectively and deliver reliable quantifications [12] . The shift towards mobile health technologies democratizes medical testing, making it feasible to monitor conditions like chronic kidney disease in resource limited settings. Additionally, the portability and ease of use of PADs, combined with the widespread availability of smartphones, position this approach as a transformative tool in the broader field of personalized medicine. As PAD technology continues to evolve, its", "to environmental variations and maintains precision, leveraging AI to interpret subtle color shifts effectively and deliver reliable quantifications [12] . The shift towards mobile health technologies democratizes medical testing, making it feasible to monitor conditions like chronic kidney disease in resource limited settings. Additionally, the portability and ease of use of PADs, combined with the widespread availability of smartphones, position this approach as a transformative tool in the broader field of personalized medicine. As PAD technology continues to evolve, its application in monitoring biomarkers such as creatinine represents a significant step forward in delivering cost effective and scalable healthcare solutions globally. Advances in quantification methods, particularly the use of smartphone cameras coupled with machine learning, present an innovative approach to enhance detection accuracy and accessibility [13] . Overall, PADs can achieve reliable and precise creatinine measurement, contributing to improved diagnostic capabilities in various settings. Proposed Improvements and Justifications Figure 4: Deng et al. (2018) demonstrate a predicate study using spray based application of reagents to a uPAD. In a predicate study, Deng et al. (2018) [24] demonstrated that aerosol spray application could be used to define hydrophobic barriers and apply reagents uniformly for iron detection, achieving consistent colorimetric responses. My approach adapts this principle for creatinine 12 detection using the Jaffe reaction. By atomizing the picric acid/NaOH reagent onto the test site, I achieved consistent color development at clinically relevant low concentrations. Unlike Deng\u2019s fabrication-focused study, my work focuses on functional equivalency with predicate creatinine assays and optimizing the colorimetric response for diagnostic purposes, particularly in a resource-limited, point-of-care context. An atomized form of picric acid offers increased surface area, which improves the rate of reaction between creatinine and the picric acid, facilitating faster and more reliable detection. This could lead to a reduction in the time required for colorimetric analysis, a significant advantage for point of care diagnostics where rapid results are critical. The enhanced reactivity of atomized picric acid would also improve the sensitivity of the uPAD, allowing it to detect lower creatinine levels with greater precision. This is particularly important for early stage detection of chronic kidney disease (CKD), where small changes in creatinine levels can be indicative of kidney dysfunction and require timely intervention. The atomization process may also help achieve more uniform sample spreading across the uPAD. This uniformity can reduce variability in test results and ensure greater reproducibility. Atomized picric acid also has the", "acid would also improve the sensitivity of the uPAD, allowing it to detect lower creatinine levels with greater precision. This is particularly important for early stage detection of chronic kidney disease (CKD), where small changes in creatinine levels can be indicative of kidney dysfunction and require timely intervention. The atomization process may also help achieve more uniform sample spreading across the uPAD. This uniformity can reduce variability in test results and ensure greater reproducibility. Atomized picric acid also has the potential to enhance the stability of the colorimetric signal, reducing the likelihood of signal fading over time, which can be a challenge with traditional detection methods. The stability of the reaction over time ensures that the uPAD provides reliable results even in fluctuating environmental conditions, such as temperature or humidity variations, which may be encountered in field settings. Incorporating atomized picric acid into the uPAD system is expected to improve the overall efficiency of creatinine detection by enhancing the sensitivity, speed, and reproducibility of the colorimetric reaction. These improvements will increase the accuracy of the uPAD as a diagnostic tool for chronic kidney 13 disease, making it a more effective, rapid, and reliable method for monitoring kidney function in clinical and field settings. The following reagents and materials were used to carry out the experiments. The budget for these materials was under $500. Picric acid is applied to the reaction site, and creatinine is applied to the other square of the uPAD. Sodium hydroxide is combined with the picric acid for stability and optimal reaction conditions. The uPAD is constructed with the filter paper (printed with hydrophobic thermal ink) as the top layer, a membrane paper to allow capillary action of fluid from the sample loading square to the reaction zone square, and a bottom laminated cardstock layer for support. Analytical balances were used to measure reagents prior to making solutions of creatinine. Reagents \u25cf 1.3% stock picric acid solution (for dilution) \u25cf Sodium hydroxide (NaOH) pellets (\u226598% purity) \u25cf Distilled water \u25cf Creatinine powder (\u226598% purity) \u25cf 50:50 reagent mix of 2 M NaOH and 0.04 M picric acid Materials and Consumables \u25cf Whatman Grade 4 filter paper \u25cf Plasma separation membrane \u25cf Laminated cardstock sheets (for backing layer) \u25cf Itari P831 Thermal Printer \u25cf Double sided adhesive tape \u25cf 100 mL Erlenmeyer flasks (for reagent preparation) \u25cf Volumetric flask (1 L, for creatinine solution preparation) \u25cf Analytical balance", "(for dilution) \u25cf Sodium hydroxide (NaOH) pellets (\u226598% purity) \u25cf Distilled water \u25cf Creatinine powder (\u226598% purity) \u25cf 50:50 reagent mix of 2 M NaOH and 0.04 M picric acid Materials and Consumables \u25cf Whatman Grade 4 filter paper \u25cf Plasma separation membrane \u25cf Laminated cardstock sheets (for backing layer) \u25cf Itari P831 Thermal Printer \u25cf Double sided adhesive tape \u25cf 100 mL Erlenmeyer flasks (for reagent preparation) \u25cf Volumetric flask (1 L, for creatinine solution preparation) \u25cf Analytical balance (for verifying reagent mass) \u25cf Glass weighing tray (for NaOH) \u25cf Fine mist spray bottle (for reagent application) \u25cf Micropipette and sterile pipette tips (for precise liquid handling) \u25cf Ice bath or beaker of cold water (for NaOH solution prep) \u25cf Plastic tweezers or forceps (for device assembly) 14 \u25cf Smartphone with fixed height imaging stand, ring light (for controlled illumination during image capture) \u25cf Computer with Python installed for RGB analysis, ImageJ software (for RGB validation and comparison) Protocols Followed: These protocols were used to construct the solutions for creatinine and picric acid, as well as deposition of the reagents to the uPAD. \u25cf Prepared 0.05 M picric acid by combining 70.5 mL of 1.3% stock solution with 29.5 mL of distilled water in a 100 mL Erlenmeyer flask, mixed thoroughly, and stored in the acid cabinet of the PIV lab. \u25cf Prepared 2 M NaOH by measuring 7.9 g of sodium hydroxide pellets onto a glass tray. Pellets were added slowly to ~50 mL of distilled water in a 100 mL Erlenmeyer flask placed in an ice or cold water bath to control the exothermic reaction. After complete dissolution, distilled water was added up to the 100 mL mark. \u25cf Created the working reagent by mixing 7.5 \u03bc L of 2 M NaOH with 7.5 \u03bc L of 0.05 M picric acid, forming a 50:50 mixture. \u25cf Deposited 15 \u03bc L of the 50:50 reagent solution directly onto the reaction zone of the filter paper using a pipette and allowed it to dry overnight at room temperature. For the spray method, used a spray bottle to apply reagent to separate devices, ensuring the mass of reagent deposited matched pipetted samples, verified using an analytical balance. \u25cf Constructed each device with three layers: a top layer of Whatman Grade 4 filter paper, a plasma separation membrane in the middle, and a laminated cardstock backing on the bottom for rigidity.", "the reaction zone of the filter paper using a pipette and allowed it to dry overnight at room temperature. For the spray method, used a spray bottle to apply reagent to separate devices, ensuring the mass of reagent deposited matched pipetted samples, verified using an analytical balance. \u25cf Constructed each device with three layers: a top layer of Whatman Grade 4 filter paper, a plasma separation membrane in the middle, and a laminated cardstock backing on the bottom for rigidity. Layers were attached using double sided tape, with channels aligned precisely. \u25cf Prepared a 1 mg/dL creatinine solution by dissolving 10 mg of creatinine in 1 L of distilled water using a volumetric flask. This was used as the test analyte. 15 Imaging and Analytical Method Validation All images were acquired using a fixed height phone stand with locked camera settings and controlled lighting via a ring light. Sample devices were consistently positioned within a marked frame to eliminate spatial variance. RGB values were extracted from a central 50\u00d750 pixel crop. Validation of automated Python based RGB extraction was conducted by comparing results against ImageJ for 30 randomized samples. Reagent Volume and Concentration Consistency To confirm that reagent volume differences did not confound the comparison between application methods, the mass of picric acid solution delivered by both pipetting and spraying was verified using an analytical balance. For each trial, the weight of the device was recorded before and after reagent application. The difference in mass confirmed that both delivery methods deposited equivalent volumes of picric acid solution across all tests. This validation ensures that observed differences in signal intensity are attributable to application method and not to differences in reagent quantity. Experimental Controls and Exclusions Only the 5 minute and 10 minute timepoints were analyzed, as the 1 minute point produced insufficient chromatic change. Equal volumes of picric acid solution were delivered by either pipette or spray, with application verified via analytical balance. The experiment focuses on \u0394 RGB (change in color intensity), not raw values. 16 Channel Response Patterns During a yellow to orange transition, green channel values are expected to drop, while red is expected to increase. The blue channel is not expected to show significant change and was excluded from further analysis. Figure 5a: ImageJ values closely matched the values given by the python module (bottom) 17 Figure 5b: Sprayed samples (left) and pipetted samples (right) were", "via analytical balance. The experiment focuses on \u0394 RGB (change in color intensity), not raw values. 16 Channel Response Patterns During a yellow to orange transition, green channel values are expected to drop, while red is expected to increase. The blue channel is not expected to show significant change and was excluded from further analysis. Figure 5a: ImageJ values closely matched the values given by the python module (bottom) 17 Figure 5b: Sprayed samples (left) and pipetted samples (right) were of comparable weights before and after picric acid was sprayed. (After pictures shown) Figure 5c: Top and bottom sides of the same sample were used for positive and negative controls. 25 uL of 1 mg/dL creatinine was administered, and photos were taken 10 minutes after and before. (Left, pipetted samples; right, sprayed samples) 18 Table 1: Control RGB measurements Pipetted Controls R G B (+) Before 206 175 76 (+) After 212 174 68 ( ) Before 218 219 215 ( ) After 222 220 217 Sprayed (+) Before 210 172 108 (+) After 214 169 64 ( ) Before 212 208 213 ( ) After 216 206 214 Data Validation To verify the accuracy of the automated RGB measurements obtained using the custom Python module, I conducted a direct comparison against manual values generated in ImageJ for a representative subset of both pipetted and sprayed samples. As shown in Figure 5, red, green, and blue intensity values were extracted from identical regions using both methods. The bottom axis in Figure 5 represents the samples tested, which were assigned a letter and number (e.g., A10, B7. A full sampling of the data was the goal). The resulting data showed strong agreement across all channels, with nearly perfect overlap between ImageJ and Python outputs, particularly for red and green intensities. These results confirm the reliability of the Python script and validate its use for consistent, high-throughput image quantification in place of manual 19 measurement. Figure 6 shows the expected R,G,B values for positive and negative controls. Negative controls used water as the analyte for the uPAD, while Positive controls used a known concentration of creatinine that could elicit a chromatic change (5mg/dL). 20 Results Analysis was performed to evaluate the effect of reagent delivery method (pipetting vs. spraying) and reaction development time (5 min vs. 10 min) on the performance of the paper-based creatinine biosensor. Performance was assessed using changes in", "manual 19 measurement. Figure 6 shows the expected R,G,B values for positive and negative controls. Negative controls used water as the analyte for the uPAD, while Positive controls used a known concentration of creatinine that could elicit a chromatic change (5mg/dL). 20 Results Analysis was performed to evaluate the effect of reagent delivery method (pipetting vs. spraying) and reaction development time (5 min vs. 10 min) on the performance of the paper-based creatinine biosensor. Performance was assessed using changes in green (\u0394 G) and red (\u0394 R) channel intensity following the Jaffe reaction. A decrease in green (negative \u0394 G) and an increase in red (positive \u0394 R) signal indicates expected progression toward an orange colorimetric complex. Figure 7: Levene\u2019s test for equal variance shows sprayed samples have lowest \u0394 G variance at 5 minutes and lower variance overall. Levene\u2019s test (Figure 7) revealed that sprayed samples exhibited significantly lower variance in green channel intensity (\u0394 G) than pipetted samples at 5 minutes (p < 0.001). This suggests that the spray method delivers a more mechanically consistent volume of reagent across devices, resulting in tighter clustering of green channel outcomes. However, while lower variance implies higher reproducibility, the accompanying lack of red channel development, essential for visible signal, indicates a possible limitation in spray-based delivery. 21 Figure 8: Box Plots of Green (left) and Red (right) Channels, at 5 and 10 minutes post creatinine administration. Figure 9: Mean plots of \u2206R and \u2206G. 22 Figures 8 and 9 provide a detailed breakdown of the red and green channel changes across both methods and time points. The box plots in Figure 8 show that the pipet method induced a substantial decrease in green intensity (\u0394 G) and a concurrent increase in red intensity (\u0394 R), especially at 10 minutes, which is consistent with a transition toward orange. In contrast, the spray method resulted in minimal red enhancement and either a smaller or even positive change in green, particularly at the 5-minute mark. These trends are reinforced in the interval plots of Figure 9, where mean \u0394 R for the pipet group is strongly negative while \u0394 G remains near zero or slightly negative, indicating a selective increase in red that is not accompanied by green elevation. Together, these results support the conclusion that pipetting is more effective in driving a red-dominant chromatic shift. Table 2: 2 Sample t-test results for the", "or even positive change in green, particularly at the 5-minute mark. These trends are reinforced in the interval plots of Figure 9, where mean \u0394 R for the pipet group is strongly negative while \u0394 G remains near zero or slightly negative, indicating a selective increase in red that is not accompanied by green elevation. Together, these results support the conclusion that pipetting is more effective in driving a red-dominant chromatic shift. Table 2: 2 Sample t-test results for the Green (above) and Red (below) channels Test Null hypothesis H \u2080 : \u03bc \u2081 - \u00b5 \u2082 = 0 Alternative H \u2081 : \u03bc \u2081 - \u00b5 \u2082 \u2260 0 T-Value DF P-Value 9.29 314 0 Test Null H \u2080 : \u03bc \u2081 - \u00b5 \u2082 = 0 Alternative H \u2081 : \u03bc \u2081 - \u00b5 \u2082 \u2260 0 T-Value DF P-Value 117.9 226 0 Two-sample t-tests were conducted independently for \u0394 G and \u0394 R, pooling 5- and 10-minute data within each channel to compare overall mean values between pipetted and sprayed groups. For the green channel, pipetted samples showed more positive mean \u0394 G values than sprayed samples (t = 9.29, df = 314, p < 0.001), indicating less green suppression. For the red channel, pipetted samples again showed significantly higher mean \u0394 R values (t = 117.9, df = 23 226, p < 0.001), confirming substantially greater red signal activation. However, due to skewed distributions and unequal variances, especially in \u0394 G, these results reflect differences in mean values, not necessarily consistent shifts across all samples. Figure 11: Scatterplot of \u0394 G versus \u0394 R. A scatterplot (Figure 11) of \u0394 R versus \u0394 G revealed distinct clustering patterns by reagent delivery method and timepoint. Pipetted samples, at both 5 and 10 minutes, primarily occupied the upper-right quadrant (positive \u0394 R and positive or near-zero \u0394 G), indicating strong red activation but limited or no green suppression. This implies that while the reaction proceeds toward a visibly orange endpoint, it does so without balanced chromogenic development. Sprayed samples at 5 minutes appeared in the lower-left quadrant (negative \u0394 R and negative \u0394 G), showing green suppression without corresponding red activation. At 10 minutes, sprayed samples shifted toward positive \u0394 G but retained negative \u0394 R values, suggesting diminishing green suppression and persistent failure to initiate red signal development. 24 Figure 12: \u0394 G- \u0394 R Composite Metric", "This implies that while the reaction proceeds toward a visibly orange endpoint, it does so without balanced chromogenic development. Sprayed samples at 5 minutes appeared in the lower-left quadrant (negative \u0394 R and negative \u0394 G), showing green suppression without corresponding red activation. At 10 minutes, sprayed samples shifted toward positive \u0394 G but retained negative \u0394 R values, suggesting diminishing green suppression and persistent failure to initiate red signal development. 24 Figure 12: \u0394 G- \u0394 R Composite Metric Analysis It is important to clarify that the original yellow regions in this study are not idealized or standardized as pure yellow in the RGB sense (i.e., R=255, G=255, B=0), but instead consist of pixels where the red and green values are both high and relatively close to each other. This distinction is critical when interpreting increases in red intensity following creatinine exposure. Since RGB values are capped at 255, a region that already contains a high red component has limited capacity for further increase unless the green component decreases proportionally. Therefore, understanding that the baseline \u201cyellow\u201d is not perfect yellow but rather have R and G values that are large and close to equal. The composite allows for a more realistic assessment of how an increase in red, particularly when accompanied by a decrease in green, produces an orange shift. Without this context, red channel changes near the upper limit could be misinterpreted. To interpret chromatic transitions more effectively, it is useful to consider the composite metric \u0394 G \u2013 \u0394 R, which captures the relative change in green versus red channel intensity. A negative value in this metric implies that either the green channel decreased or the red channel 25 increased, producing a net movement toward orange, a result that cannot be discerned from evaluating \u0394 G or \u0394 R in isolation. In this context (Figure 12), the pipetted groups at both 5 and 10 minutes demonstrate negative mean values of \u0394 G \u2013 \u0394 R, consistent with the fact that the red signal was elevated in relation to the green signal. This dual-channel effect, an increase in red and concurrent decrease in green, is what drives a true perceptual shift toward orange. Conversely, the spray groups show strongly positive values for this metric, indicating an decrease in red intensity that either outpaced or entirely overshadowed any decrease in green, resulting in a net yellow or greenish shift.", "mean values of \u0394 G \u2013 \u0394 R, consistent with the fact that the red signal was elevated in relation to the green signal. This dual-channel effect, an increase in red and concurrent decrease in green, is what drives a true perceptual shift toward orange. Conversely, the spray groups show strongly positive values for this metric, indicating an decrease in red intensity that either outpaced or entirely overshadowed any decrease in green, resulting in a net yellow or greenish shift. While neither method achieved a textbook orange endpoint, the pipet application produced the only net-negative composite values at both time points, implying that the chromatic balance shifted more effectively in the desired direction. This suggests that pipetting may promote a more favorable redistribution of chromatic intensity by enhancing red signal while attenuating green, thereby enabling a relatively stronger transition toward orange. Figure 13: Kruskal-Wallis Test Results Kruskal Wallis Results for Green Channel Descriptive Statistics Group N Median Mean Rank Z-Value 1 162 12.0204 205.7 8.30 2 162 3.0384 119.3 -8.30 Overall 324 162.5 Test Null hypothesis H\u2080: All medians are equal Alternative hypothesis H\u2081: At least one median is different 26 Method DF H-Value P-Value Not adjusted for ties 1 68.81 0.000 Adjusted for ties 1 68.81 0.000 Kruskal Wallis Results for Red Channel Descriptive Statistics Group N Median Mean Rank Z-Value 1 162 13.250 243.5 15.56 2 162 -205.317 81.5 -15.56 Overall 324 162.5 Test Null hypothesis H\u2080: All medians are equal Alternative hypothesis H\u2081: At least one median is different Method DF H-Value P-Value Not adjusted for ties 1 242.25 0.000 Adjusted for ties 1 242.25 0.000 Kruskal-Wallis Test Results To corroborate the findings of the parametric analysis and account for non-normal distributions, Kruskal-Wallis tests were performed independently for the green (\u0394 G) and red (\u0394 R) channels. For the green channel, a significant difference in medians was observed between pipetted and sprayed samples (H = 68.81, p < 0.001), with pipetted samples exhibiting a higher median \u0394 G (12.02) than sprayed samples (3.04). This confirms that pipetting resulted in less effective green suppression, while spraying achieved lower and more directionally appropriate \u0394 G values. 27 For the red channel, the Kruskal-Wallis test revealed an even more pronounced difference (H = 242.25, p < 0.001). Pipetted samples had a median \u0394 R of 13.25, indicating strong red channel activation. In contrast, sprayed samples exhibited a sharply negative median", "< 0.001), with pipetted samples exhibiting a higher median \u0394 G (12.02) than sprayed samples (3.04). This confirms that pipetting resulted in less effective green suppression, while spraying achieved lower and more directionally appropriate \u0394 G values. 27 For the red channel, the Kruskal-Wallis test revealed an even more pronounced difference (H = 242.25, p < 0.001). Pipetted samples had a median \u0394 R of 13.25, indicating strong red channel activation. In contrast, sprayed samples exhibited a sharply negative median \u0394 R of \u2212205.32, indicating complete failure to initiate the red signal development essential for a visible endpoint. It remains possible that certain outlier values reflect measurement artifacts or edge-case behaviors rather than generalizable trends. This strongly negative value points to inconsistencies during image capture, even though massive efforts were taken to standardize measurements. Another reason for these inconsistencies is likely reagent concentration, namely picric acid. The concentration used, based on previous protocols, was likely too high in relation to the creatinine concentration used. These nonparametric results reinforce the conclusion that pipetting produces a more complete and perceptible chromogenic shift, while spray-based delivery, though mechanically consistent, does not achieve the necessary red channel activation under current conditions. This study was conducted under controlled lab conditions using a limited number of replicates and simplified reagent formulations. As such, findings should not be generalized beyond the current experimental context without further validation. These statistical comparisons are intended to illustrate trends within the dataset, not to establish definitive performance benchmarks for diagnostic deployment. Discussion This study evaluated the performance of a paper-based analytical device for creatinine detection by examining how reagent delivery method (pipetting versus spraying) and reaction time (5 versus 10 minutes) influenced colorimetric shifts in green (\u0394 G) and red (\u0394 R) channels 28 during the Jaffe reaction. The expected endpoint for a successful assay is a decrease in green signal and an increase in red, resulting in an orange chromogenic complex. Statistical analysis revealed a trade-off between mechanical consistency and chemical completeness. Levene\u2019s test indicated that sprayed samples exhibited significantly lower variance in green channel intensity at both timepoints, with the lowest variance observed at 5 minutes (p < 0.001). This suggests that spraying provides a more uniform and controlled reagent distribution, likely due to the fine and even application of the atomized solution. In contrast, pipetting may introduce variability through droplet pooling or inconsistent coverage, leading to greater heterogeneity", "complex. Statistical analysis revealed a trade-off between mechanical consistency and chemical completeness. Levene\u2019s test indicated that sprayed samples exhibited significantly lower variance in green channel intensity at both timepoints, with the lowest variance observed at 5 minutes (p < 0.001). This suggests that spraying provides a more uniform and controlled reagent distribution, likely due to the fine and even application of the atomized solution. In contrast, pipetting may introduce variability through droplet pooling or inconsistent coverage, leading to greater heterogeneity in green channel outcomes. Two-sample t-tests comparing pooled timepoints showed that pipetted samples had significantly higher mean \u0394 G and \u0394 R values than sprayed samples (p < 0.001 for both). The higher \u0394 R indicates stronger red channel activation in pipetted samples, which is favorable for producing the visible endpoint. However, the elevated \u0394 G suggests an increase in green intensity, which is opposite of the expected direction for the yellow to orange transition. This implies that pipetting initiates red signal development, but may interfere with optimal green suppression, possibly due to oversaturation or uneven distribution of reagent. It is important to note that the t-tests rely on assumptions of normal distribution and equal variance, which were not met here, especially for the green channel. Because of this, these results must be interpreted with caution. To address this, Kruskal-Wallis tests were performed as a nonparametric alternative. These results supported the same directional conclusions. For the green channel, pipetted samples had a higher median \u0394 G (12.02) than sprayed samples (3.04), suggesting that spraying more effectively reduces green intensity, which aligns with the expected chromogenic progression. For the red channel, pipetted samples exhibited a median \u0394 R of 13.25 compared to 29 \u2212205.32 for sprayed samples. This confirmed that red signal development occurred only with pipetting. The extreme negative \u0394 R value in sprayed samples was verified in the raw data and likely reflects a sharp loss in red channel intensity. It may result from poor reaction progression or measurement artifact. This outcome indicates that sprayed application may not only fail to initiate red development, but may also disrupt baseline red values through reagent dilution or signal instability. These findings were further illustrated by the scatterplot of \u0394 R versus \u0394 G. Pipetted samples clustered in the upper right quadrant with positive \u0394 R and near-zero or positive \u0394 G, showing strong red activation but limited green suppression. This pattern", "intensity. It may result from poor reaction progression or measurement artifact. This outcome indicates that sprayed application may not only fail to initiate red development, but may also disrupt baseline red values through reagent dilution or signal instability. These findings were further illustrated by the scatterplot of \u0394 R versus \u0394 G. Pipetted samples clustered in the upper right quadrant with positive \u0394 R and near-zero or positive \u0394 G, showing strong red activation but limited green suppression. This pattern appeared consistent across both timepoints, suggesting that once red activation occurs, it is not heavily time-dependent. Sprayed samples at 5 minutes appeared in the lower left quadrant with negative \u0394 R and \u0394 G, showing green suppression without red activation. At 10 minutes, sprayed samples shifted upward in \u0394 G but remained negative in \u0394 R, suggesting that longer reaction time reduced green suppression and still failed to trigger red development. This disconnect implies that spray-based application initiates only the first step of the Jaffe reaction but does not support its completion. Green suppression may occur under lower levels of reagent contact, but red development may require more intense local reaction conditions. This suggests that spraying, while effective for uniform delivery, does not produce the necessary conditions for full chromogenic transition. Together, these results demonstrate a clear separation between mechanical precision and chemical efficacy. Spraying enhances reproducibility and reliably reduces green channel intensity, but it fails to trigger the red shift that marks assay completion. Pipetting delivers a more successful endpoint, but at the cost of variability in both color channels. These results 30 suggest that pipetting, although less precise, is currently the only method that can fully drive the Jaffe reaction under the tested formulation and device conditions . The above findings are exploratory in nature and do not claim generalizability beyond the conditions tested. Future Directions The findings of this study underscore both the strengths and limitations of current reagent delivery methods in paper-based creatinine sensing. While pipetted application outperformed spraying in generating a complete and perceptually valid orange color shift, it did so with increased variability and uneven green suppression. Spray, on the other hand, demonstrated reliable and directionally correct green channel suppression, but failed to activate the red channel a critical shortcoming that prevented full chromogenic development. These results suggest that signal quality is not defined by intensity alone, but by the coordinated behavior of both", "current reagent delivery methods in paper-based creatinine sensing. While pipetted application outperformed spraying in generating a complete and perceptually valid orange color shift, it did so with increased variability and uneven green suppression. Spray, on the other hand, demonstrated reliable and directionally correct green channel suppression, but failed to activate the red channel a critical shortcoming that prevented full chromogenic development. These results suggest that signal quality is not defined by intensity alone, but by the coordinated behavior of both channels and the reproducibility of their responses. The statistical stability of green suppression in sprayed samples supports the idea that reproducibility may be more valuable than raw signal magnitude, especially in low-resource or point-of-care settings. The challenge is therefore not to abandon spraying, but to modify its formulation and application parameters to support full red development while preserving its consistent green suppression. From a mechanistic perspective, the difference in performance between methods can be attributed to fluid dynamics and reagent distribution. Pipetted samples likely benefit from steep local concentration gradients and deeper reagent penetration, triggering rapid reaction kinetics and fast chromogenic transition. Spray application, by contrast, disperses reagent more thinly and uniformly, producing slower, more homogeneous interactions that initiate but do not 31 complete the reaction. This aligns with observed data: pipetted samples are intense but inconsistent; sprayed samples are consistent but incomplete. These results point toward multiple avenues for future optimization. Increasing spray volume, adjusting droplet size or velocity, or modifying solvent composition may enhance penetration depth and local reagent concentration. Likewise, reformulating the spray reagent for example by increasing picric acid concentration or tuning pH could promote red channel activation without disrupting the existing reproducibility of green suppression. The goal is not to make spray mimic pipet, but to engineer a new regime that combines the best aspects of both methods: consistent application, balanced chromogenic development, and robust perceptual output. Additionally, the delayed or incomplete red activation observed in sprayed samples suggests that spatial interaction between analyte and reagent including factors like substrate wetting, pore structure, and localized pH buffering may be more critical than previously understood. The convergence of \u0394 G values between methods at the 10-minute mark shows that given enough time, green suppression tapers off across conditions. Similarly, the \u0394 R values decrease in magnitude at 10 minutes as well, confirming that 5 minutes is the optimal reaction time. This study challenges the assumption that", "in sprayed samples suggests that spatial interaction between analyte and reagent including factors like substrate wetting, pore structure, and localized pH buffering may be more critical than previously understood. The convergence of \u0394 G values between methods at the 10-minute mark shows that given enough time, green suppression tapers off across conditions. Similarly, the \u0394 R values decrease in magnitude at 10 minutes as well, confirming that 5 minutes is the optimal reaction time. This study challenges the assumption that pipetting is the gold standard for paper-based diagnostics. Rather, pipetting is simply better matched to the current reagent and substrate conditions. With proper adjustments, spray-based application has the potential to outperform pipetting across all practical metrics including reproducibility, ease of use, cost, and automation potential making it the more scalable choice for real-world deployment. 32 In summary, while spray underperformed in its current state, its consistency and proper green suppression indicate that the underlying mechanism is sound. The solution is not to replace spray with pipet, but to engineer the chemistry around spray to resolve its red channel limitations. Doing so would unlock the full potential of automated, low-cost, and reliable paper-based diagnostics that do not depend on human precision and can function robustly in uncontrolled environments. 33 References: 1. Arnold C, Bissonnette L, Choi DH, Fang C, Fu E, Govorov AO, Hauck TS, He Y, Kit Anan W, Lee WG, et al. Advances in paper based point of care diagnostics. Biosensors and Bioelectronics. 2013; 54:585\u2013597. https://www.sciencedirect.com/science/article/pii/S095656631300777X 2. Beyond Wax Printing: Fabrication of Paper Based Microfluidic Devices Using a Thermal Transfer Printer. Ryan A. Ruiz, Jorge L. Gonzalez, Miguel Vazquez Alvarado, Nathaniel W. Martinez, Andres W. Martinez. Analytical Chemistry. 2022; 94(25):8833\u20138837. https://doi.org/10.1021/acs.analchem.2c01534 3. Chin Chung Tseng, Song Yu Lu, Szu Jui Chen, Ju Ming Wang, Lung Ming Fu, Yi Hong Wu. Microfluidic aptasensor POC device for determination of whole blood potassium. Analytica Chimica Acta. 2022; 1203:339722. https://doi.org/10.1016/j.aca.2022.339722 4. Dal Dosso F, Decrop D, P\u00e9rez Ruiz E, et al. Creasensor: SIMPLE technology for creatinine detection in plasma. Anal Chim Acta. 2018; 1000:191\u2013198. https://doi.org/10.1016/j.aca.2017.11.026 5. Deng M, Liao C, Wang X, Chen S, Qi F, Zhao X, Yu P. A paper-based colorimetric microfluidic sensor fabricated by a novel spray painting prototyping process for iron analysis. Microchemical Journal. 2018; 139:290\u2013296. https://doi.org/10.1016/j.microc.2018.03.006 6. FASTSTATS \u2013 Kidney Disease. Centers for Disease Control and Prevention. April 28, 2024. Accessed August 8, 2024. https://www.cdc.gov/nchs/fastats/kidney-disease.htm 7. Fu LM,", "F, Decrop D, P\u00e9rez Ruiz E, et al. Creasensor: SIMPLE technology for creatinine detection in plasma. Anal Chim Acta. 2018; 1000:191\u2013198. https://doi.org/10.1016/j.aca.2017.11.026 5. Deng M, Liao C, Wang X, Chen S, Qi F, Zhao X, Yu P. A paper-based colorimetric microfluidic sensor fabricated by a novel spray painting prototyping process for iron analysis. Microchemical Journal. 2018; 139:290\u2013296. https://doi.org/10.1016/j.microc.2018.03.006 6. FASTSTATS \u2013 Kidney Disease. Centers for Disease Control and Prevention. April 28, 2024. Accessed August 8, 2024. https://www.cdc.gov/nchs/fastats/kidney-disease.htm 7. Fu LM, Tseng CC, Ju WJ, Yang RJ. Rapid Paper Based System for Human Serum Creatinine Detection. Inventions. 2018; 3(2):34. https://doi.org/10.3390/inventions3020034 8. Habdias A. Silva Neto, Iana VS Arantes, Andr\u00e9 L. Ferreira, Guida HM do Nascimento, Gabriel N. Meloni, William R. de Araujo, Thiago RLC Paix\u00e3o, Wendell KT Coltro. Recent advances on paper based microfluidic devices for bioanalysis. TrAC Trends in Analytical Chemistry. 2023; 158:116893. https://doi.org/10.1016/j.trac.2022.116893 9. He Y, Zhang S, Zhang X, Baloda M, Gurung AS, Xu H, Zhang X, Liu G. Ultrasensitive nucleic acid biosensor based on enzyme gold nanoparticle dual label and lateral flow strip biosensor. Biosens Bioelectron. 2011; 26(5):2018\u20132024. 34 https://doi.org/10.1016/j.bios.2010.08.079 10. J V. Reaction of alkaline sodium picrate with creatinine: I. Kinetics and mechanism of formation of the mono creatinine picric acid complex. Clinical Chemistry. 1976. https://pubmed.ncbi.nlm.nih.gov/10095/ 11. Jafry AT, Lim H, Kang SI, Suk JW, Lee J. A comparative study of paper based microfluidic devices with respect to channel geometry. European Polymer Journal. 2015; https://www.sciencedirect.com/science/article/abs/pii/S0927775715304313 12. Jie Hu, ShuQi Wang, Lin Wang, Fei Li, Belinda Pingguan Murphy, Tian Jian Lu, Feng Xu. Advances in paper based point of care diagnostics. Biosensors and Bioelectronics. 2014; 54:585\u2013597. https://doi.org/10.1016/j.bios.2013.10.075 13. K\u0131l\u0131\u00e7 V, Horzum N, Ertugrul Solmaz M. From Sophisticated Analysis to Colorimetric Determination: Smartphone Spectrometers and Colorimetry. IntechOpen. 2020. https://doi.org/10.5772/intechopen.82227 14. Liu J, Kong X, Wang H, Zhang Y, Fan Y. Roll to roll wax transfer for rapid and batch fabrication of paper based microfluidics. Microfluidics and Nanofluidics. 2019; 24. https://doi.org/10.1007/s10404-019-2310-2 15. O\u2019Seaghdha CM, Lyass A, Massaro JM, et al. A risk score for chronic kidney disease in the general population. Am J Med. 2012; 125(3):270\u2013277. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3285426/ 16. Quon H, Grossman CE, King RL, Putt M, Donaldson K, Kricka LJ, Finlay J, Zhu T, Dimofte A, Malloy K, Cengel KA, Busch TM. Interference with the Jaff\u00e9 method for creatinine following 5-aminolevulinic acid administration. Photodiagnosis and Photodynamic Therapy. 2010; 7(4):268\u2013274. ISSN 1572-1000. https://doi.org/10.1016/j.pdpdt.2010.07.008 17. Roy W. Bonsnes, Hertha H. Taussky.", "24. https://doi.org/10.1007/s10404-019-2310-2 15. O\u2019Seaghdha CM, Lyass A, Massaro JM, et al. A risk score for chronic kidney disease in the general population. Am J Med. 2012; 125(3):270\u2013277. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3285426/ 16. Quon H, Grossman CE, King RL, Putt M, Donaldson K, Kricka LJ, Finlay J, Zhu T, Dimofte A, Malloy K, Cengel KA, Busch TM. Interference with the Jaff\u00e9 method for creatinine following 5-aminolevulinic acid administration. Photodiagnosis and Photodynamic Therapy. 2010; 7(4):268\u2013274. ISSN 1572-1000. https://doi.org/10.1016/j.pdpdt.2010.07.008 17. Roy W. Bonsnes, Hertha H. Taussky. On the Colorimetric Determination of Creatinine by the Jaffe Reaction. Journal of Biological Chemistry. 1945; 158(3):581\u2013591. https://doi.org/10.1016/S0021-9258(19)51334-5 18. Srisawasdi P, Chaichanajarernkul U, Teerakanjana N, Vanavanan S, Kroll MH. Exogenous interferences with Jaffe creatinine assays: Addition of sodium dodecyl sulfate to reagent eliminates bilirubin and total protein interference. J Clin Lab Anal. 2010; 35 24:123\u2013133. https://doi.org/10.1002/jcla.20350 19. Stevens LA, Levey AS. Measured GFR as a Confirmatory Test for Estimated GFR. J Am Soc Nephrol. 2009; 20(11):2305\u20132313. https://doi.org/10.1681/ASN.2009020171 20. Strong EB, Knutsen C, Wells JT, et al. Wax printed fluidic time delays for automating multi step assays in paper based microfluidic devices (micropads). Micromachines. 2019; 4(1):20. https://www.mdpi.com/2411-5134/4/1/20 21. Tambaru D, Rupilu RH, Nitti F, Gauru I, Suwari. Development of paper based sensor coupled with smartphone detector for simple creatinine determination. AIP Conference Proceedings. 2017. https://doi.org/10.1063/1.4978168 22. Tseng CC, Yang RJ, Ju WJ, Fu LM. Microfluidic paper based platform for whole blood creatinine detection. Chemical Engineering Journal. 2018; https://www.sciencedirect.com/science/article/abs/pii/S1385894718307642 23. Yetisen AK, Akram MS, Lowe CR. Paper based microfluidic point of care diagnostic devices. Lab Chip. 2013; 13(12):2210\u20132251. https://doi.org/10.1039/c3lc50169h 36", "Flow characteristics of Pediatric-sized Mechanical Heart Valves M.S. Biomedical Engineering Biomedical Engineering Department Charles W. Davidson College of Engineering August 7, 2025 Prepared by: Aadarsh Shivaraju Advised by: Dr. Alessandro Bellofiore San Jose State University Department Chair, Biomedical Engineering Reading Committee: Dr. Melinda Simon San Jose State University Assistant Professor, Biomedical Engineering Table of Contents List of Tables ................................................................................................................................... 2 List of Figures ................................................................................................................................. 2 Introduction .................................................................................................................................... 4 Literature Review .......................................................................................................................... 6 Research Objectives ..................................................................................................................... 11 Material and Methods ................................................................................................................. 13 I. Designing and Prototyping Pediatric-Sized Mechanical Heart Valve ........................... 13 II. Designing and Prototyping Pediatric-Sized Test Section .............................................. 16 III. Mock Circulation Loop and PIV Testing ...................................................................... 19 Results and Discussion ................................................................................................................. 24 Conclusion .................................................................................................................................... 38 References ..................................................................................................................................... 40 1 List of Tables Table 1 Experimental Conditions for Testing Table 2 Dimensional Evaluation of Mechanical Heart Valve (70\u00b0 Opening Angle) Table 3 Dimensional Evaluation of Mechanical Heart Valve (90\u00b0 Opening Angle) Table 4 Dimensional Evaluation of Test Section Table 5 Angular Dimensional Evaluation of Mechanical Heart Valve at Different Stroke Volume List of Figures Figure 1 Mechanical Heart Valve Size: Adult (Left) and Pediatric (Right) Figure 2 Solidworks model of hinge in the mechanical heart valve for the hinge Figure 3 Changes made to Hinge Region in Housing and Leaflets Figure 4 Overview of 3D Printer Process Figure 5 3D Printed Mechanical Heart Valve Figure 6 Solidworks model of Pediatric Test Section (Wireframe View) Figure 7 Solidworks model of Test Section Mold Figure 8 Sylgard 184 Pediatric Test Section Figure 9 Mock Circulation Loop Figure 10 Vivitro SuperPump (Left) and Controller (Right) Figure 11 DM Dual-head nanosecond laser used for PIV Figure 12 Phantom V2640 camera for PIV Figure 13 Average Flow Rate vs. Time Graph at 90 bpm for Single Heartbeat Figure 14 Pressure vs. Time Graph at 90 bpm for a Single Heartbeat: a) 20 mL/stroke and 70\u00b0 Opening Angle b) 20 mL/stroke and 90\u00b0 Opening Angle c) 30 mL/stroke and 70\u00b0 Opening Angle d) 30 mL/stroke and 90\u00b0 Opening Angle 2 Figure 15 PIV Velocity Maps at 90 bpm, 20 mL/stroke, 100 Ppk, and 70\u00b0 Opening Angle a) early systole b) peak systole c) late systole Figure 16 PIV Velocity Maps at 90 bpm, 20 mL/stroke, 100 Ppk, and 90\u00b0 Opening Angle a) early systole b) peak systole c) late systole Figure 17 PIV Velocity Maps at 90 bpm, 30 mL/stroke, 120 Ppk, and 70\u00b0 Opening", "c) 30 mL/stroke and 70\u00b0 Opening Angle d) 30 mL/stroke and 90\u00b0 Opening Angle 2 Figure 15 PIV Velocity Maps at 90 bpm, 20 mL/stroke, 100 Ppk, and 70\u00b0 Opening Angle a) early systole b) peak systole c) late systole Figure 16 PIV Velocity Maps at 90 bpm, 20 mL/stroke, 100 Ppk, and 90\u00b0 Opening Angle a) early systole b) peak systole c) late systole Figure 17 PIV Velocity Maps at 90 bpm, 30 mL/stroke, 120 Ppk, and 70\u00b0 Opening Angle a) early systole b) peak systole c) late systole Figure 18 PIV Velocity Maps at 90 bpm, 30 mL/stroke, 110 Ppk, and 90\u00b0 Opening Angle a) early systole b) peak systole c) late systole Figure 19 PIV Vorticity Maps at 90 bpm, 20 mL/stroke, 100 Ppk, and 70\u00b0 Opening Angle a) early systole b) peak systole c) late systole Figure 20 PIV Vorticity Maps at 90 bpm, 20 mL/stroke, 100 Ppk, and 90\u00b0 Opening Angle a) early systole b) peak systole c) late systole Figure 21 PIV Vorticity Maps at 90 bpm, 30 mL/stroke, 120 Ppk, and 70\u00b0 Opening Angle a) early systole b) peak systole c) late systole Figure 22 PIV Vorticity Maps at 90 bpm, 30 mL/stroke, 110 Ppk, and 90\u00b0 Opening Angle a) early systole b) peak systole c) late systole 3 Introduction Every year, about 1.35 million children are born worldwide with congenital heart defects (13) . Heart valve defects represent the most common congenital cardiovascular anomaly, accounting for over 25% of all congenital heart diseases (13,30) . Aortic stenosis, mitral stenosis, and regurgitation are the most common heart valve defects seen in children (29) . Aortic or mitral stenosis is a heart valve disease where the aortic or mitral valve narrows, which causes blood flow from the heart to be restricted. Aortic or mitral regurgitation occurs when blood flows backward from the aortic or mitral valve due to the valve not closing tightly. Surgical intervention may be required to repair the heart valve defects in children (2,13) . The goal of the intervention is to restore heart function to affected children by repairing the native valve to allow for tissue growth (13) . However, valve repair is often not feasible or successful, so heart valve replacement is performed on children with congenital heart valve defects (2) . There are several options for aortic or mitral valve replacement, including the Ross or Ross II procedures,", "intervention may be required to repair the heart valve defects in children (2,13) . The goal of the intervention is to restore heart function to affected children by repairing the native valve to allow for tissue growth (13) . However, valve repair is often not feasible or successful, so heart valve replacement is performed on children with congenital heart valve defects (2) . There are several options for aortic or mitral valve replacement, including the Ross or Ross II procedures, bioprosthetic valves, aortic or mitral homografts, and mechanical valves (6) . The Ross or Ross II procedure is a cardiac surgery that involves transferring the patient's pulmonary valve to the aortic or mitral position, while replacing the pulmonary valve with a biological valve (3,6) . This effectively changes the single-valve condition into one involving two valves (3,6) . While the Ross procedure is the most commonly used and preferred option for pediatric patients, it is a complex surgery that may require reintervention (6,12,29) . Bioprosthetic heart valves and aortic or mitral valve homografts are used for valve replacement in children, but they have the drawback of being less durable in young patients than in older patients due to accelerated structural deterioration from calcification and immune attack (6,13) . Mechanical heart 4 valves are durable and are often more favorable for younger patients, but require patients to take lifelong anticoagulation drugs like warfarin (6,14,25) . There are only two FDA-approved mechanical heart valves specifically designed for pediatric patients: the Abbott Masters Series 15 mm mechanical heart valve and the Medtronic Open Pivot AP 360 Supra-annular 16 mm mechanical heart valve. There is limited research on valve replacement outcomes in children under 2 years old, but the existing data indicates that mechanical heart valve replacements carry a high risk of complications, including thrombosis and bleeding (20) . Studies have also identified the supra-annular position as a risk factor for both thrombosis and the need for reoperation (20) . The supra-annular position enables the placement of a larger valve, but it comes with several drawbacks, including the need for higher levels of anticoagulant drugs, heart block, reduction of left atrial volume and compliance, compression of the circumflex coronary artery, and aneurysm formation (9,21) . In addition to the position of the pediatric valve, another consideration important for mechanical heart valves is the unique flow dynamics in pediatric patients. Pediatric patients have been shown", "the need for reoperation (20) . The supra-annular position enables the placement of a larger valve, but it comes with several drawbacks, including the need for higher levels of anticoagulant drugs, heart block, reduction of left atrial volume and compliance, compression of the circumflex coronary artery, and aneurysm formation (9,21) . In addition to the position of the pediatric valve, another consideration important for mechanical heart valves is the unique flow dynamics in pediatric patients. Pediatric patients have been shown to have higher heart rates and more hemodynamic conditions than adults. Pediatric flow conditions have been associated with higher fluid shear stress and greater potential for platelet damage compared to adult conditions (30) . The design of the mechanical heart valve is an important consideration as valves must account for the unique hemodynamic environment in children. Small design variations, such as the angle of leaflet movement during valve closure, can significantly impact flow-induced stress and the likelihood of platelet activation on the valve (11) . The goal of this study is to measure hemodynamics performances of pediatric-sized valves with different transverse angles. Flow characteristics of pediatric-sized valves will be 5 observed and comparisons will allow for insight into how the transverse angle affects hemodynamic performance. The hemodynamic performances being observed are flow rate, pressure gradient and the flow fields as the valves open and close. The data will allow for better modifications of the pediatric-sized valves for pediatric patients as the design of the valve is important to minimize platelet damage and thrombus formation. Literature Review Current approaches to heart valve replacement include Ross or Ross II procedure, bioprosthetic valves, homograph valves, and mechanical valves. The Ross procedure is a pulmonary autograft in the aortic position, where the pulmonary valve is transferred into the aortic position, and an allograft in the pulmonary position, where a biological valve is used to replace the pulmonary valve (6,10) . The Ross II procedure is a modification of the Ross procedure for the mitral position instead of the aortic position (3) . The Ross procedure has proven to have excellent hemodynamics, low occurrence of clotting and bleeding, allows for the valve to increase in size with somatic growth, and lower risk of endocarditis (6) . Endocarditis is the inflammation of the inner lining of the valve. However, current U.S. and European guidelines on aortic valve surgery are hesitant on the use of the Ross", "is a modification of the Ross procedure for the mitral position instead of the aortic position (3) . The Ross procedure has proven to have excellent hemodynamics, low occurrence of clotting and bleeding, allows for the valve to increase in size with somatic growth, and lower risk of endocarditis (6) . Endocarditis is the inflammation of the inner lining of the valve. However, current U.S. and European guidelines on aortic valve surgery are hesitant on the use of the Ross procedure as there are low patient numbers that have presented excellent long-term outcomes (1) . Dilation of the aortic root is the cause of late failure of the procedure (6,7) . The issue with Ross procedure is that the operation is complex, difficult, and highly variable for children (1,6,7) . Reoperation of the Ross procedure is not recommended as continued reoperations will cause the aortic valve to degenerate (1) . Bioprosthetic valves are made from animal tissue such as bovine, equine, or porcine pericardium or porcine aortic valves (15) . The benefits of bioprosthetic valves is that they have excellent hemodynamics, lower rates of bleeding than mechanical heart valves, and no need for 6 anticoagulation drugs (15,27) . Two types of bioprosthetic valves are stented and stentless bioprosthetic valves. Stentless valves have better durability and hemodynamics due to greater effective orifice area than stented valves, but are more difficult to implant (4,24) . Bioprosthetic valves are increasing in use than mechanical heart valves, as 80% of all implanted prosthetic heart valves are bioprosthetic valves compared to 40% in the 1990s (27) . This is due to the desire of elderly population to avoid lifelong anticoagulation drugs (27) . A disadvantage of bioprosthetic valves is that they are less durable than mechanical heart valves. This is due accelerated structural deterioration from calcification and immune attacks (13) . Homograph valves are made from biological cryopreserved human cadaveric donor valves (13) . Advantages of using homograph valves is that they offer excellent hemodynamics, good hemostasis, low risk of thromboembolism, and low risk of valve infections (16) . The disadvantages are the same as bioprosthetic valves in that they are less durable than mechanical heart valves due to accelerated structural deterioration from calcification and immune attacks (13) . Another disadvantage is there is limited accessibility to properly sized homografts (23) . Reoperation is also difficult for homograph valves due to the reimplantation of the", "of using homograph valves is that they offer excellent hemodynamics, good hemostasis, low risk of thromboembolism, and low risk of valve infections (16) . The disadvantages are the same as bioprosthetic valves in that they are less durable than mechanical heart valves due to accelerated structural deterioration from calcification and immune attacks (13) . Another disadvantage is there is limited accessibility to properly sized homografts (23) . Reoperation is also difficult for homograph valves due to the reimplantation of the coronary arteries being complicated (16) . Mechanical heart valves are made of materials consisting of titanium and carbon (12) . Mechanical heart valves offer superior durability compared to tissue valves, making them a favorable option for pediatric patients (6,14,25) . However, they require lifelong anticoagulation drugs, which carries risks of bleeding and thromboembolism (6,14,25) . The anticoagulation drug patients take is warfarin (19) . Other complications for mechanical heart valves in children are the limited size options for pediatric patients and the need for reoperation when children grow. The Abbott Masters Series 15 mm mechanical heart valve and the Medtronic Open Pivot AP 360 7 Supra-annular 16 mm mechanical heart valve are the only two FDA-approved mechanical heart valves designed for pediatric patients. Research on the outcomes of valve replacement is scarce for children younger than 2 years old, but the available data show that mechanical heart valve replacement is associated with high complication rates, thrombosis, and bleeding (20) . For adult patients, particularly those under 50-60 years old, mechanical valves are often recommended due to their excellent long-term durability (26) . The current design of the mechanical heart valve is a bileaflet design. The bileaflet mechanical valves were introduced in the late 1970s, the St. Jude Medical bileaflet valve, first implanted in 1977, quickly became the most widely used mechanical valve worldwide (4,8) . Bileaflet valves consist of two semicircular leaflets that pivot open and closed within a metal housing (4,24) . The opening angle of the leaflets relative to the annulus plane ranges from 70 to 90 degrees (24) . During forward flow, the bileaflet valves are open and there are two lateral orifice jets on sides of the leaflet facing the wall and a central orifice jet between the leaflets (22) . During reverse flow, the bileaflet valves are closed and there is leakage flow at the hinges and the b-datum plane (central opening gap between the", "housing (4,24) . The opening angle of the leaflets relative to the annulus plane ranges from 70 to 90 degrees (24) . During forward flow, the bileaflet valves are open and there are two lateral orifice jets on sides of the leaflet facing the wall and a central orifice jet between the leaflets (22) . During reverse flow, the bileaflet valves are closed and there is leakage flow at the hinges and the b-datum plane (central opening gap between the leaflet when closed) (22) . The position of a mechanical valve in a child significantly impacts its performance and associated risks. The standard position in patients is the annular positions. In children, the size of the valves varies and growth of the valve needs to be considered, which causes valve placement in supra-annular position (9,21) . In pediatric patients, supra-annular positioning has been identified as a risk factor for thrombosis and reoperation (9,21) . While supra-annular placement allows for a larger valve size, it carries several potential drawbacks, such as higher levels of anticoagulation may be required, risk of reduced left atrial volume/compliance, potential for heart block, risk of aneurysm formation, possible compression of the circumflex coronary artery (9,21) . Careful 8 preoperative imaging and sizing are crucial to determine the optimal valve position and size for each patient. The unique hemodynamic conditions in pediatric patients pose significant challenges for mechanical valve function. Pediatric patients typically have higher heart rates, more dynamic hemodynamic conditions, smaller vessel and chamber sizes. The important hemodynamic conditions for adults and pediatrics that are important are pressure gradient, stroke volume, heart rate, and cardiac output. The blood pressure gradient range for adults is from 36 to 50 mmHg and for children the range is from 18 to 22 mmHg (21,26) . The stroke volume for adults is 72 mL/stroke and for children is 27.8 mL/stroke (21,26) . The heart rate range for adults is from 43 to 98 bpm and for children the range is from 45 to 141 bpm (26) . The cardiac output generally ranges from 5-6 L/min (5) . In a study comparing the flow dynamics and blood damage between adult, child, and infant mechanical heart valves, it was found that child and infant valves had higher peak inflow velocity and shear stress than the adult valves (30) . There is shorter exposure time of shear stress on the platelets due to", "43 to 98 bpm and for children the range is from 45 to 141 bpm (26) . The cardiac output generally ranges from 5-6 L/min (5) . In a study comparing the flow dynamics and blood damage between adult, child, and infant mechanical heart valves, it was found that child and infant valves had higher peak inflow velocity and shear stress than the adult valves (30) . There is shorter exposure time of shear stress on the platelets due to higher heart rate for child and infant valves, but accumulated platelet damage was higher than adult valves (30) . The higher heart rates of children imposes more shear stress on the heart valves, which requires the valves to be durable. As the valve must function effectively across broad ranges of physiological conditions, the design requirements are more complicated than adult valves. The design of mechanical heart valves for pediatric patients requires careful consideration of the unique hemodynamic conditions in young children. A critical aspect of the design of the pediatric mechanical heart valve is the management of fluid shear stresses, which are known to contribute to platelet activation and subsequent thrombus formation. Research has demonstrated that even small design variations, such as the angle of leaflet traverse during valve 9 closure, can significantly impact the flow-induced stresses and the likelihood of platelet activation (11) . This study shows that valves with smaller traverse angles, such as 55\u00b0, produce lower shear stresses and reduce the potential for platelet activation compared to valves with larger angles, such as 64\u00b0 (11) . This suggests that optimizing the traverse angle is crucial for minimizing thrombus formation in pediatric patients, where the smaller size of the valve and the higher heart rates increase the likelihood of complications (11) . Additionally, the findings indicate that local geometrical changes to the leaflet and housing can further reduce stress concentrations, although these effects are less pronounced than those related to the traverse angle (11) . Another study found that flow was more centralized in valves with larger opening angles, but increased wall shear stresses (17) . In the context of pediatric valve design, it is also essential to minimize recirculation zones within the valve. Research has found that recirculation zones induce platelet damage (30) . A study noted that in pediatric flows, the leaflets were associated with higher shear stress and damaged platelets, and the damaged platelets were", "related to the traverse angle (11) . Another study found that flow was more centralized in valves with larger opening angles, but increased wall shear stresses (17) . In the context of pediatric valve design, it is also essential to minimize recirculation zones within the valve. Research has found that recirculation zones induce platelet damage (30) . A study noted that in pediatric flows, the leaflets were associated with higher shear stress and damaged platelets, and the damaged platelets were found far downstream of the valve (30) . This differs from adult flows, where damaged platelets are more likely to remain near the valve due to stronger recirculation zones (30) . If recirculation zones form in pediatric valves, it will lead to higher damaged platelets that increases the chances of thrombus formation. Therefore, designing the valve to reduce or eliminate recirculation zones near the leaflets can help mitigate the potential for thromboembolic complications, a critical consideration in pediatric heart valve design. The flow dynamics through the mechanical heart valves in the research have all been 2D or 3D computational fluid dynamics model simulations (11,17,22,30) . When investigating flow dynamics of mechanical heart valves in vitro, particle image velocimetry (PIV) can be used to detect leakage jets and quantify bulk fluid dynamics (18,22,30) . 10 The aim of this study is to evaluate the hemodynamics performances of pediatric-sized valves with different transverse angles, with a focus on comparing their flow characteristics. By observing the flow dynamics of pediatric-sized valves, the study will provide insight into how the transverse angle affects hemodynamic performance. The specific hemodynamic metrics being measured include flow rate, pressure gradient, and the flow fields during systole. This data will help improve the design of pediatric valves, which is critical to minimizing platelet damage and reducing the risk of thrombus formation in young patients. Research Objectives The objectives of this project is to design and prototype pediatric-size mechanical heart valves with varying transverse angles, design and prototype the test section for the pediatric sized mechanical heart valve, and compare hemodynamic performances of pediatric-sized valves with different transverse angles. Through the literature review, it was found that smaller transverse angles produce lower shear stresses and reduce potential for platelet activation (11) . Additionally, larger opening angles were found to have more centralized flow, but increased wall shear stresses (17) . These studies used adult mechanical valves, which have different hemodynamic", "valves with varying transverse angles, design and prototype the test section for the pediatric sized mechanical heart valve, and compare hemodynamic performances of pediatric-sized valves with different transverse angles. Through the literature review, it was found that smaller transverse angles produce lower shear stresses and reduce potential for platelet activation (11) . Additionally, larger opening angles were found to have more centralized flow, but increased wall shear stresses (17) . These studies used adult mechanical valves, which have different hemodynamic conditions and performances than pediatric patients. Pediatric heart valves were found to have higher peak velocity and shear stress compared to adult valves (30) . The opening angle of leaflets relative to the annulus plane ranges from 70 to 90 degrees (24) . The closing angle of the leaflets relative to the annulus plane is 30 degrees for St. Jude mechanical heart valve (11) . To observe the effects of varying transverse angles on pediatric mechanical valves, the project will focus on three specific aims. Aim 1: Design and Prototype pediatric-sized mechanical heart valves that have varying transverse angles. 11 In this aim, prototypes of the pediatric mechanical heart valves will be designed with the opening angles of 70 degrees and 90 degrees in SolidWorks. The SolidWorks files will then be sent to the 3D printer in the lab to get prototyped. It is important that leaflets fit in the housing and move smoothly. Aim 2: Design and Prototype pediatric-sized mechanical heart valve test section. In this aim, the test section and mold of the test section is modeled in SolidWorks. The mold SolidWorks files will then be sent to the 3D printer in the lab to get printed using PVB filament. The PVB mold will be smoothened using isopropyl alcohol. Sylgard 184 will be poured into the mold and cured for 3 days. The mold will then be removed after being placed in an isopropyl alcohol dissolution tank. Aim 3: Evaluate the hemodynamic performance of pediatric-sized mechanical heart valves for varying traverse angles. In this aim, the mechanical heart valve and test section will be placed on the mock circulation loop, where flow rate, pressure gradient, and velocity magnitude in flow field during systole will be measured. Flow rate will be measured using a flow meter, pressure gradient will be using gauge and differential pressure sensors, and flow fields will be measured using PIV. PIV will allow the visualization", "3: Evaluate the hemodynamic performance of pediatric-sized mechanical heart valves for varying traverse angles. In this aim, the mechanical heart valve and test section will be placed on the mock circulation loop, where flow rate, pressure gradient, and velocity magnitude in flow field during systole will be measured. Flow rate will be measured using a flow meter, pressure gradient will be using gauge and differential pressure sensors, and flow fields will be measured using PIV. PIV will allow the visualization of flow patterns through the aortic valve. Velocity magnitude and vorticity flow field images will be obtained to analyze the flow patterns. These measurements will help evaluate and compare the hemodynamic behavior between the valves with varying transverse angles. 12 Material and Methods I. Designing and Prototyping Pediatric-Sized Mechanical Heart Valve The Solidworks files of a 27.18 mm mechanical heart valve was scaled down to 15.49 mm size (Figure 1). This was done to the housing and leaflets by using the Scale tool in Solidworks and the scaling is 0.57 of the original size. Figure 1 : Mechanical Heart Valve Size: Adult (Left) and Pediatric (Right) The 15.49 mm mechanical heart valves of varying transverse angles were made by changing the inner angle of the hinge. As seen in Figure 2, the two mechanical heart valves were designed with the opening angles of 70 degrees and 90 degrees. 13 Figure 2 : Solidworks model of hinge in the mechanical heart valve for the hinge Due to shrinking the mechanical heart valve, the hinge region hole of the housing and the hinge connecting area of the leaflets are deepened. As seen in Figure 3, the modifications were made so the leaflets can be placed on the housing without falling out. Figure 3 : Changes made to Hinge Region in Housing and Leaflets Using Bambu Lab X1-Carbon 3D Printer (Figure 4a) in the Cardio Lab, the two mechanical heart valves were prototyped using Black PC filament (Figure 4b). This was done by converting the Solidworks files to STL files and sending them to the Bambu Studio Program (Figure 4c). 14 Figure 4 : Overview of 3D Printer Process In Bambu Studio, the files were exported to the Bambu Lab X1-Carbon 3D Printer, where the mechanical heart valves were 3D printed (Figure 5). The Bambu Lab X1-Carbon 3D printer is a 3D printer with a 0.4 mm nozzle that can print in multiple", "using Black PC filament (Figure 4b). This was done by converting the Solidworks files to STL files and sending them to the Bambu Studio Program (Figure 4c). 14 Figure 4 : Overview of 3D Printer Process In Bambu Studio, the files were exported to the Bambu Lab X1-Carbon 3D Printer, where the mechanical heart valves were 3D printed (Figure 5). The Bambu Lab X1-Carbon 3D printer is a 3D printer with a 0.4 mm nozzle that can print in multiple colors and multiple materials. PC filament is 1.75 mm in diameter and the source of the filament is from Polymaker. 15 Figure 5 : 3D Printed Mechanical Heart Valve The dimensions of the 3D printed mechanical heart valve were measured with a caliper and compared to the original SolidWorks model dimensions to determine whether the valve was printed successfully. The acceptance criteria for successfully printing the valves were having all dimensions within \u00b10.1mm of SolidWorks model dimensions. The caliper measured the valves five times and the average value of the measurements was used to compare with SolidWorks model dimensions. II. Designing and Prototyping Pediatric-Sized Test Section Using SolidWorks, the test section for a pediatric-sized mechanical heart valve was designed for a mock circulation loop in Cardio Lab for PIV Testing (Figure 6). Figure 6 : Solidworks model of Pediatric Test Section (Wireframe View) 16 The ends of the test section have a cylindrical shape that is 20.32 mm long and outer diameter of 25.33 mm, so it can be attached to the tubing in the mock circulation loop. The rectangular section of the test section is 100 mm long and 25.33 mm wide. A hole is made through the center of the whole test section with a diameter of 15.45mm. The hole is slightly smaller than the mechanical heart valve for snug fitting. 35 mm from the rectangular section of the test section, a smaller ring with the diameter of 14.95 mm with length of 1mm is made for the placing of mechanical heart valves. 16.5 mm from both ends of the rectangular section, 5 mm pressure port holes are made. These holes are where the pressure port connectors were placed and are 4.94 mm deep. Due to the test section being made out of Sylgard 184, a mold was created using the SolidWorks model for the test section with 3.65 mm thick walls for the rectangular section and 2", "14.95 mm with length of 1mm is made for the placing of mechanical heart valves. 16.5 mm from both ends of the rectangular section, 5 mm pressure port holes are made. These holes are where the pressure port connectors were placed and are 4.94 mm deep. Due to the test section being made out of Sylgard 184, a mold was created using the SolidWorks model for the test section with 3.65 mm thick walls for the rectangular section and 2 mm thick walls for the cylindrical section (Figure 7). The top portion of the mold is cut off, so Sylgard 184 can be poured into it. Figure 7 : Solidworks model of Test Section Mold The Solidworks file of the test section mold was converted into STL file and sent to the Bambu Studio Program, where the files were exported to Bambu Lab X1-Carbon 3D Printer to be 3D printed. The mold was printed out of 1.75 mm Blue PVB Filament from Polymaker. This 17 filament was chosen as it can be smoothen using 70% isopropyl alcohol, which will get rid of surface roughness from the 3D print. After the mold was smoothened by isopropyl alcohol, Sylgard 184 was prepared using a 9:1 ratio for base and curing agent. The total weight amount of Sylgard 184 was calculated using Mass Properties in SolidWorks on the test section model. The density of the model was changed in Mass Properties to match the density of Sylgard 184, to get the correct total weight amount of Sylgard 184 needed to fill the mold. After stirring the base and curing agent thoroughly, the Sylgard 184 is degassed in a vacuum chamber. The Sylgard is then poured into the mold, and the mold is then placed into a vacuum chamber to degas again. After degassing the Sylgard, the mold is placed outside to cure for 3 days. After 3 days of curing the Sylgard, the mold is placed into an isopropyl alcohol dissolution tank containing 99% isopropyl alcohol for the removal of the mold. After placing the mold in the isopropyl alcohol dissolution tank for 2 hours, the mold was removed and the PVB mold was peeled away from the Sylgard 184 Test Section. The image of Sylgard 184 Test Section can be seen in Figure 8. Figure 8 : Sylgard 184 Pediatric Test Section The dimensions of the Sylgard 184 pediatric test section were", "mold is placed into an isopropyl alcohol dissolution tank containing 99% isopropyl alcohol for the removal of the mold. After placing the mold in the isopropyl alcohol dissolution tank for 2 hours, the mold was removed and the PVB mold was peeled away from the Sylgard 184 Test Section. The image of Sylgard 184 Test Section can be seen in Figure 8. Figure 8 : Sylgard 184 Pediatric Test Section The dimensions of the Sylgard 184 pediatric test section were measured with a caliper and compared to the original SolidWorks model dimensions to determine whether the test sections were created successfully. The acceptance criteria for successful creation of the test section were having all dimensions within \u00b10.1mm of SolidWorks model dimensions. The 18 caliper measured the test section five times and the average value of the measurements was used to compare with SolidWorks model dimensions. Additionally, the 3D printed mechanical heart valve needs to fit in the test section. III. Mock Circulation Loop and PIV Testing The mock circulation loop is made up of several components. In Figure 9, the components that make up the mock circulation loop are the tubing for the water, the test section, resistance, compliance chambers, reservoir, differential pressure sensor, and gauge pressure sensors. There are three compliance chambers, two of them are big jars that are connected to the loop after the test section and the other compliance chamber is the glass tube before the test section. The blue container is the reservoir. One gauge pressure sensor is connected to the inlet of the test section to measure the pressure before it passes through the mechanical heart valve. The other gauge pressure sensor is connected to the outlet of the test section to measure the pressure after it passes through the mechanical heart valve. There is also a differential pressure sensor connected to the inlet and outlet of the test section. This differential pressure sensor will also measure the pressure gradient across the mechanical heart valve. The three sensors are connected to two power supplies and a computer with a LabView application. The LabView application has two files, a monitor pressure file and read pressure file. The monitor pressure file is used to monitor peak aortic pressure across the valves, and make sure that the peak aortic pressure is right for the PIV test being run. The read pressure file is used to collect pressure", "will also measure the pressure gradient across the mechanical heart valve. The three sensors are connected to two power supplies and a computer with a LabView application. The LabView application has two files, a monitor pressure file and read pressure file. The monitor pressure file is used to monitor peak aortic pressure across the valves, and make sure that the peak aortic pressure is right for the PIV test being run. The read pressure file is used to collect pressure data for the experimental run. The pressure data collected by the file is differential pressure, aortic pressure, and left ventricular pressure. A flow meter was installed on the loop after the test section of the mock circulation loop to measure the flow rate of the loop. The flow meter is connected to another power supply and computer with the LabChart application. The LabChart application is 19 used to record the flow rate collected by the flow meter. Five trials of flow rate and pressure data were collected and each trial measured 10 seconds of data. The flow rate and pressure data were collected simultaneously, but due to the trials not being time aligned, a single heartbeat was extracted from each trial. For flow rate data, heartbeats corresponding to peak flow rate were extracted and time-aligned. The heartbeats were averaged to generate a flow rate graph containing the flow rate profile for a single heartbeat for the four test conditions. For pressure data, heartbeat corresponding to peak left ventricular data was extracted and time-aligned. Calculated differential pressure gradient data was obtained by subtracting aortic pressure data from the left ventricular data. All four pressure data from all five trials for each test condition were then averaged. For each test condition, a pressure graph for a heartbeat was created with four profiles corresponding to differential pressure, left ventricular pressure, aortic pressure, and calculated differential pressure. The flow rate and pressure data was collected at a separate time than PIV data due to not being able to collect them at the same time with the current setup. The mechanical heart valve prototyped was placed inside the test section in the mock circulation loop. The test section was placed on a test section holder that was 3D printed out of the black PC filament using the 3D printer. The test section holder is there to make sure that the test section is level and", "collected at a separate time than PIV data due to not being able to collect them at the same time with the current setup. The mechanical heart valve prototyped was placed inside the test section in the mock circulation loop. The test section was placed on a test section holder that was 3D printed out of the black PC filament using the 3D printer. The test section holder is there to make sure that the test section is level and straight as twists will affect the data collected by PIV. Clamps were attached to the test section to make sure the leaflets open smoothly in the valve housing. The test section and pressure sensors were connected by 3D printed pressure port connectors made out of PLA. The blood analog solution running through the mock circulation loop is a solution made out of 54% water, 12% sodium chloride and 33% glycerol. 20 Figure 9 - Mock Circulation Loop The mock circulation loop is connected to the Vivitro SuperPump and controlled using the Vivitro SuperPump Controller, as seen in Figure 10, which will pump pulsatile flow throughout the mock circulation loop. The controller allows for the programming of the SuperPump. It was used to set the mL/stroke and bpm of SuperPump to physiological condition in children by setting the mL/Stroke to 30 mL/stroke and bpm to 90 bpm. Figure 10 - Vivitro SuperPump (Left) and Controller (Right) 21 To obtain the velocity map of the valve during systole, PIV needs to be set up for the mock circulation loop. The PIV setup consisted of a DM Dual-head nanosecond laser (Photonics industries) with high pulse energy (Nd: YLF, 527 nm) used with 2D laser pulse synchronizer, and DaVis 11 software. This can be seen in Figure 11. Figure 11 - DM Dual-head nanosecond laser used for PIV The camera used in PIV is the Phantom V2640 camera, which captures 2048\u00d71024 pixel images at a rate of 5,000 fps. The camera can be seen in Figure 12. Figure 12 - Phantom V2640 camera for PIV 22 Hollow glass particles of 8 to 12 \u03bc m in diameter from TSI were used to seed the blood analog solution in the mock circulation loop. These particles were illuminated by the laser at a wavelength of 527 nm and images of the particles were captured by the camera. The images were processed through the DaVis 11 software", "a rate of 5,000 fps. The camera can be seen in Figure 12. Figure 12 - Phantom V2640 camera for PIV 22 Hollow glass particles of 8 to 12 \u03bc m in diameter from TSI were used to seed the blood analog solution in the mock circulation loop. These particles were illuminated by the laser at a wavelength of 527 nm and images of the particles were captured by the camera. The images were processed through the DaVis 11 software to obtain velocity and vorticity maps. Velocity and vorticity maps were obtained for early systole, peak systole, and late systole. Five images of the valve at peak systole for each test condition were used to find the opening angle of the valve. The average angular dimensions of the mechanical heart valve was obtained using ImageJ software and compared to the original SolidWorks model dimensions to determine whether the valves were opening successfully. The acceptance criteria for successful opening of the mechanical were having all angular dimensions within \u00b11\u00b0 of SolidWorks model dimensions. The experimental conditions can be seen in Table 1. The testing of 2 mechanical heart valves consisted of 2 different stroke volumes for 4 seconds. The flow rate was measured in L/min by the flow meter. The pressure gradient was calculated by the two gauge pressure sensors and compared to the pressure gradient measured by differential pressure sensor. The graphs obtained from flow meters and pressure sensors were compared to each other and physiological flow conditions. The peak velocity magnitude was obtained from the velocity map during peak systole. The velocity map compared to each other and physiological flow conditions. Table 1: Experimental Conditions for Testing Stroke Volume Heart Rate Peak Aortic Pressure (26) Opening Angle Duration 20 mL/Stroke 90 bpm 100 mmHg 70\u00b0 4 second 20 mL/Stroke 90 bpm 100 mmHg 90\u00b0 4 second 30 mL/Stroke 90 bpm 100 mmHg 70\u00b0 4 second 30 mL/Stroke 90 bpm 100 mmHg 90\u00b0 4 second 23 Results and Discussion To determine the success of Aims 1 and 2, the dimensions of the mechanical heart valve were evaluated to the acceptance criteria outlined in the Material and Methods section. The results are shown in Tables 2-5. Table 2 evaluates the dimensions of the mechanical heart valve with a 70\u00b0 opening angle. The valve housing length, leaflet thickness, and length for leaflet 2 met the specifications for mechanical heart valve with a", "100 mmHg 90\u00b0 4 second 23 Results and Discussion To determine the success of Aims 1 and 2, the dimensions of the mechanical heart valve were evaluated to the acceptance criteria outlined in the Material and Methods section. The results are shown in Tables 2-5. Table 2 evaluates the dimensions of the mechanical heart valve with a 70\u00b0 opening angle. The valve housing length, leaflet thickness, and length for leaflet 2 met the specifications for mechanical heart valve with a 70\u00b0 opening angle. However, the valve housing thickness, leaflet height, and length for leaflet 1 failed to meet the specifications for mechanical heart valve with a 70\u00b0 opening angle. Table 2: Dimensional Evaluation of Mechanical Heart Valve (70\u00b0 Opening Angle) Dimension Solidworks (mm) Measured (mm) Difference (mm) Tolerance (mm) Pass/Fail Housing Length 15.491 15.494 0.003 \u00b10.1 Pass Housing Thickness 0.869 1.102 0.233 \u00b10.1 Fail Leaflet 1 Length 12.306 12.162 -0.144 \u00b10.1 Fail Leaflet 1 Height 7.738 7.402 -0.336 \u00b10.1 Fail Leaflet 1 Thickness 0.492 0.533 0.041 \u00b10.1 Pass Leaflet 2 Length 12.306 12.228 -0.078 \u00b10.1 Pass Leaflet 2 Height 7.738 7.452 -0.286 \u00b10.1 Fail Leaflet 2 Thickness 0.492 0.533 0.041 \u00b10.1 Pass Table 3 evaluates the dimensions of the mechanical heart valve with a 90\u00b0 opening angle. The valve housing length, leaflet thickness, and length for leaflet 1 met the specifications for mechanical heart valve with 90\u00b0 opening angle. However, the valve housing thickness, leaflet height, and length for leaflet 2 failed to meet the specifications for mechanical heart valve with a 90\u00b0 opening angle. 24 Table 3: Dimensional Evaluation of Mechanical Heart Valve (90\u00b0 Opening Angle) Dimension Solidworks (mm) Measured (mm) Difference (mm) Tolerance (mm) Pass/Fail Housing Length 15.491 15.494 0.003 \u00b10.1 Pass Housing Thickness 0.869 1.102 0.233 \u00b10.1 Fail Leaflet 1 Length 12.306 12.228 -0.078 \u00b10.1 Pass Leaflet 1 Height 7.738 7.417 -0.321 \u00b10.1 Fail Leaflet 1 Thickness 0.492 0.533 0.041 \u00b10.1 Pass Leaflet 2 Length 12.306 12.197 -0.109 \u00b10.1 Fail Leaflet 2 Height 7.738 7.483 -0.255 \u00b10.1 Fail Leaflet 2 Thickness 0.492 0.533 0.041 \u00b10.1 Pass Table 4 evaluates the dimensions of the test section. The side Length of the cube region and inner diameter of the test section met the specifications for the test section. However, the side height, side width and total length of the test section failed to meet the specifications for the test section. Table 4: Dimensional Evaluation of Test Section Dimension Solidworks", "-0.109 \u00b10.1 Fail Leaflet 2 Height 7.738 7.483 -0.255 \u00b10.1 Fail Leaflet 2 Thickness 0.492 0.533 0.041 \u00b10.1 Pass Table 4 evaluates the dimensions of the test section. The side Length of the cube region and inner diameter of the test section met the specifications for the test section. However, the side height, side width and total length of the test section failed to meet the specifications for the test section. Table 4: Dimensional Evaluation of Test Section Dimension Solidworks (mm) Measured (mm) Difference (mm) Tolerance (mm) Pass/Fail Side Height 28.33 27.473 -0.857 \u00b10.1 Fail Side Width 25.33 25.222 -0.108 \u00b10.1 Fail Side Length - Cube 100 100.071 0.071 \u00b10.1 Pass Inner Diameter 15.45 15.408 -0.042 \u00b10.1 Pass Total Length 144 144.75 0.75 \u00b10.1 Fail Table 5 evaluates the opening angle of the mechanical heart valve at different stroke volumes. At 30 mL/stroke, leaflet 2 of the mechanical heart valve with a 70\u00b0 opening angle met 25 the specifications. The rest of the angular measurements were not within acceptable range and deviated it by a large amount. Table 5: Angular Dimensional Evaluation of Mechanical Heart Valve at Different Stroke Volume Dimension Stroke Volume (mL/stroke) Solidworks (\u00b0) Measured (\u00b0) Difference (\u00b0) Tolerance (\u00b0) Pass/Fail Angle 1 20 70 45.576 -24.424 \u00b11 Fail Angle 2 20 70 45.607 -24.393 \u00b11 Fail Angle 1 20 90 84.456 -5.544 \u00b11 Fail Angle 2 20 90 85.503 -4.497 \u00b11 Fail Angle 1 30 70 52.203 -17.797 \u00b11 Fail Angle 2 30 70 70.731 0.731 \u00b11 Pass Angle 1 30 90 83.194 -6.806 \u00b11 Fail Angle 2 30 90 83.631 -6.369 \u00b11 Fail Based on the results from Table 2-5, Aim 1can be considered unsuccessful and Aim 2 can be considered successful. Aim 1 is considered unsuccessful due to mechanical heart valves not opening to correct opening angle when run through the mock circulation loop and some of the valve dimensions not meeting specifications. Even when one leaflet opened to the correct opening angle, the other leaflet did not open fully causing asymmetrical opening. The leaflets open to similar angles for the rest of the test conditions. These angular discrepancies can be attributed to dimensional inaccuracies from 3D printing resolution and friction at the hinge region of the valve. While some dimensions did not meet specification the overall objective of the test section was achieved. Aim 2 is considered successful due to the test section fitting", "one leaflet opened to the correct opening angle, the other leaflet did not open fully causing asymmetrical opening. The leaflets open to similar angles for the rest of the test conditions. These angular discrepancies can be attributed to dimensional inaccuracies from 3D printing resolution and friction at the hinge region of the valve. While some dimensions did not meet specification the overall objective of the test section was achieved. Aim 2 is considered successful due to the test section fitting into the mock circulation loop, mechanical heart valve fit into the test section and PIV was run using the test section and clear images were taken. 26 Aim 3 focuses on measuring and evaluating hemodynamic behavior of pediatric mechanical heart valves of varying transverse angles and comparing them to physiological behaviors. Hemodynamic behaviors measured are flow rate, pressure gradient, and the velocity flow fields during systole. Figure 13 is the average flow rate vs time graphs for each of the four test conditions for a single heartbeat. The cardiac cycle in the graphs matches up with what is expected when the heart rate is 90 bpm. At 20 mL/stroke, the peak flow rate for mechanical heart with 70\u00b0 opening angle is 7.42 L/min and with 90\u00b0 opening angle it is 8.55 L/min. At 30 mL/stroke, the peak flow rate for mechanical heart with 70\u00b0 opening angle is 12.19 L/min and with 90\u00b0 opening angle it is 13.47 L/min. Figure 13 - Average Flow Rate vs. Time Graph at 90 bpm for Single Heartbeat The peak flow rate increases at the same stroke volume when the opening angle of the valve is larger, which is expected as more fluid can pass through the valve when the opening angle is larger. The peak flow rate also increases as stroke volume increases, which is expected 27 due to cardiac output increasing. The peak flow rate for each stroke volume was close to what is in literature (30) . Figure 14 is the average pressure vs time graphs for each of the four test conditions and contains all three pressures recorded by sensors and the calculated differential pressure for a single heartbeat. At 20 mL/stroke, the mechanical heart valve with 70\u00b0 opening angle has a peak differential pressure of 19.27 mmHg, a peak left ventricular pressure of 100.69 mmHg, a peak aortic pressure of 80.84 mmHg, and a peak calculated differential pressure of 27.73", "is in literature (30) . Figure 14 is the average pressure vs time graphs for each of the four test conditions and contains all three pressures recorded by sensors and the calculated differential pressure for a single heartbeat. At 20 mL/stroke, the mechanical heart valve with 70\u00b0 opening angle has a peak differential pressure of 19.27 mmHg, a peak left ventricular pressure of 100.69 mmHg, a peak aortic pressure of 80.84 mmHg, and a peak calculated differential pressure of 27.73 mmHg. At 20 mL/stroke, the mechanical heart valve with 90\u00b0 opening angle has a peak differential pressure of 16.68 mmHg, a peak left ventricular pressure of 100.35 mmHg, a peak aortic pressure of 86.97 mmHg, and a peak calculated differential pressure of 25.56 mmHg. At 30 mL/stroke, the mechanical heart valve with 70\u00b0 opening angle has a peak differential pressure of 37.75 mmHg, a peak left ventricular pressure of 117.24 mmHg, a peak aortic pressure of 86.61 mmHg, and a peak calculated differential pressure of 45.58 mmHg. At 30 mL/stroke, the mechanical heart valve with 90\u00b0 opening angle has a peak differential pressure of 22.01 mmHg, a peak left ventricular pressure of 107.26 mmHg, a peak aortic pressure of 82.66 mmHg, and a peak calculated differential pressure of 30.74 mmHg. 28 Figure 14 - Pressure vs. Time Graph at 90 bpm for a Single Heartbeat: a) 20 mL/stroke and 70\u00b0 Opening Angle b) 20 mL/stroke and 90\u00b0 Opening Angle c) 30 mL/stroke and 70\u00b0 Opening Angle d) 30 mL/stroke and 90\u00b0 Opening Angle Differential pressure and calculated differential pressure increased when opening angle decreased due to the smaller opening angle obstructing flow of fluid passing through the valve. Differential pressure and calculated differential pressure increased when stroke volume increased due to larger amounts of fluid passing through the valve. The trend of calculated difference pressure graph is similar to the differential pressure graph during systole, but has larger pressure valves compared to differential pressure graph. The graphs are also different during diastole, where negative values are not measured by differential pressure sensors and results in sensors giving -5.5 mmHg value. The differential pressure is not in the range for pediatric patients when 29 the mechanical heart valve has an opening angle of 70\u00b0 and stroke volume of 30 mL/stroke. The calculated differential pressure is not in range for pediatric patients for all test conditions. Figure 15 is the PIV Velocity", "compared to differential pressure graph. The graphs are also different during diastole, where negative values are not measured by differential pressure sensors and results in sensors giving -5.5 mmHg value. The differential pressure is not in the range for pediatric patients when 29 the mechanical heart valve has an opening angle of 70\u00b0 and stroke volume of 30 mL/stroke. The calculated differential pressure is not in range for pediatric patients for all test conditions. Figure 15 is the PIV Velocity Magnitude Maps at early systole, peak systole, and late systole when the test condition is 90 bpm, 20 mL/stroke, 100 Ppk, and 70\u00b0 Opening Angle. Using the flowrate graph, early systole happens at 0.03 seconds, peak systole happens at 0.12 seconds, and late systole happens 0.23 seconds. Both leaflets are open at early systole, there are two side jets and a small central jet but the jets are not uniform. At peak systole, there is a diagonal flow going upwards downstream and a central backflow. The peak flow velocity magnitude is 1.64 m/s, which is higher than the normal physiological value of 1.31 m/s (30) . At late systole, the backflow downstream narrows to a central backflow when it gets closer to the valve. 30 Figure 15 - PIV Velocity Maps at 90 bpm, 20 mL/stroke, 100 Ppk, and 70\u00b0 Opening Angle a) early systole b) peak systole c) late systole Figure 16 is the PIV Velocity Magnitude Maps at early systole, peak systole, and late systole when the test condition is 90 bpm, 20 mL/stroke, 100 Ppk, and 90\u00b0 opening angle. Using the flowrate graph, early systole happens at 0.03 seconds, peak systole happens at 0.12 seconds, and late systole happens 0.23 seconds. Both leaflets are open at early systole and side orifice jets are looking uniform, but there is no central orifice jet. At peak systole, there are three orifice jets and the central orifice jet is moving up further downstream. The peak flow velocity magnitude is 1.94 m/s, which is higher than the normal physiological value of 1.31 m/s (30) . At late systole, the backflow downstream splits to two side orifice backflow when it gets closer to the valve. Figure 16 - PIV Velocity Maps at 90 bpm, 20 mL/stroke, 100 Ppk, and 90\u00b0 Opening Angle 31 a) early systole b) peak systole c) late systole Figure 17 is the PIV Velocity Magnitude Maps at early", "up further downstream. The peak flow velocity magnitude is 1.94 m/s, which is higher than the normal physiological value of 1.31 m/s (30) . At late systole, the backflow downstream splits to two side orifice backflow when it gets closer to the valve. Figure 16 - PIV Velocity Maps at 90 bpm, 20 mL/stroke, 100 Ppk, and 90\u00b0 Opening Angle 31 a) early systole b) peak systole c) late systole Figure 17 is the PIV Velocity Magnitude Maps at early systole, peak systole, and late systole when the test condition is 90 bpm, 30 mL/stroke, 120 Ppk, and 70\u00b0 opening angle. Using the flowrate graph, early systole happens at 0.03 seconds, peak systole happens at 0.12 seconds, and late systole happens 0.23 seconds. Both leaflets are open at early systole, but the jets are not uniform and there is no central orifice jet. At peak systole, there are central jets downstream and there is central backflow jet near the valve. The peak flow velocity magnitude is 1.80 m/s and is located in the central backflow jet. The velocity magnitude is higher than the normal physiological value of 1.31 m/s (30) . At the late systole, there is a central backflow near the valve. There are also side orifice jets, the bottom jet has progressed further downstream Figure 17 - PIV Velocity Maps at 90 bpm, 30 mL/stroke, 120 Ppk, and 70\u00b0 Opening Angle 32 a) early systole b) peak systole c) late systole Figure 18 is the PIV Velocity Magnitude Maps at early systole, peak systole, and late systole when the test condition is 90 bpm, 30 mL/stroke, 110 Ppk, and 90\u00b0 opening angle. Using the flowrate graph, early systole happens at 0.03 seconds, peak systole happens at 0.12 seconds, and late systole happens 0.23 seconds. Both leaflets are open at early systole, but the jets are not uniform and there is no central orifice jet. At peak systole, there is a huge central jet that stretches from valve to further downstream. The peak flow velocity magnitude is 2.76 m/s and is located in the jets near the valve. The velocity magnitude is higher than the normal physiological value of 1.31 m/s (30) . At late systole, there is a backflow throughout the region. Figure 18 - PIV Velocity Maps at 90 bpm, 30 mL/stroke, 110 Ppk, and 90\u00b0 Opening Angle a) early systole b) peak systole c) late systole", "is a huge central jet that stretches from valve to further downstream. The peak flow velocity magnitude is 2.76 m/s and is located in the jets near the valve. The velocity magnitude is higher than the normal physiological value of 1.31 m/s (30) . At late systole, there is a backflow throughout the region. Figure 18 - PIV Velocity Maps at 90 bpm, 30 mL/stroke, 110 Ppk, and 90\u00b0 Opening Angle a) early systole b) peak systole c) late systole 33 Figure 19 is the PIV Vorticity Maps at early systole, peak systole, and late systole when the test condition is 90 bpm, 20 mL/stroke, 100 Ppk, and 70\u00b0 opening angle. Using the flowrate graph, early systole happens at 0.03 seconds, peak systole happens at 0.12 seconds, and late systole happens 0.23 seconds. At early systole, there are streamlines curving around the jets and the regions with higher vorticity value. After the jets, the streamlines are straight. At peak systole, there are tiny circular regions with high vorticity values surrounding the jets. There seems to be a lot of turbulent flow in these regions. At late systole, there are not many regions with high vorticity values, and in these regions there are curved streamlines signaling recirculation. They are near the valve, as the backflow is straight further downstream. Figure 19 - PIV Vorticity Maps at 90 bpm, 20 mL/stroke, 100 Ppk, and 70\u00b0 Opening Angle a) early systole b) peak systole c) late systole 34 Figure 20 is the PIV Vorticity Maps at early systole, peak systole, and late systole when the test condition is 90 bpm, 20 mL/stroke, 100 Ppk, and 90\u00b0 opening angle. Using the flowrate graph, early systole happens at 0.03 seconds, peak systole happens at 0.12 seconds, and late systole happens 0.23 seconds. At early systole, there are straight streamlines after the curved streamlines for the side orifice jets. High vorticity values are at the curved regions. At peak systole, there are many regions on the sides and where the leaflets are. Lots of turbulent flow throughout these regions, coming from the three orifice jets hitting the leaflet wall and wall of the test section. At late systole, there are regions at the top and bottom where there are curved and circular streamlines signaling recirculation. Figure 20 - PIV Vorticity Maps at 90 bpm, 20 mL/stroke, 100 Ppk, and 90\u00b0 Opening Angle a) early systole b)", "regions. At peak systole, there are many regions on the sides and where the leaflets are. Lots of turbulent flow throughout these regions, coming from the three orifice jets hitting the leaflet wall and wall of the test section. At late systole, there are regions at the top and bottom where there are curved and circular streamlines signaling recirculation. Figure 20 - PIV Vorticity Maps at 90 bpm, 20 mL/stroke, 100 Ppk, and 90\u00b0 Opening Angle a) early systole b) peak systole c) late systole 35 Figure 21 is the PIV Vorticity Maps at early systole, peak systole, and late systole when the test condition is 90 bpm, 30 mL/stroke, 120 Ppk, and 70\u00b0 Opening Angle. Using the flowrate graph, early systole happens at 0.03 seconds, peak systole happens at 0.12 seconds, and late systole happens 0.23 seconds. At early systole, there are straight streamlines after the curved streamlines for the side orifice jets. High vorticity values are at the curved regions. At peak systole, there are tiny high vorticity regions throughout the whole test section. Lots of turbulent flow due to a high number of tiny vorticity regions and curved and circular streamlines. At late systole, there is some turbulent flow due to the top backflow jet and flow jet at the bottom. Streamlines also show curved lines. Figure 21 - PIV Vorticity Maps at 90 bpm, 30 mL/stroke, 120 Ppk, and 70\u00b0 Opening Angle a) early systole b) peak systole c) late systole 36 Figure 22 is the PIV Vorticity Maps at early systole, peak systole, and late systole when the test condition is 90 bpm, 30 mL/stroke, 110 Ppk, and 90\u00b0 Opening Angle. Using the flowrate graph, early systole happens at 0.03 seconds, peak systole happens at 0.12 seconds, and late systole happens 0.23 seconds. At early systole, there are straight streamlines after the curved streamlines for the side orifice jets. High vorticity values are at the curved regions. At peak systole, there is a tiny high vorticity region surrounding the huge central flow jet. The high vorticity regions are caused by flow hitting the walls. Streamlines are mostly straight in the central flow area. At late systole, there are central regions at the top and bottom where there are curved and circular streamlines signaling recirculation. Figure 22 - PIV Vorticity Maps at 90 bpm, 30 mL/stroke, 110 Ppk, and 90\u00b0 Opening Angle a) early systole b)", "curved regions. At peak systole, there is a tiny high vorticity region surrounding the huge central flow jet. The high vorticity regions are caused by flow hitting the walls. Streamlines are mostly straight in the central flow area. At late systole, there are central regions at the top and bottom where there are curved and circular streamlines signaling recirculation. Figure 22 - PIV Vorticity Maps at 90 bpm, 30 mL/stroke, 110 Ppk, and 90\u00b0 Opening Angle a) early systole b) peak systole c) late systole 37 The peak flow velocity magnitude increased as the stroke volume increased, which makes sense as more fluid is being pushed through the loop. The peak flow velocity magnitude also increased as opening angle increased, which is consistent with findings reported in literature (17) . There was lots of turbulent flow at peak systole, due to flow not being streamline. One of the limitations of this project is the prototype mechanical heart didn\u2019t open to the correct opening angle. The opening angles of between the two valves differed significantly, which allowed for evaluation of how opening angle affects the results. However, the opening angles used in the results were not consistent with values in the aims and actual pediatric valves. This will cause results to not be consistent with the pediatric physiological conditions. Another limitation that affected the results was no sinus region in the test section. The sinus region helps with containing the recirculation by holding vortices within the sinus. By having no sinus region, the vortices formed downstream causing turbulent flow. A limitation of this project is that only one trial was done on each test condition. This was due to PIV taking too much data on the hard drive of the computer. Could only take one trial as there were multiple groups in the CardioLab that needed to use PIV. Based on the current limitation of the project, future research should focus on figuring out ways to 3D print the valve with leaflets that open smoothly or running the tests with fixed leaflets at varying angles. Running PIV tests with fixed leaflets will help show the difference between flow when angles are different Another option for future search is creating a new test section with a sinus region as it will help minimize the recirculation. Conclusion The overall goal of this project was to prototype and design pediatric mechanical heart valves with", "on figuring out ways to 3D print the valve with leaflets that open smoothly or running the tests with fixed leaflets at varying angles. Running PIV tests with fixed leaflets will help show the difference between flow when angles are different Another option for future search is creating a new test section with a sinus region as it will help minimize the recirculation. Conclusion The overall goal of this project was to prototype and design pediatric mechanical heart valves with varying transverse angles, prototype and design pediatric test section, and evaluate 38 and compare the hemodynamic performance of pediatric-sized mechanical heart valves with varying traverse angles. Aim 1 of the project was not achieved successfully due to the mechanical heart valves not opening to correct angles and some dimensions of valve not meeting specifications. Aim 2 of the project was achieved successfully due to the test section working in the mock circulation loop, the mechanical heart valve fitting into the test section, and PIV was run smoothly using the test section. Aim 3 was achieved successfully due to being able evaluate and compare the hemodynamic performance of pediatric-sized mechanical heart valves with varying traverse angles. The results showed that peak flow rate and peak velocity magnitudes were increased, while pressure gradient decreased when transverse angle was larger. Due to aim 1 not being achieved successfully, there were some limitations to the project such as not getting the correct opening angles for valves, no sinus regions, and only one PIV trial for data. 39 References 1. Aboud, A., E. I. Charitos, B. Fujita, U. Stierle, J.-C. Reil, V. Voth, M. Liebrich, M. Andreas, T. Holubec, C. Bening, M. Albert, P. Fila, J. Ondrasek, P. Murin, R. Lange, H. Reichenspurner, U. Franke, A. Gorski, A. Moritz, G. Laufer, W. Hemmer, H.-H. Sievers, and S. Ensminger. Long-term outcomes of patients undergoing the Ross procedure. J. Am. Coll. Cardiol. 77:1412\u20131422, 2021. 2. Alsoufi, B., C. Manlhiot, M. Al-Ahmadi, B. W. McCrindle, A. Kalloghlian, G. Siblini, Z. Bulbul, and Z. Al-Halees. Outcomes and associated risk factors for mitral valve replacement in children \u2606 . Eur. J. Cardiothorac. Surg. , 2011.doi:10.1016/j.ejcts.2010.12.060 3. Athanasiou, T., A. Cherian, and D. Ross. The Ross II procedure: Pulmonary autograft in the mitral position. Ann. Thorac. Surg. 78:1489\u20131495, 2004. 4. Bloomfield, P. Choice of heart valve prosthesis. Br. Heart J. 87:583\u2013589, 2002. 5. Bruss, Z. S., and A. Raja. Physiology, Stroke", "C. Manlhiot, M. Al-Ahmadi, B. W. McCrindle, A. Kalloghlian, G. Siblini, Z. Bulbul, and Z. Al-Halees. Outcomes and associated risk factors for mitral valve replacement in children \u2606 . Eur. J. Cardiothorac. Surg. , 2011.doi:10.1016/j.ejcts.2010.12.060 3. Athanasiou, T., A. Cherian, and D. Ross. The Ross II procedure: Pulmonary autograft in the mitral position. Ann. Thorac. Surg. 78:1489\u20131495, 2004. 4. Bloomfield, P. Choice of heart valve prosthesis. Br. Heart J. 87:583\u2013589, 2002. 5. Bruss, Z. S., and A. Raja. Physiology, Stroke Volume. StatPearls Publishing, 2022.at <https://www.ncbi.nlm.nih.gov/books/NBK547686/> 6. David, T. E. Aortic valve replacement in children and young adults. J. Am. Coll. Cardiol. 67:2871\u20132873, 2016. 7. David, T. E., C. David, A. Woo, and C. Manlhiot. The Ross procedure: Outcomes at 20 years. J. Thorac. Cardiovasc. Surg. 147:85\u201394, 2014. 8. DeWall, R. A., N. Qasim, and L. Carr. Evolution of mechanical heart valves. Ann. Thorac. Surg. 69:1612\u20131621, 2000. 9. Eltayeb, O. M., W. J. Readdy, M. C. Mong\u00e9, J. M. Forbess, A. E. Sarwark, A. Patel, and C. L. Backer. Mitral valve replacement in infants using a 15-mm mechanical valve. Ann. 40 Thorac. Surg. 108:552\u2013557, 2019. 10. Etnel, J. R. G., L. C. Elmont, E. Ertekin, M. M. Mokhles, H. J. Heuvelman, J. W. Roos-Hesselink, P. L. de Jong, W. A. Helbing, A. J. J. C. Bogers, and J. J. M. Takkenberg. Outcome after aortic valve replacement in children: A systematic review and meta-analysis. J. Thorac. Cardiovasc. Surg. 151:143-152.e3, 2016. 11. Govindarajan, V., H. S. Udaykumar, L. H. Herbertson, S. Deutsch, K. B. Manning, and K. B. Chandran. Impact of design parameters on bileaflet mechanical heart valve flow dynamics. J. Heart Valve Dis. 18:535\u2013545, 2009. 12. Harris, C., B. Croce, and C. Cao. Tissue and mechanical heart valves. Ann. Cardiothorac. Surg. 4:399, 2015. 13. Hofferberth, S. C., M. Y. Saeed, L. Tomholt, M. C. Fernandes, C. J. Payne, K. Price, G. R. Marx, J. J. Esch, D. W. Brown, J. Brown, P. E. Hammer, R. W. Bianco, J. C. Weaver, E. R. Edelman, and P. J. del Nido. A geometrically adaptable heart valve replacement. Sci. Transl. Med. 12:, 2020. 14. IJsselhof, R. J., M. G. Slieker, M. G. Hazekamp, R. Accord, H. van Wetten, F. Haas, and P. H. Schoof. Mitral valve replacement with the 15-mm mechanical valve: A 20-year multicenter experience. Ann. Thorac. Surg. 110:956\u2013961, 2020. 15. Jiang, Y., S. Wang, J. Bian, S. Chen, and Y. Shao. Mechanical versus bioprosthetic", "E. Hammer, R. W. Bianco, J. C. Weaver, E. R. Edelman, and P. J. del Nido. A geometrically adaptable heart valve replacement. Sci. Transl. Med. 12:, 2020. 14. IJsselhof, R. J., M. G. Slieker, M. G. Hazekamp, R. Accord, H. van Wetten, F. Haas, and P. H. Schoof. Mitral valve replacement with the 15-mm mechanical valve: A 20-year multicenter experience. Ann. Thorac. Surg. 110:956\u2013961, 2020. 15. Jiang, Y., S. Wang, J. Bian, S. Chen, and Y. Shao. Mechanical versus bioprosthetic aortic valve replacement in middle-aged adults: A systematic review and meta-analysis. J. Cardiovasc. Dev. Dis. 10:90, 2023. 16. Kim, J. Y., J. B. Kim, S.-H. Jung, S. J. Choo, C. H. Chung, and J. W. Lee. Long-term outcomes of homografts in the aortic valve and root position: A 20-year experience. Korean J. Thorac. Cardiovasc. Surg. 49:258\u2013263, 2016. 41 17. King, M. J., T. David, and J. Fisher. Three-dimensional study of the effect of two leaflet opening angles on the time-dependent flow through a bileaflet mechanical heart valve. Med. Eng. Phys. 19:235\u2013241, 1997. 18. Klusak, E., A. Bellofiore, S. Loughnane, and N. J. Quinlan. High-resolution measurements of velocity and shear stress in leakage jets from bileaflet mechanical heart valve hinge models. J. Biomech. Eng. 137:111008, 2015. 19. Kumar, T., A. Singh, S. Thakre, S. Acharya, S. Shukla, and S. Kumar. Scientific evolution of artificial heart valves: A narrative review. Cureus 15:, 2023. 20. Mashali, M. H., A. F. Elmahrouk, Z. Ahmad, O. Abdulrahman, A. F. Galleon, A. Al-Kouatli, A. A. Jamjoom, and R. A. Zahr. Mechanical mitral valve endurance in children under 2 years. Cardiothorac. Surg. 32:, 2024. 21. Mills, M., M. John, R. Tang, M. P. Fundora, R. Keesari, K. Kanter, K. Maher, and P. Chai. Mitral valve replacement in infants and children: Experience using a 15-mm mechanical valve. Ann. Thorac. Surg. 116:322\u2013329, 2023. 22. Min Yun, B., C. K. Aidun, and A. P. Yoganathan. Blood damage through a bileaflet mechanical heart valve: a quantitative computational study using a multiscale suspension flow solver. J. Biomech. Eng. 136:101009, 2014. 23. Notenboom, M. L., A. Schuermans, J. R. G. Etnel, K. M. Veen, P. C. van de Woestijne, F. R. Rega, W. A. Helbing, A. J. J. C. Bogers, and J. J. M. Takkenberg. Paediatric aortic valve replacement: a meta-analysis and microsimulation study. Eur. Heart J. 44:3231\u20133246, 2023. 24. Pibarot, P., and J. G. Dumesnil. Prosthetic heart valves. Circulation 119:1034\u20131048, 2009. 42", "mechanical heart valve: a quantitative computational study using a multiscale suspension flow solver. J. Biomech. Eng. 136:101009, 2014. 23. Notenboom, M. L., A. Schuermans, J. R. G. Etnel, K. M. Veen, P. C. van de Woestijne, F. R. Rega, W. A. Helbing, A. J. J. C. Bogers, and J. J. M. Takkenberg. Paediatric aortic valve replacement: a meta-analysis and microsimulation study. Eur. Heart J. 44:3231\u20133246, 2023. 24. Pibarot, P., and J. G. Dumesnil. Prosthetic heart valves. Circulation 119:1034\u20131048, 2009. 42 25. Sharabiani, M. T. A., D. M. Dorobantu, A. S. Mahani, M. Turner, A. J. Peter Tometzki, G. D. Angelini, A. J. Parry, M. Caputo, and S. C. Stoica. Aortic valve replacement and the Ross operation in children and young adults. J. Am. Coll. Cardiol. 67:2858\u20132870, 2016. 26. de Simone, G., R. B. Devereux, S. R. Daniels, G. Mureddu, M. J. Roman, T. R. Kimball, R. Greco, S. Witt, and F. Contaldo. Stroke volume and cardiac output in normotensive children and adults. Circulation 95:1837\u20131843, 1997. 27. Velho, T. R., R. M. Pereira, F. Fernandes, N. C. Guerra, R. Ferreira, and \u00c2. Nobre. Bioprosthetic aortic valve degeneration: A review from a basic science perspective. Rev. Bras. Cir. Cardiovasc. 37:239, 2022. 28. Voges, I., M. Jerosch-Herold, J. Hedderich, E. Pardun, C. Hart, D. D. Gabbert, J. H. Hansen, C. Petko, H.-H. Kramer, and C. Rickers. Normal values of aortic dimensions, distensibility, and pulse wave velocity in children and young adults: a cross-sectional study. J. Cardiovasc. Magn. Reson. 14:77, 2012. 29. Yoganathan, A. P., M. Fogel, S. Gamble, M. Morton, P. Schmidt, J. Secunda, S. Vidmar, and P. del Nido. A new paradigm for obtaining marketing approval for pediatric-sized prosthetic heart valves. J. Thorac. Cardiovasc. Surg. 146:879\u2013886, 2013. 30. Yun, B. M., D. B. McElhinney, S. Arjunon, L. Mirabella, C. K. Aidun, and A. P. Yoganathan. Computational simulations of flow dynamics and blood damage through a bileaflet mechanical heart valve scaled to pediatric size and flow. J. Biomech. 47:3169\u20133177, 2014. 43", "Characterizing the Detection Limits and Sensitivity of Paper Based Microfluidic Devices for Chronic Kidney Disease BS. Biomedical Engineering Program Biomedical Engineering Charles W. Davidson College of Engineering March 13, 2025 Aasiya Jabbar, Praagna Doddaballapur, Janane Sivakumar Advisor: Alessandro Bellofiore, Ph.D Faculty Advisor San Jose State University Associate Professor and Chair Abstract Chronic Kidney Disease is one of the leading causes of death in the United States [1] . In fact, it was estimated that 35.5 million people in the United States are diagnosed with the condition [2] . Chronic Kidney disease is defined as the progressive degradation of the kidneys and their function [3] . Currently, there is no known cure for the condition. As a result, treatment for the condition often involves the monitoring and regulating of the progression of the condition and the state of the Kidney. Regular monitoring of the condition is especially important for CKD patients as it can lead to early detection, which leads to earlier intervention, resulting in better outcomes for patients. However, current monitoring methods are limited in nature. They often require extensive laboratory procedures and equipment, which are only available in hospital and clinical settings, ultimately limiting patients in their ability to regulate their condition. To address this, a paper-based diagnostic device (\u03bc PAD), employing colorimetric detection, was developed to provide CKD patients with an easy-to-understand means of monitoring their condition, at their convenience. The device achieves this by measuring serum creatinine in the patient\u2019s blood. Serum creatinine is a common byproduct of muscle metabolism and is regularly filtered out by the kidneys. A large concentration of creatinine in the blood serves as an indicator of poor kidney function. Creatinine is measured by the device using the Jaffe\u2019s reaction, which yields a red-color change in the presence of creatinine in the sample. In this study, paper-based microfluidic devices (\u03bc PAD) were developed, and the detection limits and sensitivity of these devices was assessed across both phosphate buffer saline solutions and porcine blood. The performance of the device was evaluated across all ranges of creatinine concentrations: Low, Medium and High, in both mediums. The sensitivity of the device was determined across all ranges. The performance of the device was evaluated by the degree of color change for each respective concentration of creatinine solution applied and the degree of color change was quantified using the RGB values in Python. Through this analysis, the performance", "assessed across both phosphate buffer saline solutions and porcine blood. The performance of the device was evaluated across all ranges of creatinine concentrations: Low, Medium and High, in both mediums. The sensitivity of the device was determined across all ranges. The performance of the device was evaluated by the degree of color change for each respective concentration of creatinine solution applied and the degree of color change was quantified using the RGB values in Python. Through this analysis, the performance of the device in typical in-use conditions can be determined, which can indicate the effectiveness of this device as a diagnostic tool for kidney disease monitoring. Keywords: Paper-based Microfluidic Devices, Chronic Kidney Disease, Colorimetric Detection, Jaffe Reaction, Serum Creatinine, Porcine Blood, RGB, Python 1 Table of Contents Executive Summary ......................................................................................................................... 4 Introduction ...................................................................................................................................... 5 Literature Review: ........................................................................................................................... 7 Proposed Solution: ........................................................................................................................... 9 Statement of Need: ........................................................................................................................... 9 Materials and Methods: .................................................................................................................. 10 Materials: ................................................................................................................................. 10 Methods: .................................................................................................................................. 12 1. Design of \u03bc PAD Device ................................................................................................. 12 2. Fabrication and Assembly of \u03bc PAD Device .................................................................. 13 3. Jaffe Reaction Reagent Preparation ............................................................................... 13 4. Preparation of the Calibration Curve for the Porcine Blood Samples ........................... 14 5. Determination of Creatinine levels in Porcine Blood Samples ..................................... 14 6. Testing Device to Determine Sensitivity, and Detection Limits .................................... 15 Results and Discussion: ................................................................................................................. 16 Saline Creatinine Solution Testing of \u03bc PAD: .......................................................................... 16 1. Sensitivity of \u03bc PAD Device using Saline Creatinine Solutions: ................................... 16 2. Detection Limits of \u03bc PAD Device using Saline Creatinine Solutions: .......................... 19 a. Lower Detection Limit Analysis: ............................................................................. 19 b. Upper Detection Limit Analysis: ............................................................................. 25 Porcine Blood Testing of \u03bc PAD ............................................................................................... 28 3. Calibration Curve for Determination of Creatinine Concentration in Porcine Blood ... 29 4. Sensitivity of the \u03bc PAD Device using Porcine Blood Samples ..................................... 31 5. Detection Limits of \u03bc PAD Device using Porcine Blood Samples ................................. 35 a. Lower Detection Limit Analysis: ............................................................................. 35 b. Upper Detection Limit Analysis .............................................................................. 36 Sources of Error: ...................................................................................................................... 40 Conclusion: .................................................................................................................................... 44 Future Work ................................................................................................................................... 45 Cost Analysis ................................................................................................................................. 45 Safety: ............................................................................................................................................ 46 FDA Considerations: ..................................................................................................................... 47 Acknowledgements: ....................................................................................................................... 49 2 References ...................................................................................................................................... 50 Appendix A: Python Script for Extracting RGB Values from Test Images ................................... 53 Appendix B: Calculations for the Amount of Creatinine needed for Lower Detection Limit Testing of \u03bc PAD Device for Porcine Blood ................................................................................... 54 Appendix C:", "Detection Limit Analysis: ............................................................................. 35 b. Upper Detection Limit Analysis .............................................................................. 36 Sources of Error: ...................................................................................................................... 40 Conclusion: .................................................................................................................................... 44 Future Work ................................................................................................................................... 45 Cost Analysis ................................................................................................................................. 45 Safety: ............................................................................................................................................ 46 FDA Considerations: ..................................................................................................................... 47 Acknowledgements: ....................................................................................................................... 49 2 References ...................................................................................................................................... 50 Appendix A: Python Script for Extracting RGB Values from Test Images ................................... 53 Appendix B: Calculations for the Amount of Creatinine needed for Lower Detection Limit Testing of \u03bc PAD Device for Porcine Blood ................................................................................... 54 Appendix C: Standard Operating Procedure for Picric Acid ......................................................... 55 3 Executive Summary This paper presents the development of a microfluidic paper based analytical device, designed to help those with chronic kidney disease track and monitor creatinine levels in the body to ultimately detect early signs or prevent the onset of kidney disease. A key challenge in point of care diagnostics was the detection efficiency as well as the device being a smaller, less convenient size to work with. Therefore, we decided to modify the device by making the reaction pads slightly bigger and the hydrophobic portion slightly thicker to ensure no sample is lost. By doing this, we were able to work with a wider surface to observe our findings. Additionally, Our findings demonstrate that the enhanced \u03bc PAD provides more reliable detection and sensitivity while maintaining the affordability and simplicity of conventional paper based-platforms. With these improvements, users will be able to obtain a more accurate description and value of their creatinine levels, which will strongly influence the future decisions they make to improve their overall health and wellbeing. 4 Introduction Chronic Kidney Disease (CKD) is a global public health concern that affects thousands of people across the world. Based on the CDC, 14% of U.S adults are affected by CKD, and 9 in 10 adults with CKD don\u2019t know they have it. 1 in 3 adults with severe CKD are unaware they are affected by CKD. This raises a critical and prevalent gap present through all these numbers. The diagnosis of CKD, which is a disease that has the potential to permanently harm an individual, isn\u2019t simple and an easy process for most people to perform or to have someone perform for them. Tests to diagnose CKD often are conducted in hospital settings, where waitlists for appointments with a medical professional can take months, resulting in patients not asking for diagnosis tests or not taking an appointment at all.", "critical and prevalent gap present through all these numbers. The diagnosis of CKD, which is a disease that has the potential to permanently harm an individual, isn\u2019t simple and an easy process for most people to perform or to have someone perform for them. Tests to diagnose CKD often are conducted in hospital settings, where waitlists for appointments with a medical professional can take months, resulting in patients not asking for diagnosis tests or not taking an appointment at all. The \u03bc PAD is an inexpensive and compact device that can cover this void in this system by providing all patients, especially CKD diagnosed patients, opportunities to check their own kidney health in their homes and at their convenience. The \u03bc PAD is a simple-to-use and reliable method for one to check their risk for CKD or monitor their health. The \u03bc PAD design consists of two squares: an input square for test sample insertion and an output square for a Jaffe\u2019s reaction and colorimetric reaction to occur. In order to use this device, patients would simply place a drop of their blood in the input square of the \u03bc PAD and wait for the \u03bc PAD to change color. Then, using an app, they would take a picture of the color change to identify the creatinine concentration in their blood. This provides the patient with information regarding whether they\u2019re at risk of being diagnosed with CKD or their kidney\u2019s condition, in the case they are already diagnosed with CKD. The patients can do this at home on their own without the need of direct professional medical assistance, making it a helpful tool for patients to confirm their health in their own time. The creation of the \u03bc PAD uses mostly versions of paper materials that undergo extensive treatments for sample testing. The \u03bc PAD is created using three main layers: the filter paper layer on top, the plasma separation membrane layer in the middle, and a plastic backing in the bottom layer. The \u03bc PAD\u2019s filter paper layer includes thermal printer ink, which provides the inked regions of the paper with hydrophobicity. The \u03bc PAD\u2019s filter paper layer is later treated with picric acid and an alkaline medium in the output square. The plasma separation membrane layer of the device allows the separation of red blood cells and fibrinogen from the whole blood test sample. The bottom plastic backing", "membrane layer in the middle, and a plastic backing in the bottom layer. The \u03bc PAD\u2019s filter paper layer includes thermal printer ink, which provides the inked regions of the paper with hydrophobicity. The \u03bc PAD\u2019s filter paper layer is later treated with picric acid and an alkaline medium in the output square. The plasma separation membrane layer of the device allows the separation of red blood cells and fibrinogen from the whole blood test sample. The bottom plastic backing layer provides the device with a stiff layer to support the above layers of the device. After each of these layers have been cut and treated as needed, the \u03bc PAD is ready to be used to test samples for creatinine concentrations. Typically, the blood sample would be placed in the input square, resulting in a capillary action of the blood fluid moving toward the output square of the device, where the picric acid and alkaline treated region lies. When the test sample reaches the output square, over the span of approximately five minutes, the Jaffe\u2019s reaction starts to occur. This colorimetric reaction displays a yellow-orange color change in the output square of the device, displaying the creatinine concentration present in 5 the sample through the color intensity and the type of color present. After analyzing the image of the color change, the creatinine concentration in the test sample can be identified. This provides the patient with mind relieving information about the condition of their kidney. Overall, the development of the \u03bc PAD device in this manner serves a small role in a larger trend towards making healthcare reliable and convenient with in-home tests. The usage of the \u03bc PAD empowers people to have a chance to care for their health and wellbeing better. 6 Literature Review: In recent years, there has been a rise in the prevalence of Chronic Kidney Disease (CKD), worldwide. It is estimated that over 697 million people globally have a reduced glomerular filtration rate, which is characteristic of people with the disease [14] . In plain terms, Chronic Kidney Disease (CKD) is a condition whereby the kidney sustains damage that impairs proper kidney function [15] . One key bioindicator of this disease is creatinine. Creatinine is a metabolic waste product that is typically filtered out by the kidneys and excreted in the urine [18] . High levels of creatinine indicate an impairment of Kidney function", "million people globally have a reduced glomerular filtration rate, which is characteristic of people with the disease [14] . In plain terms, Chronic Kidney Disease (CKD) is a condition whereby the kidney sustains damage that impairs proper kidney function [15] . One key bioindicator of this disease is creatinine. Creatinine is a metabolic waste product that is typically filtered out by the kidneys and excreted in the urine [18] . High levels of creatinine indicate an impairment of Kidney function 3 . Currently, CKD is primarily monitored through the level of creatinine in the blood or urine, although other bodily fluids such as sweat, saliva, cerebral spinal fluid, and tears can also be used [17] . Blood is the most ideal out of all the options as it provides the most clinically relevant data [17] . For a healthy kidney, blood creatinine levels should be 0.5-1.0 mg/dL for women and 0.7-1.2 mg/dL for men [18] . However, this range varies depending on the muscle mass of the individual, that is, individuals with higher muscle mass tend to have higher creatinine levels [18] . The most common method used to measure creatinine concentration is the Jaff\u00e9 assay [18] . However, due to the poor specificity of the Jaff\u00e9 assay to other biomolecules, this method is often replaced by enzymatic reactions [17] . In addition to this, other techniques of measurement include chromatography, pH sensors, spectrophotometry, and more [18] . Due to the complexity and ornateness of the measurement processes, creatinine levels must be measured by a specialist and in a laboratory setting as the process requires complex machinery, and has a slow experimental time [18] . Enzymatic reactions are also not ideal as they have a high cost associated with them, along with large instrumentation and reagent volume [17] . This complicates and hinders the patient\u2019s ability to monitor the condition. Consequently, as the level of CKD cases rises, there is an increasing need for the development of new methods of monitoring the condition that will allow the affected patients to monitor it independently and at their own leisure. Colorimetric detection is one of the most used techniques for \u00b5PAD. This is due to its ability to provide quick results, its simple instruction of use, stability, and convenience. This detection can be seen by the movement of the analyte solution, such as the toxic ions or creatinine in this case, where", "an increasing need for the development of new methods of monitoring the condition that will allow the affected patients to monitor it independently and at their own leisure. Colorimetric detection is one of the most used techniques for \u00b5PAD. This is due to its ability to provide quick results, its simple instruction of use, stability, and convenience. This detection can be seen by the movement of the analyte solution, such as the toxic ions or creatinine in this case, where it reacts to the specific reagents to produce a color change to visually determine the creatinine levels. This determination can be done by scanners, phone cameras, other cameras, then transferred to a PC or phone for analysis [14] . As one method may either not be enough to conclude the test, or be precise enough, other methods are available. One other method that is quite useful is fluorescent detection, where the reaction is between the dye and targeted molecules. This helps us also visually observe the color change to detect the amount of creatinine in the sample. Other methods of detection include electrochemical detection, chemiluminescence detection, electrochemiluminescence detection, 7 nanoparticle based sensor detection, and spectrometry detection [14] . Although all these methods have a different procedure, all have the same purpose to detect the amount of creatinine to ultimately figure out how to reduce the amount in the sample, which benefits those who need monitoring and/or reduction of creatinine levels. Jaffe\u2019s reaction became common in the 19th century. However, it was said that the chemical analysis was nonspecific, so other methods appeared to improve the specificity of this analysis. Adsorption to Fuller\u2019s earth, ion-exchange chromatography, liquid chromotography, fragmentography, enzymatic hydrolysis, and kinetic analysis along with different alternatives such as the substitution of picrate have been suggested over time. Jaffe\u2019s reaction consists of a method used to visually detect creatinine, allowing the waste product to appear as an orange color. This detection of color has to do with the utilization of creatinine with picric acid in an alkaline medium, an aqueous solution with a pH of higher than 7. In this case, 1% of picric acid is used with 0.75 N NaOH in protein free filtrate (PFF). After 15 minutes to allow the right shade to fully develop, it is measured at 520 nm via photometer. There were three experiments done to successfully detect the amount of creatinine. The purpose of", "detection of color has to do with the utilization of creatinine with picric acid in an alkaline medium, an aqueous solution with a pH of higher than 7. In this case, 1% of picric acid is used with 0.75 N NaOH in protein free filtrate (PFF). After 15 minutes to allow the right shade to fully develop, it is measured at 520 nm via photometer. There were three experiments done to successfully detect the amount of creatinine. The purpose of the first experiment was to determine the ideal concentration of NaOH needed for maximum color development in the standard creatinine stock. The second experiment was meant to determine the amount of NaOH concentration needed to detect maximum color development for measurements of creatinine in urine. The third experiment was done to determine the concentration of NaOH needed for maximum color development of creatinine in serum [18] . As there are numerous methods to detect creatinine levels in samples, it is crucial to have a set of safety precautions and measures that are advisable to assist the process of gaining precise and accurate results. This can mean the usage of materials and equipment that can provide accurate results without unexpected explosion or harm during experimentation. For example, during the quantification of creatinine in a sample, the usage of filter paper would be ineffective as the material doesn't provide clear observable color-changing results. For better viewing of color detection, modified filter paper is considered a better option [19] . Although color detection can be measured through digital imaging and other means, this can provide the person conducting the experiment visual conviction of any numerical data collected. If creatinine detection is performed using 3D or regular \u00b5PADs, there are more specific precautions necessary for safe data collection. Although there are countless methods of creatinine detection, \u00b5pads are considered relatively effective due to their low cost, ease of use, and disposable qualities [20] . In order to obtain a high color intensity on the \u00b5pads during and after Jaffe\u2019s reaction for results, it would be advisable to have the reaction run at 37 C, regardless of creatinine concentration [18] . With a high color intensity for data collection on the \u00b5PAD, the results can be clearly collected by either being visible to the naked eye or processed through a digital camera. 8 Methods that require rigorous safety measures taken during data collection from biofluidic", "In order to obtain a high color intensity on the \u00b5pads during and after Jaffe\u2019s reaction for results, it would be advisable to have the reaction run at 37 C, regardless of creatinine concentration [18] . With a high color intensity for data collection on the \u00b5PAD, the results can be clearly collected by either being visible to the naked eye or processed through a digital camera. 8 Methods that require rigorous safety measures taken during data collection from biofluidic experiments can include lab-on-chip devices. These devices are compact and implantable if needed for point of care treatments. Lab-on-chip devices can be easily sterilized through ultraviolet radiation and autoclaving and heat [22] . For usage of this device, it includes several components such as heat exchangers for temperature control, electrical or mechanical micropumps for microfluidic changes, mixers, and other compact equipment for microfluidic manipulations [22] . Temperature control is especially important as it can affect the processes of ongoing reactions in the lab-on-chip device, such as evaporation or heating of compact parts in the device [22] . By taking account of these safety measurements, data can be collected efficiently and effectively. The goal of this work is to obtain the sample to then analyze the levels of creatinine via the numerous methods available, and of course allow the patients to know the amount of creatinine in their body for their own benefit. Working with others and using the necessary methods to determine and analyze creatinine levels consists of some risks. To reduce those risks, it is imperative that the right protocols are taken to prevent any complications in the future that may hinder the service of detection. Proposed Solution: The aim of this study is to manufacture and test the effectiveness of paper-based microfluidic devices (\u03bc PAD) in detecting blood serum creatinine. The device was fabricated by printing the hydrophobic barriers on the filter paper using the Itari Portable Thermal Printer, cutting Plasma membrane paper in the desired dimensions, and adhering the two layers to a PVC plastic card backing. The effectiveness of the device was evaluated by testing the sensitivity and detection limits of the color change in the device against various saline creatinine solutions and blood samples of known creatinine concentrations. To simplify the work of future researchers, a method of determining the concentration of creatinine in porcine blood samples, which is essential for evaluating the performance of", "Printer, cutting Plasma membrane paper in the desired dimensions, and adhering the two layers to a PVC plastic card backing. The effectiveness of the device was evaluated by testing the sensitivity and detection limits of the color change in the device against various saline creatinine solutions and blood samples of known creatinine concentrations. To simplify the work of future researchers, a method of determining the concentration of creatinine in porcine blood samples, which is essential for evaluating the performance of the device, has been developed. Creatinine levels in the porcine blood samples was determined by creating a standard curve of the expected creatinine levels in a healthy porcine species and measuring the absorbance of the color change in these solutions by spectrophotometry. This plot can be used as a reference in future research for determining the concentration of creatinine in various porcine blood samples. Additionally, since sheep blood better resembles human blood, the findings and methods from this project can be used to obtain data which can potentially bring this prototype closer to being a functional medical device for human use. Statement of Need: Recently, there is a growing need for low-cost, easy-to-use POC diagnostic devices for Chronic Kidney Disease monitoring due to the limited accessibility of current monitoring 9 techniques. In general, the monitoring of Kidney function is critical to the management and treatment of CKD. However, existing monitoring systems consists primarily of in-hospital laboratory testing such as Urine-Albumin Creatinine Ratio (uACR), Blood Urea Nitrogen (BUN), and Estimated Glomerular Filtration Rate (eGFR) [29] . These tests are typically done by medical professionals, and require specific assays. As a result, patients become dependent on hospitals and medical professionals for proper disease treatment and management. This reduces patient autonomy in the treatment of the condition and creates extra burden for patients as they need to make frequent hospital visits. Paper based Microfluidic Devices (uPADs) offer an effective alternative to this issue by offering patients a means of monitoring their creatinine levels within the comfort of their home. Through this, patients develop greater autonomy in the treatment of the condition and are able to make better life-style choices as a result. Its ability to provide immediate information regarding their blood creatinine levels, through an easy-to-understand interface allows patients to recognize potential health risks early, allowing them to take preventative action. Moreover, its cheap manufacturing costs ensures accessibility to patients of all demographics.", "by offering patients a means of monitoring their creatinine levels within the comfort of their home. Through this, patients develop greater autonomy in the treatment of the condition and are able to make better life-style choices as a result. Its ability to provide immediate information regarding their blood creatinine levels, through an easy-to-understand interface allows patients to recognize potential health risks early, allowing them to take preventative action. Moreover, its cheap manufacturing costs ensures accessibility to patients of all demographics. Therefore, the development of a low-cost, effective Point of Care (POC) device for detecting creatinine can greatly improve the quality of care for Chronic Kidney Disease patients. Materials and Methods: Materials : Materials Purpose 8 in x 11 in Filter Paper Sheets To provide a hydrophobic layer to concentrate fluid to one region PALL Vivid Plasma Separation Membrane Paper Separation of Plasma content from whole blood sample PVC Plastic Cards To provide a sturdy platform for the microfluidic device Scissors Size and adjust dimensions for microfluidic layers by cutting as needed Exacto Knife To provide precision in cutting materials Double Sided Scotch Tape To attach two layers of microfluidic device; especially for the filter paper layer to the plasma separation membrane paper layer 10 Itari Thermal Printer To provide hydrophobicity to regions of the filter paper Centrifuge Machine The spin to the whole blood for plasma layer separation from hemoglobin P20 and P200 Micropipette/Micropipette Tips For hydrophobicity testing with deionized water 10 mL Serological Pipettes To create blood sample solutions and saline creatinine solutions of various creatinine concentrations. 14 mL plastic test tubes lids To store saline creatinine solution and blood samples with varying concentrations of creatinine in each tube in the refrigerator. 50mL plastic test tubes with lids These tubes are used for centrifuging purposes and/or storing blood samples in the refrigerator 2 100mL beakers This is used for measuring blood, phosphate buffer saline, plasma, or other creatinine-related solutions 1.3% Picric Acid Stock Solutions Used for pretreating the reaction square of the microfluidic device after further dilution Dry, Solid NaOH Crystals Used in dissolved form for pretreating the reaction square of the microfluidic device DI water Used for picric acid and NaOH dilutions and for blank value data testing Solid Creatinine Used for creating solutions with varying creatinine concentrations in blood and in saline solutions Phosphate Buffer Solutions Used for creating dilutions of saline creatinine solutions 100mL of Porcine", "1.3% Picric Acid Stock Solutions Used for pretreating the reaction square of the microfluidic device after further dilution Dry, Solid NaOH Crystals Used in dissolved form for pretreating the reaction square of the microfluidic device DI water Used for picric acid and NaOH dilutions and for blank value data testing Solid Creatinine Used for creating solutions with varying creatinine concentrations in blood and in saline solutions Phosphate Buffer Solutions Used for creating dilutions of saline creatinine solutions 100mL of Porcine Blood Used for plasma extraction for data collection and as test sample on the microfluidic device 11 PPE (gloves, safety goggles, lab coats) Worn for body and face protection as any work is done in a lab setting Fume Hood Operated for better air ventilation during safe usage of explosive and toxic chemicals, such as picric acid Light box/Ring light stand To provide better environment for images captured during reaction, especially in reaction region of the microfluidic device iPhone 13 Camera Tool used to capture images of the microfluidic device for data collection Methods: 1. Design of \u03bc PAD Device The \u03bc PAD device used in this study was designed in three components: Filter paper layer, Membrane paper layer, and plastic backing. The filter paper layer was made using Grade 1, 11uM, 8 in x 11 in rectangular filter paper. Its design consists of a 2.56 cm by 1.45 cm rectangular square with two 0.7 cm by 5.5 cm squares at the center of the device. The layer serves to identify the sample input region (square on the left), where the sample is added, and the reaction zone (square on the right) where the color change is presented. It also serves to separate and control phases of the reaction, and remove red blood cells from plasma in the input sample. The reaction zone is treated with the Jaffe Reaction reagents and assumes a bright yellow color once the region has been treated. The input and reaction zones of the device are separated by a black, hydrophobic barrier that serves to prevent the contamination of the reagents in these regions. The membrane paper layer consists of a 1.05 cm by 0.5 cm rectangular strip of PALL Vivid Membrane Separation paper. This strip was placed directly underneath the two squares in the filter paper, to guide the flow of the liquid substrate from the input zone to the reaction zone of the", "been treated. The input and reaction zones of the device are separated by a black, hydrophobic barrier that serves to prevent the contamination of the reagents in these regions. The membrane paper layer consists of a 1.05 cm by 0.5 cm rectangular strip of PALL Vivid Membrane Separation paper. This strip was placed directly underneath the two squares in the filter paper, to guide the flow of the liquid substrate from the input zone to the reaction zone of the filter paper via capillary action. A plastic backing is placed underneath the membrane paper to prevent fluid from seeping out from under the membrane paper and to add sturdiness to the device. Previous designs of the \u03bc PAD device utilized smaller dimensions for the device overall to maximize the number of devices that can be printed. In this model, the device dimensions were increased to provide a greater surface area for applying the adhesive that is used to assemble the device and simplifying the device assembly process. The 12 device was designed to have two distinct input and reaction zones to simplify the user interface and make it easier for the user to use the device. 2. Fabrication and Assembly of \u03bc PAD Device The fabrication of the \u03bc PAD device consists of 3 components, this being the filter paper, membrane paper and the plastic backing. The filter paper layer was fabricated by first printing the filter paper layer design on a 8 in x 11 in filter paper sheet, using the Itari Thermal Printer. To conserve paper and reduce waste, multiple devices were printed on the same sheet. Once printed, the sheet was baked for 30 minutes at 100 \u2103 to ensure the printed area is completely hydrophobic. Each device was then cut out using an Exacto knife. Next, the reaction zone of each device sample was treated with a 1:1 mixture of 5uL of 2M NaOH and 5uL of 0.4M Picric Acid solution. This was done in a fume hood and the devices were then left to dry overnight in the fume hood. Once dry, the reaction zone of the device assumes a bright yellow color. The membrane paper layer and plastic backing did not undergo any form of pretreatment as it was cut to the desired shape and did not incorporate any printed or treated regions. In general, the assembly of the device is as follows: the", "of 0.4M Picric Acid solution. This was done in a fume hood and the devices were then left to dry overnight in the fume hood. Once dry, the reaction zone of the device assumes a bright yellow color. The membrane paper layer and plastic backing did not undergo any form of pretreatment as it was cut to the desired shape and did not incorporate any printed or treated regions. In general, the assembly of the device is as follows: the filter paper filter paper is placed on top of the membrane paper strip, which is placed on top of a plastic card backing. To stick the layers together, thin strips of double sided tape first are cut to the dimensions of the hydrophobic barrier size and attached to the hydrophobic regions of the filter paper layer. The double sided tape strips are cut thinly to ensure that it fits on all four sides of the device borders without interfering with the process of detecting creatinine by reacting with the test solution. Considering that this phase is the most time consuming, alternative adhesion methods are currently being explored to improve the efficiency and speed of the assembly process. Next, the membrane paper layer strips are attached to the filter paper. The entirety is then attached to a PVC plastic card backing using thinly cut double sided tape. To reduce waste, nine test device samples were attached to a single plastic backing piece. 3. Jaffe Reaction Reagent Preparation The 10uL mixture of 2M NaOH and 0.4M Picric Acid on the reaction pad is what facilitates the Jaffe\u2019s reaction and allows for the color change to occur. To prepare the 2M NaOH solution, approximately 7.9 g of NaOH crystals was weighed and added to a 250mL beaker containing 100mL DI water. Due to the heat releasing nature of the reaction, the solution inside the 250mL beaker increased in temperature slightly. The solution was continuously stirred until the solution became clear and the solid dissolved. The solution cooled down after 5-7 minutes. 13 To prepare the 0.4M Picric Acid solution, 70.5mL of 1.3% Picric Acid Stock Solution was added to a 250mL beaker with 29.5mL of DI water. Due to the volatile nature of concentrated picric acid, the procedure was done under a fume hood and all necessary safety measures were taken. The saline creatinine solutions used for testing were prepared by measuring out", "continuously stirred until the solution became clear and the solid dissolved. The solution cooled down after 5-7 minutes. 13 To prepare the 0.4M Picric Acid solution, 70.5mL of 1.3% Picric Acid Stock Solution was added to a 250mL beaker with 29.5mL of DI water. Due to the volatile nature of concentrated picric acid, the procedure was done under a fume hood and all necessary safety measures were taken. The saline creatinine solutions used for testing were prepared by measuring out various amounts of creatinine and adding them to phosphate buffer solution. To prepare a 0.1 mg/dL saline creatinine solution, for example, 0.1 mg of solid creatinine was added to 100mL of phosphate buffer solution. This process was repeated for all creatinine solutions used in testing. 4. Preparation of the Calibration Curve for the Porcine Blood Samples A calibration curve was made to determine the creatinine concentration in heparinized porcine blood samples. This measure was taken in order to properly gauge the effectiveness of the \u03bc PAD device in determining creatinine levels in blood samples. In order to determine the concentration of creatinine in blood samples, the normal range of serum creatinine in porcine blood samples was first determined. Based on literature, it was found that the serum creatinine levels of a healthy porcine species ranges from 0.6-1.6 mg/dL [4] . Using this range, nine 30 mL standard saline creatinine solutions with concentrations ranging from 0.1mg/dL to 1.6mg/dL were prepared, with concentration increments of 0.1mg/dL. A wider range was used as a buffer, in the case that the measured concentrations are lower than expected. Next, 10mL of a 50:50 solution of 0.4M Picric Acid and 2M NaOH solution was added to each saline creatinine solution. This was done to induce a quantifiable color change in the standard solutions. A 1:3 Picric Acid/NaOH to creatinine ratio was used to match that of the solutions used in the \u03bc PAD. After combining the saline creatinine solutions with Picric Acid and NaOH solutions, the solutions were allowed to sit for 5 minutes, to reach maximum color intensity. The absorbance of each of these samples were then measured using the spectrophotometer present in ENGR 233 at a wavelength of 520nm. Three replicate measurements were made for each concentration to obtain more accurate measurements. These replicate measurements were then averaged and the resulting absorbance data was then plotted on Microsoft Excel, as a function of concentration,", "saline creatinine solutions with Picric Acid and NaOH solutions, the solutions were allowed to sit for 5 minutes, to reach maximum color intensity. The absorbance of each of these samples were then measured using the spectrophotometer present in ENGR 233 at a wavelength of 520nm. Three replicate measurements were made for each concentration to obtain more accurate measurements. These replicate measurements were then averaged and the resulting absorbance data was then plotted on Microsoft Excel, as a function of concentration, to obtain the calibration curve. 5. Determination of Creatinine levels in Porcine Blood Samples Prior to using Porcine Blood in the testing of the \u03bc PAD device, the creatinine levels in the sample need to be determined. To achieve this, the blood sample was first centrifuged and the blood plasma was extracted. Picric Acid and NaOH was then added to the blood plasma to induce a color change in the translucent plasma solution. The mixture was allowed to sit at room temperature for 5 minutes to obtain the maximum color change. The absorbance of the mixture was then measured at a wavelength of 520 14 nm. Three replicates absorbance measurements were made to obtain a more accurate absorbance measurement. The average absorbance of the sample was then compared with the calibration curve developed previously to determine the approximate estimate of the concentration of creatinine in the sample. 6. Testing Device to Determine Sensitivity, and Detection Limits The \u03bc PAD contains two pads, one being the insertion pad and the other being the reaction pad, which is cured with the alkaline picrate solution. Testing the device consists of adding the solution, initially, saline creatinine solution, then later on blood, to the insertion pad and photographing the color change on the reaction pad of the device at specified intervals. To prepare the devices for testing, multiple sets of samples were taped on plastic backing cards. Each plastic backing card yielded 9 samples. 6 cards, each containing 9 samples were taped onto the paper, labeled with the intended concentration, then placed into a lightbox and captured photos, at different time points. An iPhone 13 was then placed in a fixed position and above the light box, at a magnification of 4.1x. Once the setup was complete, 30uL of test solution was pipetted onto the reaction zone of the device. Images were then taken of the device before application, at 0 minutes, 3 minutes,", "6 cards, each containing 9 samples were taped onto the paper, labeled with the intended concentration, then placed into a lightbox and captured photos, at different time points. An iPhone 13 was then placed in a fixed position and above the light box, at a magnification of 4.1x. Once the setup was complete, 30uL of test solution was pipetted onto the reaction zone of the device. Images were then taken of the device before application, at 0 minutes, 3 minutes, 6, minutes, and 9 minutes after the solution was added to the device, inside the lightbox. The images taken before applying the solution were used to determine the initial degree of color change that occurred in the device. The light box allowed for better consistency in terms of the background and lighting, and prevented interference from lighting of the environment. This procedure was used throughout all tests done in this study, with varied concentration ranges for the test solutions. For sensitivity testing with Saline creatinine solutions and Porcine Blood, a concentration range of 0-15 mg/dL was used for Saline creatinine tests and a range of 0.1847-54 mg/dL was used. For the Saline tests, the concentrations measured were 0 mg/dL, 0.4685 mg/dL, 0.9375 mg/dL, 1.875 mg/dL, 3.75 mg/dL, 7.5 mg/dL, and 15 mg/dL. For the Porcine Blood tests for Sensitivity, the following creatinine concentrations were used: 0.1847 mg/dL, 11 mg/dL, 21 mg/dL, 32.75 mg/dL, 43 mg/dL, 54 mg/dL. The differences in the ranges are due to the limitations in the amount of solid creatinine that needed to be added into the test blood solutions to create test samples with concentrations below 15 mg/dL. A 0 mg/dL concentration could not be measured for Porcine Blood Samples due to untreated Porcine blood samples having a baseline creatinine concentration of 0.1847 mg/dL. Hence, this was the smallest concentration measured in this range. This is discussed in greater detail in Results, Section 5 . In Upper detection Limit testing, a concentration range of 20-60 mg/dL, at approximately 5 mg/dL increments, was used for both Porcine blood tests and Saline Creatinine tests. The same concentration range was used to maintain consistency between the two tests. For lower detection limit testing with Saline Creatinine Solutions, a concentration range of 1.0-2.6 mg/dL, at 0.2 mg/dL increments was used. These ranges 15 were used to reflect findings in literature, indicating a lower detection limit of 2.0 mg/dL and an", "Upper detection Limit testing, a concentration range of 20-60 mg/dL, at approximately 5 mg/dL increments, was used for both Porcine blood tests and Saline Creatinine tests. The same concentration range was used to maintain consistency between the two tests. For lower detection limit testing with Saline Creatinine Solutions, a concentration range of 1.0-2.6 mg/dL, at 0.2 mg/dL increments was used. These ranges 15 were used to reflect findings in literature, indicating a lower detection limit of 2.0 mg/dL and an upper detection limit of 60 mg/dL for these devices [23, 24] . 7. Processing of Image Data After the images of the reaction site were taken, the images were processed to extract RGB values. This was done by cropping the image of the reaction pad to include only the color-changed region and running it through a Python script that extracted the RGB values of each image and saved it to a csv file. The code is provided in Appendix A . These values were then normalized manually using the image data from the negative control (Images of the sample taken before creatinine was added). The resulting data was then organized into their respective time-points. This was done using Microsoft Excel. Results and Discussion: Saline Creatinine Solution Testing of \u03bc PAD: Prior to testing the \u03bc PAD device with Porcine Blood, the functionality of the device needs to be determined. This is done to ensure that the device is performing as expected or has the potential to yield the desired results, prior to testing in real-world conditions. In order to evaluate the device\u2019s general performance, the sensitivity and detection limits of the device are determined using Saline Creatinine Solutions. These solutions are used as they are easily customizable to any desired creatinine concentration. 1. Sensitivity of \u03bc PAD Device using Saline Creatinine Solutions: Next, the sensitivity of the \u03bc PAD device was determined using the procedures highlighted in Methods, Section 6 . To evaluate the sensitivity of the device, a linear regression was performed on the resulting data for each time point, after the saline solution was added to the device. Sensitivity is defined as the extent to which a device can accurately respond to a known stimulus [7] . In this case, sensitivity is defined as the degree to which the device is able to produce a color change that is proportional to the creatinine concentration of the solution being", ". To evaluate the sensitivity of the device, a linear regression was performed on the resulting data for each time point, after the saline solution was added to the device. Sensitivity is defined as the extent to which a device can accurately respond to a known stimulus [7] . In this case, sensitivity is defined as the degree to which the device is able to produce a color change that is proportional to the creatinine concentration of the solution being added to the device. Since it is known that the Jaff\u00e9 Reaction produces a color change that is directly proportional to the level of creatinine in the solution added, a linear relationship is expected between the saline creatinine solutions (stimulus) and the color change produced (response). Hence, a linear regression was performed on the color change that occurred at each time point. The resulting data is presented in Figures 1-4 . Since the color change was quantified using Image analysis, which yields RGB values for each color channel, the sensitivity was analyzed for the color change in each channel. Figure 1: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 0 minutes after the Solution was Added 16 Figure 2: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 3 minutes after the Solution was Added Figure 3: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 6 minutes after the Solution was Added 17 Figure 4: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 9 minutes after the Solution was Added The linear regression plots in Figures 1-4 revealed that the device had the highest sensitivity at 0 minutes, yielding R 2 values of 0.7761, 0.6564, and 0.6089, for the Red, Green, and Blue Color Channels, respectively. The device reported that the second best sensitivity in the color change readings occurred at 3 minutes after adding the saline creatinine solutions, reporting R 2 values of 0.4754 for the color change in the Red Channel, 0.5089 for the Green Channel, and 0.0053 for the Blue Channel. The device was found to have the worst sensitivity in the readings at 6 minutes after the solution was applied to the device, reporting R 2 values of 0.4436 for the 18 Red Channel, 0.4990 for the Green Color Channel, and 0.1729 for the Blue Color Channel. Moreover, it was found that the color change in the Blue color", "of 0.4754 for the color change in the Red Channel, 0.5089 for the Green Channel, and 0.0053 for the Blue Channel. The device was found to have the worst sensitivity in the readings at 6 minutes after the solution was applied to the device, reporting R 2 values of 0.4436 for the 18 Red Channel, 0.4990 for the Green Color Channel, and 0.1729 for the Blue Color Channel. Moreover, it was found that the color change in the Blue color channel varied significantly across the different time points, whereas the Red and Green color channel trends remained relatively consistent. Although the highest sensitivity occurred immediately after the saline solution was added to the test device, the nature of the driving reaction of the device, that is, the Jaff\u00e9 Reaction suggests that this result is not necessarily accurate. According to literature, the Jaff\u00e9 Reaction requires at least 5 minutes to reach completion [7] . As a result, it is unlikely that the reaction had reached completion immediately after the sample was added to the input square. Moreover, considering the fact that the test solution is not directly added to the reaction square of the device, by design of the \u03bc PAD device, it is also likely that the solution had not completely reached the reaction square, immediately after the solution was added to the device. Experimental observations also support this theory as no color change or wetting of the reaction square was observed at this time point. Therefore, t = 3 minutes is considered to be the point of highest sensitivity for this device. The results also revealed that the device has the least sensitivity at 6 minutes, immediately following the time point with the highest sensitivity. This trend was found to be consistent with research findings reporting a drop in sensitivity of the Jaff\u00e9 Reaction after it reaches its maxima [7] , due to an increase in vulnerability of the reaction to interference that was observed. Consequently, this drop in sensitivity that occurred at 6 minutes is likely a result of this phenomenon. Variability in the blue channel data throughout the results is likely attributed to the type of color change induced by the Jaff\u00e9 Reaction. The Jaff\u00e9 Reaction is known to produce a yellow-orange color change in the reaction square. Consequently, the color change is likely to be most reflected in the Red and Green Color Channels in the", "reaction to interference that was observed. Consequently, this drop in sensitivity that occurred at 6 minutes is likely a result of this phenomenon. Variability in the blue channel data throughout the results is likely attributed to the type of color change induced by the Jaff\u00e9 Reaction. The Jaff\u00e9 Reaction is known to produce a yellow-orange color change in the reaction square. Consequently, the color change is likely to be most reflected in the Red and Green Color Channels in the images taken of the reaction square, and very minimally in the Blue Color Channel (depending on the lighting conditions, and time of day). Therefore, the color change occurring in this channel is not considered to be a reliable indicator of sensitivity, and is considered as experimental noise. Despite being able to derive conclusive results from the present dataset, the overall correlation between the color change and the creatinine concentration in the test solution was weak. The R 2 values derived from this test were consistently less than 0.5, indicating a poor fit. Moreover, the data itself was visually scattered, not indicative of a particular distribution, aside from linear. This is likely due to limitations or experimental errors that result in variance in the present dataset. Possible Sources of Error contributing to variability in the data is discussed in the Sources of Error section. 2. Detection Limits of \u03bc PAD Device using Saline Creatinine Solutions: a. Lower Detection Limit Analysis: In addition to determining the upper detection limit, the lower detection limit of the \u03bc PAD was determined. This was done to determine the minimum concentration of creatinine that 19 could be detected by the device. This is essential to determining the effectiveness and assessing performance of a device and whether it has the potential to effectively detect creatinine levels within the ranges needed for proper CKD monitoring. The lower detection limit was determined using the procedures described in Methods, Section 6 . In order to determine the lower detection limit of the device, a 2-sample t-test for the color change at each concentration was performed. This test consisted of comparing the mean color change that occurred at each concentration and comparing it to the results of a negative control sample. In this case, the negative control sample was the mean color change that occurred when a test solution containing no creatinine was added to the device (purely Phosphate Buffer Solution).", "6 . In order to determine the lower detection limit of the device, a 2-sample t-test for the color change at each concentration was performed. This test consisted of comparing the mean color change that occurred at each concentration and comparing it to the results of a negative control sample. In this case, the negative control sample was the mean color change that occurred when a test solution containing no creatinine was added to the device (purely Phosphate Buffer Solution). The lowest concentration that was found to have a statistically significant difference was determined to be the lower detection limit of the device. For this test, the null hypothesis states that the means of the color change produced by various concentrations are statistically significantly different from the negative control. The alternative hypothesis for this test is that the means of the two groups are not statistically different from one another. Prior to running this statistical test, the normality of the present dataset must be determined as it is a requirement for the 2-sample t-test. The results of this normality test are provided in Figures 5-9, for both the negative control and the color change at each time point. Figure 5: Normality test Results for Negative Control Figure 6: Normality test Results for Color Change at 0 minutes 20 Figure 7: Normality test Results for Color Change at 3 minutes Figure 8: Normality test Results for Color Change at 6 minutes 21 Figure 9: Normality test Results for Color Change at 9 minutes The probability plots reveal that the data is not normally distributed at all the time points, rather the distribution varies across all time points. As a result, the method of concentration comparison at the non-normally distributed time points must consist of non-parametric tests since it doesn\u2019t fulfill the criteria for a 2-sample t-test, specifically for the color change at 0 and 3 minutes. As a result, a Mann-U-Whitney test was used to determine the lower detection limit at 0 and 3 minutes. This test serves as the non-parametric equivalent of the 2-sample t test, using median comparison instead of mean comparison. The comparisons were done on Minitab Statistical Software. Figure 10: Mann-U-Whitney Results for Color Change at 0 minutes Concentration P-value 22 1.0mg/dL 0.016 1.2 mg/dL 0.023 1.4 mg/dL 0.004 1.6 mg/dL 0.008 1.8 mg/dL 0.013 2.0 mg/dL 0.121 2.2 mg/dL 0.061 2.4 mg/dL 0.295 2.6 mg/dL 0.038", "test was used to determine the lower detection limit at 0 and 3 minutes. This test serves as the non-parametric equivalent of the 2-sample t test, using median comparison instead of mean comparison. The comparisons were done on Minitab Statistical Software. Figure 10: Mann-U-Whitney Results for Color Change at 0 minutes Concentration P-value 22 1.0mg/dL 0.016 1.2 mg/dL 0.023 1.4 mg/dL 0.004 1.6 mg/dL 0.008 1.8 mg/dL 0.013 2.0 mg/dL 0.121 2.2 mg/dL 0.061 2.4 mg/dL 0.295 2.6 mg/dL 0.038 Figure 11: Mann-U-Whitney Results for Color Change at 3 minutes Concentration P-value 1.0mg/dL 0.004 1.2 mg/dL 0.013 1.4 mg/dL 0.045 1.6 mg/dL 0.048 1.8 mg/dL 0.019 2.0 mg/dL 0.037 2.2 mg/dL 0.026 2.4 mg/dL 0.096 2.6 mg/dL 0.032 Figure 12: 2-Sample t-Test Results for Color Change at 6 minutes 23 Concentration P-value 1.0mg/dL 0.006 1.2 mg/dL 0.009 1.4 mg/dL 0.009 1.6 mg/dL 0.727 1.8 mg/dL 0.245 2.0 mg/dL 0.519 2.2 mg/dL 0.019 2.4 mg/dL 0.056 2.6 mg/dL 0.050 Figure 13: 2-Sample t-Test Results for Color Change at 9 minutes Concentration P-value 1.0mg/dL 0.015 1.2 mg/dL 0.023 1.4 mg/dL 0.064 1.6 mg/dL 0.376 1.8 mg/dL 0.021 2.0 mg/dL 0.051 2.2 mg/dL 0.011 2.4 mg/dL 0.063 2.6 mg/dL 0.295 24 Across all time points, it was found that the first instance of a p-value less than 0.05 was at a concentration of 1.0mg/dL, indicating that there was a statistically significant difference between the color change in the negative control and device, based on the null hypothesis. Therefore, this value is considered to be the lower detection limit of the device. Here, the first instance reflects the lowest concentration wherein a color change was detected. Unlike the other tests involving Saline Creatinine Solutions, there was no difference between the lower detection limit value detected across the different time points, although there were slight variations in the p-values. Considering the entire Lower detection limit dataset as a whole, it was found that there was a great amount of variation in the p-value reported for the different concentrations. Some concentrations reported very large p-values and some time-points reported multiple concentrations with significant differences. These observations indicate that there was no consistent trend that was observed in the reported p-values and that a high degree of variability exists. This indicates that the data collected from this test is not very precise. The causes of this variability is discussed in the Sources of Error section. Originally, Signal-to-Noise", "a great amount of variation in the p-value reported for the different concentrations. Some concentrations reported very large p-values and some time-points reported multiple concentrations with significant differences. These observations indicate that there was no consistent trend that was observed in the reported p-values and that a high degree of variability exists. This indicates that the data collected from this test is not very precise. The causes of this variability is discussed in the Sources of Error section. Originally, Signal-to-Noise Ratio was to be used as the method of determining the lower detection limit of the device. However, due to the degree of variability in the data, this method was found to be incompatible as it yielded concentration values that were beyond the concentrations used for the testing range of this test. Therefore, 2-sample t-Test and Mann-Whitney-U test was used in its place. b. Upper Detection Limit Analysis: In order to evaluate the performance of the device for higher creatinine concentrations and quantify its range of optimal performance, the upper detection limit of the device was determined using Saline creatinine samples. For this test, Saline creatinine solutions were used as a means of determining whether the device is able to function as expected. This was done by following the procedures highlighted in Methods, Section 6 . The results from the tests are provided in Figures 14-17 . Figure 14: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 0 minutes after the Solution was Added 25 Figure 15: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 3 minutes after the Solution was Added 26 Figure 16: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 6 minutes after the Solution was Added Figure 17: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 9 minutes after the Solution was Added 27 Based on the results provided in Figures 14-17 , it was found that the device performs poorly when larger concentrations of creatinine are added. This is demonstrated by the consistently low R 2 values that were observed for the color change in each channel, across the 4 time points. On average, these R 2 values were found to be less than 0.3, indicating a poor linear correlation. Although other fit types, such as exponential, or polynomial increased the R 2 values slightly, increasing them by 0.1 at most, this type of relationship is not", "device performs poorly when larger concentrations of creatinine are added. This is demonstrated by the consistently low R 2 values that were observed for the color change in each channel, across the 4 time points. On average, these R 2 values were found to be less than 0.3, indicating a poor linear correlation. Although other fit types, such as exponential, or polynomial increased the R 2 values slightly, increasing them by 0.1 at most, this type of relationship is not expected for these variables as literature indicates a strong linear relationship between creatinine concentration and color change in the Jaff\u00e9 reaction [8] . Moreover, the color change in the Red and Green Color Channels reported a negative slope. This indicates that the degrees of color change were found to decrease as the concentration of creatinine in the test solution increased. This contradicts the experimental observations made during the testing process wherein the device exhibited rapid wetting and color change at higher concentrations than at lower concentrations. As a result of the high variability in the linear relationship in the Red and Green Color Channel data, which are key indicators of the color change in the device, it can be concluded that the Upper detection Limit of the device for Saline Creatinine Solutions is 20 mg/dL. Beyond this concentrations, a high degree of variability is observed in the color change that occurs, which is indicative of the device\u2019s unreliability at those concentrations. Comparing the R 2 values across the different time-points, it can be seen that at 0 minutes, the R 2 values were relatively high, potentially indicating an increase in device performance at this time point. However, given that the Jaff\u00e9 reaction takes at least 5 minutes to develop a maximum color change [8] , it is unlikely that the color change observed at this point is reflective of the actual color change produced by the analyte. Therefore, the color change at this time-point was considered unreliable. The unreliability and variability observed in the present dataset is likely due to various limitations in the experimental procedure that interfered with the color change observed. The factors contributing to this are discussed in the Sources of Error section. The highest R 2 values were observed at 3 minutes after the sample was added to the device, indicating that the reaction reached a maximum sensitivity at this time, for this concentration range. These", "change at this time-point was considered unreliable. The unreliability and variability observed in the present dataset is likely due to various limitations in the experimental procedure that interfered with the color change observed. The factors contributing to this are discussed in the Sources of Error section. The highest R 2 values were observed at 3 minutes after the sample was added to the device, indicating that the reaction reached a maximum sensitivity at this time, for this concentration range. These findings were consistent with the point of maximum sensitivity that was observed for other Saline Solution tests. Porcine Blood Testing of \u03bc PAD After testing the device with Saline Creatinine Solutions, the performance of the device was tested using Porcine Blood Solutions. This was done to analyze the device\u2019s performance in real-use conditions whereby the user will input a blood sample into the device to measure blood creatinine levels. Through this analysis, the viability and effectiveness of this device for these applications can be determined. 28 3. Calibration Curve for Determination of Creatinine Concentration in Porcine Blood Prior to testing with blood, the baseline concentration of creatinine in the blood must be determined. The method used to determine this is highlighted in Methods, Section 5 . The resulting calibration curve is presented in Figure 18 . Figure 18: Calibration Curve for Serum Creatinine Determination in Porcine Blood Samples The plot from Figure 18 reveals a linear relationship between the average absorbance and saline creatinine concentration, yielding an R 2 value of 0.912. This indicates that approximately 86.73% of the variability in the data arises from a linear relationship between the two variables. This linear relation is expected since it is known that absorbance is directly proportional to solution concentration, as stated by the Beer-Lambert Law [5] . Moreover, since the Jaff\u00e9 Reaction produces a color change directly proportional to creatinine levels in the sample, a linear output was expected [6] . As a result of the R 2 value being greater 0.8, this curve can be used to approximate the concentration of creatinine in porcine blood samples, based on their absorbance. Although a relatively good R 2 value was obtained (e.g. > 0.9), a higher R 2 value was expected for this relationship (e.g. R 2 > 0.95) . This deviation is likely due to variability in the data collection process resulting in variability in the absorbance values. In order", ". As a result of the R 2 value being greater 0.8, this curve can be used to approximate the concentration of creatinine in porcine blood samples, based on their absorbance. Although a relatively good R 2 value was obtained (e.g. > 0.9), a higher R 2 value was expected for this relationship (e.g. R 2 > 0.95) . This deviation is likely due to variability in the data collection process resulting in variability in the absorbance values. In order to produce a measurable color change in the standard saline creatinine solutions used to make the calibration curve, the Jaff\u00e9 reaction was induced on the saline solutions. This reaction is known to be time-dependent, with color intensity increasing rapidly over time [7] . As a result, slight inconsistencies in the absorbance measurement times, such as measuring absorbance a couple seconds before or after the reaction reaches maximum color change, are likely to yield absorbance measurements that are slightly higher than the true value. This phenomenon was also observed experimentally, whereby the absorbance values continuously changed as time passed. Consequently, this is the likely cause of the R 2 value being lower than expected. Despite these 29 limitations, the calibration curve can still be used to provide a reasonable estimate of the creatinine concentrations in the Porcine Blood Samples. Next, the concentration of creatinine in the given blood sample was determined, as well as the concentration of creatinine in standard Porcine Blood solutions of increasing creatinine concentrations. Methods described in Methods, Section 4 were followed to make this curve. This was done to determine the baseline concentration of creatinine in the given Porcine Blood Sample, which is essential for making standard Blood creatinine solutions, and to validate the calibration curve created and test whether this relationship can be used to determine the concentration of creatinine in the blood. The absorbance measurements taken of Porcine blood of various expected concentrations are provided in Figure 19 below. The absorbance measurements taken of Porcine Blood before creatinine was added yielded an average absorbance of 0.569. Using the relationship derived from the calibration curve (e.g. the equation of the best fit line), it was found that the concentration of creatinine in the sample is approximately 0.1847 mg/dL, which is much lower than expected physiological range of 0.6-1.6mg/dL. However, since this was within the range of the calibration curve, this concentration is considered to be accurate.", "in Figure 19 below. The absorbance measurements taken of Porcine Blood before creatinine was added yielded an average absorbance of 0.569. Using the relationship derived from the calibration curve (e.g. the equation of the best fit line), it was found that the concentration of creatinine in the sample is approximately 0.1847 mg/dL, which is much lower than expected physiological range of 0.6-1.6mg/dL. However, since this was within the range of the calibration curve, this concentration is considered to be accurate. Based on this value, a set of standard Porcine Blood creatinine solutions can be made. To further validate the calibration curve against the standard porcine blood samples, the absorbance of these standard solutions were measured and compared against the calibration curve in Figure 18 . The concentration of each solution was determined based on the measured absorbance by using the equation of the best fit line in Figure 18. Figure 19: Table of Average Absorbance Measurements and Calculated Creatinine Concentrations for Various Known Creatinine Concentrations of Porcine Blood Amount of Solid Creatinine added to 9 mL of Blood (mg) Calculated Creatinine Concentration (mg/dL) Average Absorbance Curve-Based Creatinine Concentration (mg/dL) 0 mg - 0.569 0.1847 0.982 mg 11 2.441 1.634 1.964 mg 21 2.394 1.598 2.946 mg 32.755 1.724 1.079 3.928 mg 43 0.771 0.346 4.91 mg 54 1.615 0.995 30 The absorbance measurements reported in Figure 19 indicate a concentration for the creatinine solutions that are much less than the hand calculated concentration measurements, reporting calculations much less than 2 mg/dL, even though the calculated concentrations of solutions added are within 11-54 mg/dL. Since these values are calculated based on the baseline concentration of the Porcine Blood and the mass of solid creatinine added to the Porcine Blood Solutions, the calculated concentrations are considered to be a more accurate representation of the concentration of the Porcine Blood Solutions. This discrepancy is likely due to the concentrations of the Standard Porcine Blood Samples being too large to be measured accurately by the spectrophotometer. According to literature, UV-Vis Spectrophotometers, similar to those used to measure absorbance of the standard solutions, have an optimal performance range for absorbance measurements between 0.1-1.0; when the spectrophotometer reports absorbance measurements beyond this range, at high concentrations, absorbance measurements are found to be not as reliable due to instrument limitations [10, 11] . Moreover, studies reveal that the Beer-Lambert does not hold for highly concentrated solutions, rather", "Blood Samples being too large to be measured accurately by the spectrophotometer. According to literature, UV-Vis Spectrophotometers, similar to those used to measure absorbance of the standard solutions, have an optimal performance range for absorbance measurements between 0.1-1.0; when the spectrophotometer reports absorbance measurements beyond this range, at high concentrations, absorbance measurements are found to be not as reliable due to instrument limitations [10, 11] . Moreover, studies reveal that the Beer-Lambert does not hold for highly concentrated solutions, rather absorbance is non-linearly related to the sample concentration due to a shading effect that was observed in neighboring chromophores in the solution [12] . As a result, this variation in the absorbance measurements for these concentrations are expected. These variations can be resolved by measuring much lower concentrations, whose absorbance is within the operating range of the UV-VIS spectrophotometer. Therefore, these absorbance values for these concentrations and the corresponding concentrations that were derived from Figure 18 , are inconclusive. However, due to limitations in blood availability for experiments and time-constraints, Porcine Blood standard solutions of the correct concentration range could not be made to properly validate the the calibration plot as it would require larger-than available amounts of blood in order to prepare standard Porcine Blood Solutions of smaller creatinine concentrations that could be properly detected by the spectrophotometer. This is needed since the analytical scales available in the BME labs do not have the proper resolution to measure out masses less than 0.1mg, which is required to make these standard solutions. The addition of saline creatinine solutions of a certain concentration was also not viable as it would interfere with the hematocrit of the blood, and therefore the transparency of the solution and its measured absorbance. Additional interfering factors, such as poor plasma separation and computational errors in the development of the Standard Porcine Blood Creatinine solutions for validation, are discussed in further detail in the Sources of Error section. 4. Sensitivity of the \u03bc PAD Device using Porcine Blood Samples The sensitivity of the \u03bc PAD Device in detecting creatinine levels in Porcine Blood was also examined. For this analysis, the same method used to determine the sensitivity of creatinine in Saline Creatinine Solutions was used to determine the sensitivity of the device in detecting creatinine levels in Porcine Blood. The details of this method are provided in Methods, Section 6 . The results from the sensitivity analysis", "section. 4. Sensitivity of the \u03bc PAD Device using Porcine Blood Samples The sensitivity of the \u03bc PAD Device in detecting creatinine levels in Porcine Blood was also examined. For this analysis, the same method used to determine the sensitivity of creatinine in Saline Creatinine Solutions was used to determine the sensitivity of the device in detecting creatinine levels in Porcine Blood. The details of this method are provided in Methods, Section 6 . The results from the sensitivity analysis are provided in Figures 19-22. 31 Figure 19: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 0 minutes after the Solution was Added Figure 20: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 3 minutes after the Solution was Added Figure 21: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 6 minutes after the Solution was Added 32 Figure 22: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 9 minutes after the Solution was Added The linear regressions modeled in Figures 19-22 reveal that the device reached peak sensitivity 6 minutes after the sample was administered to the device, yielding R 2 values of 0.5771, 0.5061, and 0.6689 for Red, Green, and Blue Channels, respectively. This was higher than the peak sensitivity that was observed in Saline Creatinine Solutions where the maximum sensitivity occurred at 3 minutes. This is likely due to the diffusion rate of the fluid through the device\u2019s plasma separation membrane layer. Given that the Saline creatinine solutions were made using a large quantity of DI water which has a low viscosity of 1 cP, the solution travels through this separation layer much faster than the Whole Porcine Blood samples which has a viscosity of ~3.5 cP and travels much slower through the Plasma membrane layer [8] . As a result, the creatinine in the Blood likely takes longer to reach the reaction square and reaction with the Jaff\u00e9 33 Reaction Reagents than the creatinine in the Saline Solution. Ultimately, this results in a slower maximum sensitivity time in Porcine Blood than in Saline Solutions. This was also observed experimentally whereby the blood applied to the device took longer to be absorbed into the device than the Saline Solution. Unlike the Saline Solutions, there was no increase in sensitivity at 0 minutes, rather the device reported having the worst sensitivity at this time, with R 2 values of 0.0008", "33 Reaction Reagents than the creatinine in the Saline Solution. Ultimately, this results in a slower maximum sensitivity time in Porcine Blood than in Saline Solutions. This was also observed experimentally whereby the blood applied to the device took longer to be absorbed into the device than the Saline Solution. Unlike the Saline Solutions, there was no increase in sensitivity at 0 minutes, rather the device reported having the worst sensitivity at this time, with R 2 values of 0.0008 for the Red Channel, 0.0021 for the Green Channel, and 0.115 for the Blue Channel. This was consistent with expected results at 0 minutes for the device, given the device design and the nature of the Jaff\u00e9 Reaction. The R 2 values for the sensitivity at 3 minutes and 9 minutes were very similar for Porcine blood samples, with the R 2 values at 9 minutes being slightly lower than those at 3 minutes. Both R 2 values for 3 and 9 minutes were the lowest values reported for this test. The drop in R 2 values and therefore sensitivity of the device at 9 minutes is consistent with literature findings that suggest that sensitivity of the Jaff\u00e9 Reaction declines after the point of maximum sensitivity has been reached [7] . These findings are also similar to the trends in the R 2 values in the Porcine Blood samples whereby the R 2 value dropped after they reached their highest sensitivity. Across all the time points, however, the lower variability was observed in the R 2 values for the Blue Color Channel, compared to the Saline creatinine tests. The best-fit line was more consistently linear for these tests compared to the Saline creatinine tests, where the trends are highly variable. Considering that nearly identical testing conditions were used for the Porcine Blood Tests and the Saline Creatinine Solution-based tests, this increased linearity is likely due to random chance, especially considering that the color change observed in the device is not reflected by this color channel. Comparing the R 2 values at maximum sensitivity for Porcine Blood and Saline Creatinine Solutions, it was found the Porcine Blood Samples yielded higher R 2 values than those found in Saline Creatinine Solutions, which yielded R 2 values of 0.4754, 0.5089, and 0.0053 for Red, Green and Blue Color Channels. This indicates that the device has a higher sensitivity in detection in Porcine Blood", "that the color change observed in the device is not reflected by this color channel. Comparing the R 2 values at maximum sensitivity for Porcine Blood and Saline Creatinine Solutions, it was found the Porcine Blood Samples yielded higher R 2 values than those found in Saline Creatinine Solutions, which yielded R 2 values of 0.4754, 0.5089, and 0.0053 for Red, Green and Blue Color Channels. This indicates that the device has a higher sensitivity in detection in Porcine Blood than Saline. This was not consistent with findings in literature which predicts a lower sensitivity in the blood due to interference from other molecules in the blood such as proteins, glucose levels, and bilirubin, and environmental factors such as temperature, and pH [9] . One possible explanation for this phenomenon is the presence of impurities in the Saline Creatinine solution, resulting in the presence of impurities in solution that may have interfered with the reaction. Another contributing factor could arise from the Porcine blood sensitivity test having only 5 distinct concentrations whereas the Saline Solution-based Sensitivity test uses 9 distinct concentrations. This difference in data points likely resulted in the blood samples having a correlation that does not completely capture the entire behavior of the \u03bc PAD device in measuring Porcine blood samples, resulting in a fit that has less variability in the data. Although it was found that the Porcine Blood samples were better detected by the device than the Saline creatinine solutions, and had less variability overall, the R 2 values were relatively low, indicating poor sensitivity. This is likely due to limitations in the experimental design and 34 procedure that may have resulted in inconsistencies in the data obtained. The sources of these limitations are discussed in detail in the Sources of Error section. 5. Detection Limits of \u03bc PAD Device using Porcine Blood Samples a. Lower Detection Limit Analysis: Although it was originally planned to experimentally derive the lower detection limit of the device using Porcine Blood, limitations in the amount of blood available and machine resolution limitations made it difficult to measure the Lower Detection Limit of the \u03bc PAD device using Porcine Blood. One of the critical components to measuring the lower detection limits of the device using Porcine Blood is the synthesis of the Porcine Blood Standard Solutions. This is done by determining the concentration of creatinine already present in the blood and", "derive the lower detection limit of the device using Porcine Blood, limitations in the amount of blood available and machine resolution limitations made it difficult to measure the Lower Detection Limit of the \u03bc PAD device using Porcine Blood. One of the critical components to measuring the lower detection limits of the device using Porcine Blood is the synthesis of the Porcine Blood Standard Solutions. This is done by determining the concentration of creatinine already present in the blood and adding a certain quantity of solid creatinine anhydrous to the blood to obtain the desired concentration of creatinine in the blood. Since a concentration range of 0 mg/dL to 2.6 mg/dL was used to test lower detection limits for Saline Creatinine Solutions, a similar concentration range of creatinine was desired for the Porcine blood test samples for this test. Previous tests using the Calibration Curve discussed in Section 1 revealed that the Porcine Blood Samples given had a pre-existing creatinine concentration of 0.1847mg/dL. As a result, the desired concentration range for lower detection limit testing was 0.1847 mg/dL to ~4.0 mg/dL. When the calculations were done to determine the amount of solid creatinine that needed to be added to create 9mL of 4.0 mg/dL standard Porcine Blood solution for testing which is the highest creatinine solution needed to be made, it was found that 0.000343377g of solid creatinine needed to be added to the existing blood solution to obtain the desired concentration. The calculations for this mass are provided in Appendix B . This would mean that all the masses of solid creatinine needed to be added to make the other standard blood creatinine solutions would be less than this value. Since all of the analytical scales in the BME department do not have the resolution to measure out a mass that small, only having a smallest resolution of 0.0001mg, this concentration of Standard Porcine Blood Creatinine solutions could not be made without requiring a higher volume of blood. The calculations in Appendix B revealed that approximately 100mL of Porcine Blood was necessary to have a solid creatinine mass that could be somewhat accurately measured out using the most precise balance in the BME lab. Since a majority of the 400mL of Porcine blood that was given was used for other study objectives, approximately 50mL of Porcine blood remained. Using this, only one standard Porcine Blood creatinine solution could be", "be made without requiring a higher volume of blood. The calculations in Appendix B revealed that approximately 100mL of Porcine Blood was necessary to have a solid creatinine mass that could be somewhat accurately measured out using the most precise balance in the BME lab. Since a majority of the 400mL of Porcine blood that was given was used for other study objectives, approximately 50mL of Porcine blood remained. Using this, only one standard Porcine Blood creatinine solution could be made, which is a grossly inadequate sample size for estimating the lower detection limit of the device for Porcine Blood. Moreover, the smallest concentration of Blood creatinine solution that could be made was 11 mg/dL, which required 0.000270g of creatinine to be added to the solution to make a 2.5mL solution. Weighing out masses lower than this was found to be very difficult since the mass value would not remain constant and using a larger volume of blood was also risky since the amount of blood available was limited. The calculations for this mass measurement are presented in Appendix B . 35 Alternatives to adding solid creatinine to the blood samples to reach the desired concentration consist of adding saline solutions of a certain concentration of creatinine to whole Porcine Blood, to obtain the desired creatinine concentration. However, this approach was not taken due to the effects this method would have on the Hematocrit value of blood, which may have unintended effects on the color change induced by the Jaff\u00e9 reaction, and thereby disrupt results obtained. By adding a saline creatinine solution of known concentration, the amount of red-blood cells in the solution (e.g. Hematocrit) value decreases, leading to a solution that is lighter in color and less viscous than actual blood. This causes the blood to lose its material properties, resulting in the tests that yield values that are not representative of how the device performs with blood. In this way, it defeats the purpose of testing the device with blood\u2013to simulate real-life use cases of the device\u2013by using a model that does not have the capacity to properly capture the behavior of blood in the device. Hence, this alternative method was not used. Other limitations such as time and plasma separation paper availability also played a role. The limited supply of plasma membrane paper limited the amount of devices that could be made. Time constraints of the project", "this way, it defeats the purpose of testing the device with blood\u2013to simulate real-life use cases of the device\u2013by using a model that does not have the capacity to properly capture the behavior of blood in the device. Hence, this alternative method was not used. Other limitations such as time and plasma separation paper availability also played a role. The limited supply of plasma membrane paper limited the amount of devices that could be made. Time constraints of the project also limited the amount of tests that could be done. As a result, priority was given to tests that could provide a more accurate representation of the device\u2019s performance such as sensitivity and upper detection limit testing of Porcine blood samples. b. Upper Detection Limit Analysis In order to properly evaluate the performance of the device at higher concentrations, the upper detection limits were also tested using Porcine Blood Samples. This was done to determine whether the trends observed in the upper detection limit testing for Saline creatinine solutions are consistent with those found in Porcine blood testing. The methods used to test the upper detection limit of the device are provided in Methods, Section 6. The resulting data for the tests are provided in Figures 23-26 . Figure 23: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 0 minutes after the Solution was Added 36 Figure 24: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 3 minutes after the Solution was Added 37 Figure 25: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 6 minutes after the Solution was Added Figure 26: Color Change in \u03bc PAD Device for Saline Creatinine Solutions 9 minutes after the Solution was Added 38 Based on the results found in Figures 23-26 , it was found that device performance is significantly better for higher concentrations in Porcine Blood samples than Saline Creatinine Solutions. This is evident in the R 2 values obtained for each timepoint being higher than those found in the Saline Creatinine Solution testing. For example, at t = 9 minutes, the upper detection limit plot for Saline Creatinine solutions reported R 2 values of 0.0431 for the Red Color Channel, 0.002 for the Green Color Channel, and 0.129 for the Blue Color Channel. In contrast, the Porcine Blood tests for Upper detection limit yielded R 2 values of 0.4538, 0.4568, and 0.2965 for Red,", "the R 2 values obtained for each timepoint being higher than those found in the Saline Creatinine Solution testing. For example, at t = 9 minutes, the upper detection limit plot for Saline Creatinine solutions reported R 2 values of 0.0431 for the Red Color Channel, 0.002 for the Green Color Channel, and 0.129 for the Blue Color Channel. In contrast, the Porcine Blood tests for Upper detection limit yielded R 2 values of 0.4538, 0.4568, and 0.2965 for Red, Green and Blue Color Channels, respectively. As a result, the Porcine Blood tests report a better linear correlation between color change and the creatinine concentration of the solution. Across all the time points, it was found that the highest R 2 value occurred at 6 minutes after the solution was added, yielding R 2 values of 0.5309 for the Red Color Channel, 0.1877 for the Green Color Channel, and 0.5533 for the Blue Color Channel. This indicates that the color change was optimal and reached its maxima at 6 minutes after the solution was added. Similar to the upper detection limit test for saline creatinine, the time of maximum sensitivity was consistent with the sensitivity tests on Porcine Blood, whereby the sensitivity reached peak color change at 6 minutes. In addition to this, there was also no rise in color change at 0 minutes, rather the correlation between color change and concentration was the worst at this time point, reporting R 2 values less than 0.3. This trend was also observed in the Sensitivity testing done using Porcine Blood. A decrease in the R 2 value was observed in the time points after the device reached maximum sensitivity. This was consistent with the trends seen in the sensitivity and upper detection limit tests for Saline and Porcine Blood. This was also found to be consistent with literature whereby there is a decrease in the sensitivity after the point of maximum sensitivity is reached for the Jaff\u00e9 reaction [8] . Moreover, it was found that the slopes of the color change channels were positive for all time points in the Porcine Blood samples, indicating a direct correlation between color change and creatinine concentration in the Porcine Blood samples. This was expected since it is known that the Jaff\u00e9 Reaction induces a color change that is directly proportional to the creatinine concentration in the solution. This contrasts the results found for Saline", "sensitivity is reached for the Jaff\u00e9 reaction [8] . Moreover, it was found that the slopes of the color change channels were positive for all time points in the Porcine Blood samples, indicating a direct correlation between color change and creatinine concentration in the Porcine Blood samples. This was expected since it is known that the Jaff\u00e9 Reaction induces a color change that is directly proportional to the creatinine concentration in the solution. This contrasts the results found for Saline creatinine solutions which displayed a negative correlation between color change and concentration. Less variability between color channels was also observed in the plots with the trendlines for all the color channels being comparatively closer together than in the Saline Solution tests. This reveals that the device was able to detect color change with more precision in the blood samples than in the saline solutions. This trend was also observed in sensitivity testing of the device, whereby the Porcine Blood data displayed less variability compared to Saline creatinine solutions. Overall, since a direct, linear correlation was observed for the color change in Porcine blood samples, without the presence of any plateaus in the plot of the data, it can be determined that the upper detection limit of the device is 55 mg/dL as the plot reveals a change in color in the device that is consistent with the relationship observed. That is, the color change in the 39 device for this concentration is closer to the best-fit line. Beyond this concentration, that is, at 60 mg/dL the change in RGB values is much larger, and consistently produces a color change that is not in-line with the best-fit line, indicating that it does not follow a linear relationship. However, considering that this is only one data point, future testing should evaluate the performance of the device beyond this concentration to assess the variability of the color change for concentrations beyond 55 mg/dL. This upper detection limit value for Porcine Blood samples was found to be significantly higher than that of Saline Creatinine Solutions, which reported an upper detection limit of 20 mg/dL. These differences in the upper detection limit of Porcine Blood and Saline creatinine solutions are likely due to experimental errors or the presence of interfering factors in the Saline Creatinine tests that resulted in higher variability in that dataset and consequently a lower upper detection limit. In general, since the", "This upper detection limit value for Porcine Blood samples was found to be significantly higher than that of Saline Creatinine Solutions, which reported an upper detection limit of 20 mg/dL. These differences in the upper detection limit of Porcine Blood and Saline creatinine solutions are likely due to experimental errors or the presence of interfering factors in the Saline Creatinine tests that resulted in higher variability in that dataset and consequently a lower upper detection limit. In general, since the Jaff\u00e9 reaction is by nature, a non-specific reaction, it is very vulnerable to interference from other molecules in the solution. However, it is expected that the Saline creatinine solutions provide a more accurate estimate of the detection limits a performance of the device as the solution can be customized to remove any molecules that may potentially interfere with the results. This cannot be done in Porcine Blood samples as any form of treatment may affect the material properties of the blood and thus the quality of the results. One possible explanation for this phenomenon is the presence of interfering molecules in the Porcine blood that increase the color change produced by the Jaff\u00e9 reaction. According to literature, it was found that the presence of glucose, bilirubin, aceto \u2010 acetate and cephalosporin has a negative effect on the color change, resulting in a greater change in color [9] . This may have resulted in the elevated upper detection limit that was observed. Despite these findings, variability in the dataset was observed for these results, likely arising from errors made during the testing process and other interfering factors. These variables are discussed in the Sources of Error Section. Sources of Error: The data obtained from this study contained a great deal of variability, in both Saline creatinine solutions and Porcine Blood samples. This variability is likely due to various errors made in the experimental process that likely interfered with the data collected. The potential contributing factors are provided below: 1. Limited Number of Discrete Concentrations Measured for each Test (Sample Size): Although 6 replicate measurements were taken for each concentration and 216 data points for each test, only 9 distinct concentrations were measured for each test. Though replicates provide a good estimate of the color change at a particular concentration, it does not provide much insight into the general behavior of the device over different concentration ranges. Here, it is possible that", "The potential contributing factors are provided below: 1. Limited Number of Discrete Concentrations Measured for each Test (Sample Size): Although 6 replicate measurements were taken for each concentration and 216 data points for each test, only 9 distinct concentrations were measured for each test. Though replicates provide a good estimate of the color change at a particular concentration, it does not provide much insight into the general behavior of the device over different concentration ranges. Here, it is possible that the present dataset is too small to provide an accurate estimate of the behavior of the device, rather a wider concentration range is needed to capture the general trends of the performance over time. This is likely to cause a great deal of noise in the data as observed. 40 Moreover, the sensitivity testing using Porcine Blood consisted of only 5 distinct concentrations due to material limitations (discussed in a later section). This further reduces the sample size of the data, resulting in a relationship that is less representative of the device\u2019s performance. 2. Poor Print Quality and Assembly of the Test Devices Used: Since the devices are hand made, inconsistencies in the filter paper layer size and plasma separation paper layer were observed. Misalignment in the filter paper layer and the plasma separation paper layer, are likely to affect the quality of the results by disrupting the flow of the analyte from the input zone to the reaction area, resulting in only a fraction of the fluid reaching the reaction zone a producing a color change that is much less than the amount expected for the analyte. In addition to this, poor quality printing of the filter paper layer whereby the ink smudges onto the reaction zone was observed to impact the color change. Specifically, the smudging of the ink, even to a small degree, was observed to make the input zone of the paper slightly hydrophobic. As a result, not all of the fluid applied to the input area was absorbed, or did not become absorbed at the same rate as the other fluids. This is likely to impact the corresponding color change as not all of the fluid will reach the reaction zone in time. 3. Dim Lighting of the Test Environment: Another interfering factor is the lighting of the environment where the images were taken. Since the images taken of the device at these concentrations were", "result, not all of the fluid applied to the input area was absorbed, or did not become absorbed at the same rate as the other fluids. This is likely to impact the corresponding color change as not all of the fluid will reach the reaction zone in time. 3. Dim Lighting of the Test Environment: Another interfering factor is the lighting of the environment where the images were taken. Since the images taken of the device at these concentrations were at various times of the day, it is likely that the lack of light in the surrounding environment lowered the brightness of the image, resulting in the images of reaction squares appearing darker or brighter than observed. Although a light box was used to prevent these effects, it is likely that the box did not entirely cancel out the effects of the ambient lighting of the surrounding lab environment, yielding a white casting of the device image was observed at higher concentrations. Figure 27 provides an image of the increase of whitecasting that was observed in the device images at higher concentrations than lower concentrations taken 2 hours apart. Figure 27: Image Comparison of the \u03bc PAD at 25 mg/dL v. 60 mg/dL Creatinine Concentrations for Upper Detection Limit Testing [25 mg/dL] [60 mg/dL] 41 As illustrated in Figure 27 , the white-casting of the image at 60 mg/dL gives a more monotone appearance to the reaction square compared to the 25 mg/dL. Upon noticing this during the testing process, efforts were made to reduce this effect, such as the use of flash and increasing screen and image brightness, however, these changes were not able to effectively remove the white-casting. Since the data is obtained from analyzing the color intensity of the images, these variations in the image quality have a significant effect on the RGB values obtained. Brighter images capture the color change more vividly, resulting in a larger detected color change. Darker images have higher RGB values in the RGB scale (since Black is assigned 255 and White is assigned 0 for computer-based image analysis), and does not capture color change to the same extent. This yields inaccurately higher RGB values, and lower color change. 4. Improper Dispensing of Fluid onto the Input Zone and Image Capturing Times Improper dispensing of the fluid on the device, resulting in splattering of the analyte directly onto the reaction zone, can", "detected color change. Darker images have higher RGB values in the RGB scale (since Black is assigned 255 and White is assigned 0 for computer-based image analysis), and does not capture color change to the same extent. This yields inaccurately higher RGB values, and lower color change. 4. Improper Dispensing of Fluid onto the Input Zone and Image Capturing Times Improper dispensing of the fluid on the device, resulting in splattering of the analyte directly onto the reaction zone, can contribute to a larger, much faster color change as the solution has prematurely been introduced into the reaction zone. Over time, this yields a larger color change in the device than expected for that concentration. Variability in the image capturing times of the reaction zone also contributes to variability in the data. Since the Jaff\u00e9 reaction is a time-dependent reaction, the intensity of the color change increases as time passes [8] . As a result, slight variations in the image capturing times, such as capturing the image too soon or too late, can cause the color intensity in the reaction square to be darker or lighter than expected, leading to inaccurate results. Although precautions were taken to ensure the images were captured at the accurate time, due to the nature of the Jaff\u00e9 reaction, slight variations can have an effect on the color change obtained. 5. Material Limitations Due to delays in the plasma membrane paper delivery, less than 1 sheet of plasma membrane paper was available. Since this remaining portion was shared across 3 groups, limitations existed in terms of how much of this paper could be used. As a result, this placed limitations on the number of devices that could be used for testing, and the number of additional tests and repeat testing that could be done for the given test. As a consequence, not all of the erroneous trends in the data could be corrected for. Moreover, there was a shortage of Porcine blood needed to test additional creatinine concentrations for blood testing. While preparing the various standard Porcine blood solutions, a total of 400mL of Porcine Blood was allocated for use. However, a majority of this volume was used for other experimental tasks in the study such as determining the creatinine concentration in the blood samples, validating the Porcine Blood calibration curve, and upper detection limit testing. As a result, there was only ~50 mL of", "there was a shortage of Porcine blood needed to test additional creatinine concentrations for blood testing. While preparing the various standard Porcine blood solutions, a total of 400mL of Porcine Blood was allocated for use. However, a majority of this volume was used for other experimental tasks in the study such as determining the creatinine concentration in the blood samples, validating the Porcine Blood calibration curve, and upper detection limit testing. As a result, there was only ~50 mL of blood available, leaving only enough blood for creating 5 distinct concentrations. Hence, given these circumstances, and material limitations, 42 only 5 different Porcine Blood concentrations were tested for Sensitivity testing using Porcine Blood. 6. The Non-Specificity of the Jaff\u00e9 Reaction in Porcine Blood Testing According to literature, the Jaff\u00e9 Reaction is a non-specific reaction that is vulnerable to interference with other biomolecules such as proteins, glucose levels, and bilirubin, as well as environmental factors such as temperature, and pH [9] . In Saline Creatinine tests, these factors are easily controlled since the solution can be customized to exclude these known affecting factors. However, in Porcine Blood Samples, these factors are difficult to control as they are extracted from an animal with baseline levels of glucose, proteins, and bilirubin. In fact, it is estimated that Healthy Porcine Blood has 66-116 mg/dL of glucose, 5500-8900 mg/dL of proteins, and 0.1-1.2 mg /dL of bilirubin in the blood [26, 27, 28] . As a result, it is likely that these molecules interfere with the Jaff\u00e9 reaction and affect the color change produced. 7. Calculation Errors in the Solid Creatinine Needed for Porcine Blood Standard Solutions In order to validate the calibration curve made to determine creatinine concentration in Porcine Blood, a series of Porcine blood Standard solutions of various concentrations need to be made. Originally, the intended creatinine concentration range that was to be used for the validation of the calibration curve was 0.1-1.0 mg/dL, however, due to calculation errors in the amount of creatinine needed to be added to the various standard solutions, the concentrations of creatinine added to the standard solutions were off by a factor of 9, meaning that the amount of creatinine added to the standard Porcine Blood Solutions were 9x larger than intended. This ultimately yielded a Porcine blood creatinine concentration range that was much higher than expected, resulting in inconsistencies in the absorbance measurements. 8. Separation Issues", "0.1-1.0 mg/dL, however, due to calculation errors in the amount of creatinine needed to be added to the various standard solutions, the concentrations of creatinine added to the standard solutions were off by a factor of 9, meaning that the amount of creatinine added to the standard Porcine Blood Solutions were 9x larger than intended. This ultimately yielded a Porcine blood creatinine concentration range that was much higher than expected, resulting in inconsistencies in the absorbance measurements. 8. Separation Issues for Blood Plasma from Porcine Blood for Calibration Curve Validation and Baseline Blood Creatinine Levels One of the major issues encountered in the measurement of creatinine levels in blood was the extraction of plasma from the blood samples. Since UV-VIS spectrophotometry is a color-based method, minimal color interference is required to ensure accurate detection of absorbance. Consequently, prior to measuring the absorbance of the blood samples using the spectrophotometer, the samples must be centrifuged to extract the plasma layer which is known to contain the creatinine. However, despite repeated efforts with various centrifuges, blood batches, and speeds, complete plasma separation of the blood was not achieved, resulting in large amounts of hemoglobin, the pigment causing agent in blood, to remain in the sample thus potentially interfering with the absorbance measurements made. In order to properly assess the degree to which the hemoglobin interferes with the creatinine measurement, the absorbance spectrum for both molecules were explored. Figure 28 contains the absorbance spectrum found for both hemoglobin and creatinine, that was found from literature [22, 23] . 43 Figure 28 : Absorbance Spectrum for Hemoglobin and Creatinine Molecules [Creatinine] [Porcine Blood Hemoglobin] Based on these findings, it can be seen that creatinine has a wider absorption spectrum, with a peak absorbance of 8.9 at ~520 nm whereas hemoglobin in Porcine blood is characterized by a sharp peak with a maximum occuring at ~400 nm, with an absorbance of 1.7. Considering this and comparing the two spectrums, it can be concluded that the overlap between the two spectrums are minimal. Therefore, the interference of hemoglobin in the detection of creatinine is expected to be minimal. Although there are small peaks in the absorbance spectrum for hemoglobin (reaching peaks at approximately 540 nm and 570 nm), the usage of highly concentrated creatinine blood solutions minimizes the effect of this interference by producing a higher peak. Research indicates that solutions of high creatinine concentrations produce", "comparing the two spectrums, it can be concluded that the overlap between the two spectrums are minimal. Therefore, the interference of hemoglobin in the detection of creatinine is expected to be minimal. Although there are small peaks in the absorbance spectrum for hemoglobin (reaching peaks at approximately 540 nm and 570 nm), the usage of highly concentrated creatinine blood solutions minimizes the effect of this interference by producing a higher peak. Research indicates that solutions of high creatinine concentrations produce stronger, higher absorbance measurements [22] . Therefore, the effect of the presence of hemoglobin in the test solutions is expected to be minimal, but it is likely that it still may have an effect. Conclusion: Overall, the study found that with Saline Creatinine Solutions, the \u00b5PAD had maximum sensitivity at 3 minutes. With Porcine Blood, the sensitivity was found to be slightly higher, reaching a maximum at 6 minutes. The lower and upper detection limits of the device were found to be 1.0 mg/dL and 20 mg/dL for the Saline Creatinine tests. The upper detection limit for the device with Porcine Blood was found to be 55 mg/dL, which is higher than that found in the Saline creatinine tests. Due to machine and material limitations, the lower detection limit of the device could not be determined for Porcine Blood Samples. Although the Calibration Curve made for Porcine blood yielded a very good correlation, the calculation errors in the validation process for this curve led to inconclusive results. The development and testing of the \u03bc PAD device in this study underscore a critical step forward in the evolution of healthcare diagnostics, particularly for those who suffer with chronic kidney disease (CKD). By creating a paper based microfluidic device aimed at affordable, accessible, and easy to use, this project demonstrates a tangible solution to the limitations posed by conventional diagnostic methods, which often require frequent trips to the hospital, expensive 44 equipment, and specialized settings. The \u00b5pads integration of Jaffe\u2019s reaction with colorimetric detection and smart-phone based image analysis allows for quantitative monitoring of creatinine levels, therefore empowering patients with a tool to manage their health at home. Through testing with saline and porcine blood, the \u03bc PAD was shown to perform reliably within the concentration ranges appropriate for CKD monitoring with measurable sensitivity and clear detection limits. Although some limitations were observed such as variability due to experimental conditions, device assembly", "and specialized settings. The \u00b5pads integration of Jaffe\u2019s reaction with colorimetric detection and smart-phone based image analysis allows for quantitative monitoring of creatinine levels, therefore empowering patients with a tool to manage their health at home. Through testing with saline and porcine blood, the \u03bc PAD was shown to perform reliably within the concentration ranges appropriate for CKD monitoring with measurable sensitivity and clear detection limits. Although some limitations were observed such as variability due to experimental conditions, device assembly inconsistencies, and limited membrane paper, the overall performance of the \u03bc PAD suggests it is a promising alternative to current diagnostic practices. Additionally, the validation process and calibration curve development offer a strong foundation for future researchers aiming to optimize or adapt the device for broader applications Ultimately, this work not only represents a prototype for low cost, point of care CKD monitoring, but also contributes to the broader vision of accessible and preventative healthcare solutions. As diagnostic technology continues to evolve, innovations like the \u03bc PAD are poised to play a pivotal role in closing gaps in healthcare access, specifically for underserved communities. Future Work Future Studies should strive to perform testing in consistent lighting conditions where the surrounding environment has adequate light to properly capture the color change occurring in the device. In addition to this, ways to improve or standardize the print quality and device fabrication should be explored to minimize the interference from fabrication quality. A wider range of concentrations, ideally greater than 9 concentrations, should be tested to ensure the behavior of the device is being properly captured. Greater degree of care should be taken when capturing the photos at specific time points to accurately capture the color change at the desired time intervals. Alternative reactions, aside from the Jaff\u00e9 reaction, such as enzymatic reactions, should also be explored to improve the specificity of the device overall. To address material limitations, future researchers should obtain materials in abundance to prevent sudden shortages that may interfere with the experimental aims. In addition to this, future work should develop better experimental designs to properly validate the calibration curve developed in this study by measuring the absorbance of Standard Porcine Blood Solutions that are within the creatinine range of the calibration curve or within the spectrophotometer\u2019s optimal performance range. Cost Analysis For running a total of 5 different tests (Upper and Lower detection Limit, and Sensitivity, for both", "materials in abundance to prevent sudden shortages that may interfere with the experimental aims. In addition to this, future work should develop better experimental designs to properly validate the calibration curve developed in this study by measuring the absorbance of Standard Porcine Blood Solutions that are within the creatinine range of the calibration curve or within the spectrophotometer\u2019s optimal performance range. Cost Analysis For running a total of 5 different tests (Upper and Lower detection Limit, and Sensitivity, for both Porcine blood Samples and Saline creatinine solutions) using 9 different sample concentrations on the microfluidic device (six replicates per concentration), the breakdown of the costs are provided below. 45 Microfluidic Device Components & Necessities Materials Cost * A4 Filter Paper $38.00 Vivid Plasma Separation GF membrane, 8\" x 11\" sheet $54.61 Plastic Card Backing $8.99 Total $101.60 Materials Cost * 2L Whole Porcine Blood $800.00 Light Box $35.00 Camera $50.00 20uL Micropipette $415.50 20uL Micropipette Tips $160.50 Total $1461.00 Overall Total $1562.60 *Some of the expenses are approximations. Safety: During experimentation, safety precautions are imperative to achieve precise and accurate results without any form of bodily harm inflicted on the researcher. Before beginning any experiments, all Standard Operating Procedures (SOP) were read and understood. The Standard Operating Procedure for Picric Acid that was followed when handling the material is provided in Appendix C . This project consisted of working with hazardous solutions, so it was crucial that the right precautions were taken to ensure safety was maintained as well as a successful project. Since Jaffe\u2019s reaction involves the interaction between creatinine and an alkaline picrate solution, it was crucial that picric acid and sodium hydroxide (NaOH) were handled in a safe manner. Picric acid is a pale yellow, odorless liquid that is highly explosive, especially when dry. It should always stay moist and kept away from metal surfaces, alkalies, or reducing agents. The Globalized Harmonized System (GHS) classification for acute toxicity of 46 picric acid is 3 out of 5, with 1 being the most toxic, therefore, to minimize any contact with these potential sources of danger and avoid inhalation of fumes, all work related to picric acid is meant to be conducted in the fume hood. Additionally, the fume hood sash was lowered the majority of the time when picric acid was inside to prevent inhalation of the toxic fumes. Mixing the NaOH into the picric acid was also", "toxicity of 46 picric acid is 3 out of 5, with 1 being the most toxic, therefore, to minimize any contact with these potential sources of danger and avoid inhalation of fumes, all work related to picric acid is meant to be conducted in the fume hood. Additionally, the fume hood sash was lowered the majority of the time when picric acid was inside to prevent inhalation of the toxic fumes. Mixing the NaOH into the picric acid was also performed in the fume hood. NaOH is a strong base that is corrosive to skin, eyes, and mucous membranes, which supports why it is crucial the necessary personal protective equipment (PPE) is worn. Along with wearing the right clothing and working in the right environments, labeling is necessary to avoid mix-ups. Wrongfully labeling containers can be extremely risky because dangerous chemical reactions can occur, harmful particles can be produced, and overall, it can harm individuals in the lab. Along with properly labeling containers, it's important they are sealed tightly and stored in the right location. The picric acid and NaOH was stored in the acid cabinet. The glass containers with picric acid were rinsed out with DI water three times, then with soap, then again with DI water to ensure no residue of picric acid was in the container. Nothing with picric acid residue went down the sink, rather, it all was poured into the labeled waste container. When centrifuging the blood to obtain the plasma extract, the tubes were made sure to be tightly capped to prevent any spilling and contamination. FDA Considerations: As \u00b5PADs transition from research prototypes to commercially available diagnostic products, they must meet certain regulatory requirements to ensure safety, effectiveness, and quality. In the United States, this responsibility falls under the requirements of the Food and Drug Administration (FDA). Understanding the FDA requirements is essential for developers seeking to bring \u00b5pads into the market, as these devices influence device classification, design controls, clinical validation, along with most market surveillance. Device Classification and Intended Use Under the 21 CFR \u00a7862.1225, devices for creatinine testing are generally classified as Class II medical devices in vitro devices (IVD). The primary intended use is for point-of-care and at-home monitoring by patients who are either at risk or are currently managing CKD. This classification implies a moderate risk level and typically requires a 510(k) premarket approval. Analytical and Clinical Performance", "these devices influence device classification, design controls, clinical validation, along with most market surveillance. Device Classification and Intended Use Under the 21 CFR \u00a7862.1225, devices for creatinine testing are generally classified as Class II medical devices in vitro devices (IVD). The primary intended use is for point-of-care and at-home monitoring by patients who are either at risk or are currently managing CKD. This classification implies a moderate risk level and typically requires a 510(k) premarket approval. Analytical and Clinical Performance Validation For compliance with FDA standards, the \u00b5PAD must undergo through analytical validation including: \u25cf Sensitivity and specificity testing across the clinical range of creatinine concentrations \u25cf Limits of detection validation, as presented in the study \u25cf Reproducibility and repeatability testing \u25cf Stability and shelf-life testing of the reagents Materials Safety and Biocompatibility 47 As the device interacts with blood, all materials, including filter paper, plasma separation membrane paper, and plastic backing, may contact the sample, whilst meeting biocompatibility standards, specifically ISO 10993 for cytotoxicity, irritation, and sensitization. Reagents such as picric acid and NaOH present safety concerns, and although they are pre-applied to the reaction pad and not directly handled by the user, comprehensive chemical safety testing must demonstrate safe disposal protocols present, compliance with hazardous material regulations, and there is no risk of user exposure. 21 CFR \u00a7820.30(g)(design validation) ensures no user exposure to risk under normal operating conditions. Human Usability Since the uPAD is designed for in-home use, human factors engineering becomes a critical component. According to the FDA\u2019s Guidance documented in \u201cApplying Human Factors and Usability Engineering to Medical Devices\u201d, usability testing is necessary, as well as validation of the device\u2019s instructions for use so users can safely and effectively operate the device without professional supervision. Software Application Since the \u00b5PAD is meant to be paired with a mobile application to analyze colorimetric results, the software is considered Software as a Medical Device Software(SaMD) under CFR \u00a7880.6310 and FDA\u2019s \u201cGuidance on Clinical Decision Support Software\u201d. It must demonstrate accuracy and reliability in image processing as well as concentration estimation. Additionally it must protect patient data privacy and comply with 21 CFR Part 11 if electronic records or signatures are used. Labeling Requirements Labeling for IVD devices must comply with 21 CFR Part 809 Subpart B and include intended use, test limitations, step by step instructions, warnings and precautions, and a description of sample handling. Labeling should", "\u00a7880.6310 and FDA\u2019s \u201cGuidance on Clinical Decision Support Software\u201d. It must demonstrate accuracy and reliability in image processing as well as concentration estimation. Additionally it must protect patient data privacy and comply with 21 CFR Part 11 if electronic records or signatures are used. Labeling Requirements Labeling for IVD devices must comply with 21 CFR Part 809 Subpart B and include intended use, test limitations, step by step instructions, warnings and precautions, and a description of sample handling. Labeling should also meet the standards under 21 CFR \u00a7801.5 since the \u00b5PAD is meant for in-home use. Quality System Regulation (QSR ) The manufacturing and post-market activities for the uPAD must comply with 21 CFR 820 which includes the requirements for design controls, production and process controls, corrective and preventative actions, document control and record keeping, and complaint handling and post market-surveillance. These regulations are intended to ensure that the uPAD is consistently designed, produced, and maintained to meet quality and safety standards. 48 Acknowledgements: Section Group Member Abstract Janane Sivakumar Executive Summary Aasiya Jabbar Introduction Praagna Doddaballapur Literature Review Praagna Doddaballapur, Aasiya Jabbar, Janane Sivakumar Proposed Solution Janane Sivakumar Materials & Methods Janane Sivakumar, Aasiya Jabbar, Praagna Doddaballapur Results and Discussion Janane Sivakumar Conclusion Aasiya Jabbar FDA Considerations Aasiya Jabbar Cost Analysis Praagna Doddaballapur Safety Aasiya Jabbar 49 References 1. CDC. Chronic Kidney Disease in the United States, 2023. , 2024.at <https://www.cdc.gov/kidney-disease/php/data-research/index.html> 2. CDC. Chronic Kidney Disease: Common, Serious, and Costly. , 2024.at <https://www.cdc.gov/kidney-disease/ckd-facts/index.html> 3. What Is Chronic Kidney Disease? , 2024.at <https://www.niddk.nih.gov/health-information/kidney-disease/chronic-kidney-disease-c kd/what-is-chronic-kidney-disease> 4. Iwase, H., T. Yamamoto, and D. K. C. Cooper. Episodes of hypovolemia/dehydration in baboons with pig kidney transplants: A new syndrome of clinical importance? Xenotransplantation 26:e12472, 2019. 5. Oshina, I., and J. Spigulis. Beer-Lambert law for optical tissue diagnostics: current state of the art and the main limitations. J. Biomed. Opt. 26:100901, 2021. 6. Liu, W.-S., Y.-T. Chung, C.-Y. Yang, C.-C. Lin, K.-H. Tsai, W.-C. Yang, T.-W. Chen, Y.-T. Lai, S.-Y. Li, and T.-Y. Liu. Serum creatinine determined by Jaffe, enzymatic method, and isotope dilution-liquid chromatography-mass spectrometry in patients under hemodialysis: Creatinine determined by different methods in patients under HD. J. Clin. Lab. Anal. 26:206\u2013214, 2012. 7. Delanghe, J. R., and M. M. Speeckaert. Creatinine determination according to Jaffe--what does it stand for? Clinical Kidney Journal 4:83\u201386, 2011. 8. (PDF) Geometrical Aspects During Formation of Compact Aggregates of Red Blood Cellsat <https://www.researchgate.net/publication/26371526_Geometrical_Aspects_During_For mation_of_Compact_Aggregates_of_Red_Blood_Cells/figures?lo=1> 9. K\u00fcme,", "Yang, T.-W. Chen, Y.-T. Lai, S.-Y. Li, and T.-Y. Liu. Serum creatinine determined by Jaffe, enzymatic method, and isotope dilution-liquid chromatography-mass spectrometry in patients under hemodialysis: Creatinine determined by different methods in patients under HD. J. Clin. Lab. Anal. 26:206\u2013214, 2012. 7. Delanghe, J. R., and M. M. Speeckaert. Creatinine determination according to Jaffe--what does it stand for? Clinical Kidney Journal 4:83\u201386, 2011. 8. (PDF) Geometrical Aspects During Formation of Compact Aggregates of Red Blood Cellsat <https://www.researchgate.net/publication/26371526_Geometrical_Aspects_During_For mation_of_Compact_Aggregates_of_Red_Blood_Cells/figures?lo=1> 9. K\u00fcme, T., B. Sa\u011flam, C. Ergon, and A. R. Sisman. Evaluation and comparison of Abbott Jaffe and enzymatic creatinine methods: Could the old method meet the new requirements? J. Clin. Lab. Anal. 32:e22168, 2018. 10. Falgueras Vallbona, B., A. Kajtazi, G. Shahtahmassebi, and Q. S. Hanley. Raw data and noise in spectrophotometry. Anal. Chim. Acta 1333:343393, 2025. 11. Why is the absorbance reading on my device (spectrometer/colorimeter) unstable or nonlinear at values above 1.0?at <https://www.vernier.com/til/2589> 12. Aramrueang, N., P. Lomwongsopon, S. Boonsong, and P. Kingklao. Improved spectrophotometric method for determination of high-range volatile fatty acids in mixed acid fermentation of organic residues. Fermentation 8:202, 2022. 50 13. Melo, M. M. P., A. Machado, A. O. S. S. Rangel, and R. B. R. Mesquita. Disposable microfluidic paper-based device for on-site quantification of urinary creatinine. Chemosensors (Basel) 11:368, 2023. 14. Tummalapalli, Sri Lekha, and Michelle M. Estrella. 2019. \u201cPredicting Risk of Kidney Disease: Is Risk-Based Kidney Care on the Horizon?\u201d JAMA: The Journal of the American Medical Association . 15. Vaidya, Satyanarayana R., and Narothama R. Aeddula. 2022. Chronic Kidney Disease . StatPearls Publishing. 16. Talalak, Kwanrutai, Julaluk Noiphung, Temsiri Songjaroen, Orawon Chailapakul, and Wanida Laiwattanapaisal. 2015. \u201cA Facile Low-Cost Enzymatic Paper-Based Assay for the Determination of Urine Creatinine.\u201d Talanta 144 (November):915\u201321. 17. Tseng, Chin-Chung, Ruey-Jen Yang, Wei-Jhong Ju, and Lung-Ming Fu. 2018. \u201cMicrofluidic Paper-Based Platform for Whole Blood Creatinine Detection.\u201d Chemical Engineering Journal 348 (September):117\u201324. 18. Jarinya Sittiwong, Fuangfa Unob. 2016. \u201cPaper-based Platform for Urinary Creatinine Detection.\u201d The Japan Society for Analytical Chemistry. 19. Suphanan Sununta, Poomrat Rattanarat, Orawon Chailapakul, Narong Praphairaksit. 2018. \u201cMicrofluidic Paper-based Analytical Devices for Determination of Creatinine in Urine Samples.\u201d The Japan Society for Analytical Chemistry. 20. Athanasios T. Giannitsis, Tallinn Estonia. 2011. \u201cMicrofabrication of biomedical lab-on-chip devices. A review.\u201d Department of Electronics, Tallinn University of Technology. 21. B D Toora, G Rajagopalan. 2001. \u201cMeasurement of creatinine by Jaffe\u2019s reaction - Determination of concentration of sodium hydroxide required", "Platform for Urinary Creatinine Detection.\u201d The Japan Society for Analytical Chemistry. 19. Suphanan Sununta, Poomrat Rattanarat, Orawon Chailapakul, Narong Praphairaksit. 2018. \u201cMicrofluidic Paper-based Analytical Devices for Determination of Creatinine in Urine Samples.\u201d The Japan Society for Analytical Chemistry. 20. Athanasios T. Giannitsis, Tallinn Estonia. 2011. \u201cMicrofabrication of biomedical lab-on-chip devices. A review.\u201d Department of Electronics, Tallinn University of Technology. 21. B D Toora, G Rajagopalan. 2001. \u201cMeasurement of creatinine by Jaffe\u2019s reaction - Determination of concentration of sodium hydroxide required for maximum color development in standard, urine and protein free filtrate of serum.\u201d Indian Journal of Experimental Biology. 22. (PDF) Spectrophotometric assay of creatinine in human serum sampleat <https://www.researchgate.net/publication/274376588_Spectrophotometric_assay_of_cre atinine_in_human_serum_sample/figures?lo=1&utm_source=google&utm_medium=orga nic> 23. (PDF) Maltose-mediated long-term stabilization of freeze- and spray- dried forms of bovine and porcine hemoglobinat <https://www.researchgate.net/publication/334201407_Maltose-mediated_long-term_stab ilization_of_freeze-_and_spray-_dried_forms_of_bovine_and_porcine_hemoglobin/figur es?lo=1> 24. Image-based creatinine monitor for chronic kidney disease at <https://drive.google.com/drive/folders/1GBGCHBP6D_pYLdX9B6OffYaICBtJjUpY> 25. Whelan, A., R. Elsayed, A. Bellofiore, and D. C. Anastasiu. Selective partitioned regression for accurate kidney health monitoring. Ann. Biomed. Eng. 52:1448\u20131462, 51 2024. 26. Eveleth, D. F. The blood chemistry of swine. J. Biol. Chem. 104:559\u2013563, 1934. 27. Hellwing, A. L. F., A.-H. Tauson, and A. Skrede. Blood parameters in growing pigs fed increasing levels of bacterial protein meal. Acta Vet. Scand. 49:33, 2007. 28. Ndabakuranye, J. P., A. E. Rajapaksa, G. Burchall, S. Li, S. Prawer, and A. Ahnood. A novel optical assay system for bilirubin concentration measurement in whole blood. , 2024.at <http://arxiv.org/abs/2406.14816> 29. Tests to Check Your Kidney Healthat <https://www.kidney.org/kidney-topics/tests-to-check-your-kidney-health> 52 Appendix A: Python Script for Extracting RGB Values from Test Images 53 Appendix B: Calculations for the Amount of Creatinine needed for Lower Detection Limit Testing of \u03bc PAD Device for Porcine Blood For a 9mL 4.0 mg/dL Porcine Blood Solution: 9 \ud835\udc5a\ud835\udc3f \u00d7 0 . 01 \ud835\udc51\ud835\udc3f 1 \ud835\udc5a\ud835\udc3f \u00d7 0 . 1847 \ud835\udc5a\ud835\udc54 1 \ud835\udc51\ud835\udc3f \u2248 0 . 016623 \ud835\udc5a\ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc4e\ud835\udc59\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc51\ud835\udc66 \ud835\udc5d\ud835\udc5f\ud835\udc52\ud835\udc60\ud835\udc52\ud835\udc5b\ud835\udc61 \ud835\udc56\ud835\udc5b 9 \ud835\udc5a\ud835\udc3f \ud835\udc5c\ud835\udc53 \ud835\udc35\ud835\udc59\ud835\udc5c\ud835\udc5c\ud835\udc51 9 \ud835\udc5a\ud835\udc3f \u00d7 0 . 01 \ud835\udc51\ud835\udc3f 1 \ud835\udc5a\ud835\udc3f \u00d7 4 . 0 \ud835\udc5a\ud835\udc54 1 \ud835\udc51\ud835\udc3f \u2248 0 . 36 \ud835\udc5a\ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 \ud835\udc56\ud835\udc5b \ud835\udc61\ud835\udc5c\ud835\udc61\ud835\udc4e\ud835\udc59 \ud835\udc61\ud835\udc5c \ud835\udc5a\ud835\udc4e\ud835\udc58\ud835\udc52 \ud835\udc4e 4 . 0 \ud835\udc5a\ud835\udc54 / \ud835\udc51\ud835\udc3f \ud835\udc46\ud835\udc5c\ud835\udc59\ud835\udc62\ud835\udc61\ud835\udc56\ud835\udc5c\ud835\udc5b \ud835\udc34\ud835\udc5a\ud835\udc5c\ud835\udc62\ud835\udc5b\ud835\udc61 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51 \ud835\udc61\ud835\udc5c \ud835\udc4e\ud835\udc51\ud835\udc51 = 0 . 36 \ud835\udc5a\ud835\udc54 \u2212 0 . 016623 \ud835\udc5a\ud835\udc54 = 0 . 343377 \ud835\udc5a\ud835\udc54 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 0 . 343377 \ud835\udc5a\ud835\udc54 \u2192 0 . 000343377 \ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 \ud835\udc61\ud835\udc5c \ud835\udc4f\ud835\udc52 \ud835\udc4e\ud835\udc51\ud835\udc51\ud835\udc52\ud835\udc51 For a 100mL 4.0mg/dL Porcine Blood Solution: 100 \ud835\udc5a\ud835\udc3f \u00d7", "0 . 01 \ud835\udc51\ud835\udc3f 1 \ud835\udc5a\ud835\udc3f \u00d7 4 . 0 \ud835\udc5a\ud835\udc54 1 \ud835\udc51\ud835\udc3f \u2248 0 . 36 \ud835\udc5a\ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 \ud835\udc56\ud835\udc5b \ud835\udc61\ud835\udc5c\ud835\udc61\ud835\udc4e\ud835\udc59 \ud835\udc61\ud835\udc5c \ud835\udc5a\ud835\udc4e\ud835\udc58\ud835\udc52 \ud835\udc4e 4 . 0 \ud835\udc5a\ud835\udc54 / \ud835\udc51\ud835\udc3f \ud835\udc46\ud835\udc5c\ud835\udc59\ud835\udc62\ud835\udc61\ud835\udc56\ud835\udc5c\ud835\udc5b \ud835\udc34\ud835\udc5a\ud835\udc5c\ud835\udc62\ud835\udc5b\ud835\udc61 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51 \ud835\udc61\ud835\udc5c \ud835\udc4e\ud835\udc51\ud835\udc51 = 0 . 36 \ud835\udc5a\ud835\udc54 \u2212 0 . 016623 \ud835\udc5a\ud835\udc54 = 0 . 343377 \ud835\udc5a\ud835\udc54 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 0 . 343377 \ud835\udc5a\ud835\udc54 \u2192 0 . 000343377 \ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 \ud835\udc61\ud835\udc5c \ud835\udc4f\ud835\udc52 \ud835\udc4e\ud835\udc51\ud835\udc51\ud835\udc52\ud835\udc51 For a 100mL 4.0mg/dL Porcine Blood Solution: 100 \ud835\udc5a\ud835\udc3f \u00d7 0 . 01 \ud835\udc51\ud835\udc3f 1 \ud835\udc5a\ud835\udc3f \u00d7 0 . 1847 \ud835\udc5a\ud835\udc54 1 \ud835\udc51\ud835\udc3f \u2248 0 . 1847 \ud835\udc5a\ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc4e\ud835\udc59\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc51\ud835\udc66 \ud835\udc5d\ud835\udc5f\ud835\udc52\ud835\udc60\ud835\udc52\ud835\udc5b\ud835\udc61 \ud835\udc56\ud835\udc5b 100 \ud835\udc5a\ud835\udc3f \ud835\udc5c\ud835\udc53 \ud835\udc35\ud835\udc59\ud835\udc5c\ud835\udc5c\ud835\udc51 100 \ud835\udc5a\ud835\udc3f \u00d7 0 . 01 \ud835\udc51\ud835\udc3f 1 \ud835\udc5a\ud835\udc3f \u00d7 4 . 0 \ud835\udc5a\ud835\udc54 1 \ud835\udc51\ud835\udc3f \u2248 4 . 0 \ud835\udc5a\ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 \ud835\udc56\ud835\udc5b \ud835\udc61\ud835\udc5c\ud835\udc61\ud835\udc4e\ud835\udc59 \ud835\udc61\ud835\udc5c \ud835\udc5a\ud835\udc4e\ud835\udc58\ud835\udc52 \ud835\udc4e 4 . 0 \ud835\udc5a\ud835\udc54 / \ud835\udc51\ud835\udc3f \ud835\udc46\ud835\udc5c\ud835\udc59\ud835\udc62\ud835\udc61\ud835\udc56\ud835\udc5c\ud835\udc5b \ud835\udc34\ud835\udc5a\ud835\udc5c\ud835\udc62\ud835\udc5b\ud835\udc61 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51 \ud835\udc61\ud835\udc5c \ud835\udc4e\ud835\udc51\ud835\udc51 = 4 . 0 \ud835\udc5a\ud835\udc54 \u2212 0 . 1847 \ud835\udc5a\ud835\udc54 = 3 . 8153 \ud835\udc5a\ud835\udc54 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 3 . 8153 \ud835\udc5a\ud835\udc54 \u2192 0 . 0038153 \ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 \ud835\udc61\ud835\udc5c \ud835\udc4f\ud835\udc52 \ud835\udc4e\ud835\udc51\ud835\udc51\ud835\udc52\ud835\udc51 For a 50 mL 4.0mg/dL Porcine Blood Solution: 50 \ud835\udc5a\ud835\udc3f \u00d7 0 . 01 \ud835\udc51\ud835\udc3f 1 \ud835\udc5a\ud835\udc3f \u00d7 0 . 1847 \ud835\udc5a\ud835\udc54 1 \ud835\udc51\ud835\udc3f \u2248 0 . 09235 \ud835\udc5a\ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc4e\ud835\udc59\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc51\ud835\udc66 \ud835\udc5d\ud835\udc5f\ud835\udc52\ud835\udc60\ud835\udc52\ud835\udc5b\ud835\udc61 \ud835\udc56\ud835\udc5b 100 \ud835\udc5a\ud835\udc3f \ud835\udc5c\ud835\udc53 \ud835\udc35\ud835\udc59\ud835\udc5c\ud835\udc5c\ud835\udc51 50 \ud835\udc5a\ud835\udc3f \u00d7 0 . 01 \ud835\udc51\ud835\udc3f 1 \ud835\udc5a\ud835\udc3f \u00d7 4 . 0 \ud835\udc5a\ud835\udc54 1 \ud835\udc51\ud835\udc3f \u2248 2 . 0 \ud835\udc5a\ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 \ud835\udc56\ud835\udc5b \ud835\udc61\ud835\udc5c\ud835\udc61\ud835\udc4e\ud835\udc59 \ud835\udc61\ud835\udc5c \ud835\udc5a\ud835\udc4e\ud835\udc58\ud835\udc52 \ud835\udc4e 4 . 0 \ud835\udc5a\ud835\udc54 / \ud835\udc51\ud835\udc3f \ud835\udc46\ud835\udc5c\ud835\udc59\ud835\udc62\ud835\udc61\ud835\udc56\ud835\udc5c\ud835\udc5b \ud835\udc34\ud835\udc5a\ud835\udc5c\ud835\udc62\ud835\udc5b\ud835\udc61 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51 \ud835\udc61\ud835\udc5c \ud835\udc4e\ud835\udc51\ud835\udc51 = 2 . 0 \ud835\udc5a\ud835\udc54 \u2212 0 . 09235 \ud835\udc5a\ud835\udc54 = 1 . 90765 \ud835\udc5a\ud835\udc54 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 1 . 90765 \ud835\udc5a\ud835\udc54 \u2192 0 . 00190765 \ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 \ud835\udc61\ud835\udc5c \ud835\udc4f\ud835\udc52 \ud835\udc4e\ud835\udc51\ud835\udc51\ud835\udc52\ud835\udc51 For a 2.5 mL 11 mg/dL Porcine Blood Solution: 2 . 5 \ud835\udc5a\ud835\udc3f \u00d7 0 . 01 \ud835\udc51\ud835\udc3f 1 \ud835\udc5a\ud835\udc3f \u00d7 0 . 1847 \ud835\udc5a\ud835\udc54 1 \ud835\udc51\ud835\udc3f \u2248 0 . 00046175 \ud835\udc5a\ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc4e\ud835\udc59\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc51\ud835\udc66 \ud835\udc5d\ud835\udc5f\ud835\udc52\ud835\udc60\ud835\udc52\ud835\udc5b\ud835\udc61 \ud835\udc56\ud835\udc5b 9 \ud835\udc5a\ud835\udc3f \ud835\udc5c\ud835\udc53 \ud835\udc35\ud835\udc59\ud835\udc5c\ud835\udc5c\ud835\udc51 2 . 5 \ud835\udc5a\ud835\udc3f \u00d7 0 . 01 \ud835\udc51\ud835\udc3f 1 \ud835\udc5a\ud835\udc3f \u00d7 11 \ud835\udc5a\ud835\udc54 1 \ud835\udc51\ud835\udc3f \u2248 0 . 275 \ud835\udc5a\ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 \ud835\udc56\ud835\udc5b \ud835\udc61\ud835\udc5c\ud835\udc61\ud835\udc4e\ud835\udc59 \ud835\udc61\ud835\udc5c \ud835\udc5a\ud835\udc4e\ud835\udc58\ud835\udc52 \ud835\udc4e 4 . 0 \ud835\udc5a\ud835\udc54 / \ud835\udc51\ud835\udc3f \ud835\udc46\ud835\udc5c\ud835\udc59\ud835\udc62\ud835\udc61\ud835\udc56\ud835\udc5c\ud835\udc5b \ud835\udc34\ud835\udc5a\ud835\udc5c\ud835\udc62\ud835\udc5b\ud835\udc61 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51 \ud835\udc61\ud835\udc5c \ud835\udc4e\ud835\udc51\ud835\udc51 = 0 . 275 \ud835\udc5a\ud835\udc54 \u2212 0 . 00046175 \ud835\udc5a\ud835\udc54 = 0 . 270 \ud835\udc5a\ud835\udc54 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 0 . 270 \ud835\udc5a\ud835\udc54 \u2192 0 . 000270 \ud835\udc54 \ud835\udc5c\ud835\udc53", ". 00046175 \ud835\udc5a\ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc4e\ud835\udc59\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc51\ud835\udc66 \ud835\udc5d\ud835\udc5f\ud835\udc52\ud835\udc60\ud835\udc52\ud835\udc5b\ud835\udc61 \ud835\udc56\ud835\udc5b 9 \ud835\udc5a\ud835\udc3f \ud835\udc5c\ud835\udc53 \ud835\udc35\ud835\udc59\ud835\udc5c\ud835\udc5c\ud835\udc51 2 . 5 \ud835\udc5a\ud835\udc3f \u00d7 0 . 01 \ud835\udc51\ud835\udc3f 1 \ud835\udc5a\ud835\udc3f \u00d7 11 \ud835\udc5a\ud835\udc54 1 \ud835\udc51\ud835\udc3f \u2248 0 . 275 \ud835\udc5a\ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 \ud835\udc56\ud835\udc5b \ud835\udc61\ud835\udc5c\ud835\udc61\ud835\udc4e\ud835\udc59 \ud835\udc61\ud835\udc5c \ud835\udc5a\ud835\udc4e\ud835\udc58\ud835\udc52 \ud835\udc4e 4 . 0 \ud835\udc5a\ud835\udc54 / \ud835\udc51\ud835\udc3f \ud835\udc46\ud835\udc5c\ud835\udc59\ud835\udc62\ud835\udc61\ud835\udc56\ud835\udc5c\ud835\udc5b \ud835\udc34\ud835\udc5a\ud835\udc5c\ud835\udc62\ud835\udc5b\ud835\udc61 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51 \ud835\udc61\ud835\udc5c \ud835\udc4e\ud835\udc51\ud835\udc51 = 0 . 275 \ud835\udc5a\ud835\udc54 \u2212 0 . 00046175 \ud835\udc5a\ud835\udc54 = 0 . 270 \ud835\udc5a\ud835\udc54 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 0 . 270 \ud835\udc5a\ud835\udc54 \u2192 0 . 000270 \ud835\udc54 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5f\ud835\udc52\ud835\udc4e\ud835\udc61\ud835\udc56\ud835\udc5b\ud835\udc56\ud835\udc5b\ud835\udc52 \ud835\udc5b\ud835\udc52\ud835\udc52\ud835\udc51\ud835\udc52\ud835\udc51 \ud835\udc61\ud835\udc5c \ud835\udc4f\ud835\udc52 \ud835\udc4e\ud835\udc51\ud835\udc51\ud835\udc52\ud835\udc51 54 Appendix C: Standard Operating Procedure for Picric Acid 55 56 57 58 59 60 61", "CharacterizingtheGrowthofBloodClotsNearMechanicalHeartValves B.S.BiomedicalEngineeringBiomedical&ChemicalEngineeringDepartmentCharlesW.DavidsonCollegeofEngineeringMay17,2024 Preparedby:Michelle HerreraHarsimran KaurAnthonyHo Advisedby:Dr. Alessandro BellofioreSanJoseStateUniversityDepartmentChair, BiomedicalEngineering TableofContents FigureList 2TableList 3ExecutiveSummary 4LiteratureReview 5BiomedicalMotivation 6StatementofNeed 6MaterialsandMethods 7I. ThrombogenicityTester(TGT)Preparation 7II.BloodPreparation 8III.TGTTesting 9IV. HeskaDataCollecting 10V. SoundWaveMeasuring 11Results 12I. ClotFormation 12II.CellCountData 14III.SoundWaveData 16Discussion 18Conclusions 19FutureWork 20Safety 21Acknowledgments 22References 23CostAnalysis 26Appendix 27 1 FigureList Figure1 StJudeRegent27mmMHVBeforeClotting Figure2 TGTSetupWithFilledTygonLoop Figure3 ResultsofClottedBloodonHeskaElementHT5Screen Figure4. BottomandTopViewofMHV1FromTestSession4 Figure5. BottomandTopViewofMHV1FromTestSession2 Figure6. BoxPlotofWhiteBloodCellCountsBeforeClottingandClotted Figure7. BoxPlotofRedBloodCellCountsBeforeClottingandClotted Figure8. BoxPlotofPlateletCountBeforeClottingandClotted Figure9. SoundWavesofTGT10MinuteWaterSimulation Figure10. BoxPlotofSoundWavesinWaterSimulation Figure11 BoxPlotofSoundWavesinBloodSimulation 2 TableList Table1 OriginalandClottedMassofMHVsThroughSession2 Table2 CostAnalysisTabledescribingtheapproximatecostsforallthematerialsandequipmentusedintheexperiments. 3 ExecutiveSummary Amechanicalheartvalve(MHV)isamedicaldevicethatreplacesadamagedordiseased heartvalveandperformsthesamefunctionsthatanormalvalvedoes.Regardingthecurrent study, thefocuswillbeonMHVsthatconsistoftwoleafletsthatopenandclose,allowingblood toflowtotheheartorthebody 7 . WhileMHVshavedurabilitiesthatarehighercomparedto otherapproaches;duetothematerialofthesurfaceofthevalve,thepatientisrequiredtobeona lifelongdosageofanticoagulantsinordertopreventorreducetheriskofbloodclotorthrombus formation 8 . Currently, thereexistsnosolutiontothisissue,forcingindividualswithMHVstobe onanticoagulantsfortherestoftheirlivesorrisktheformationofbloodclotting. Regardingtheresearch,planstoaidindeterminingwhenbloodclotsbegintoformby lookingatthechangeinbloodcompositionanddeterminingwhethertheclickingoftheheart valvebecomesquieterasclottingoccurswillbecrucial.SimulationofanMHVinahumanbody isundergonethroughtheutilizationofathrombogenicitytester(TGT),adevicecapableof simulatingpulsatilebloodflowwiththehelpofamotorthatspinstheTGTleftandright.The TGTisplacedinsideanincubatortokeepthebloodatthecorrecttemperature.Eventually, after thesimulationisrunforawhile,bloodclotswillbeginformingonthesurfaceofthevalve,and samplesofthebloodclotsaretakeninordertocharacterizethechangeinbloodcomposition fromtheoriginalbloodsamples.Thebloodsampleswillbecharacterizedthroughtheutilization ofthehematologyanalyzer, HeskaElementHT5,displayingdataintermsofredandwhiteblood cellcountsaswellasplateletcounts,allowingforcomparisonstobemade. Inadditiontocharacterizingthebloodsamples,itisbelievedthatastheclottingbeginsto form,theclickingsoundoftheMHVclosingalsobeginstodecreaseinvolume.Testingthis hypothesisrequiredtheattachmentofamicrophonetotheTGTsimulationandlisteningfora 4 changeinsoundwhenthesimulationwasrunning.Forthecentralhypothesisregardingthis topic,thereremainsapossibilitythatwhenbloodclotsdevelop,theinitialbloodcharacterization changesafterthesimulationisover,andduringtheincubationperiod,soundsareemittedthat maybedetectedbythesensitivemicrophone. LiteratureReview Mechanicalheartvalves(MHVs)areoftenusedforindividualswithheartvalveproblems, althoughtheyhaveconsiderablelimitations.Oneimportantdisadvantageisthedemandforlong-term anticoagulantmedicationtoavoidthrombusdevelopment(Harrisetal.,2015).Nevertheless, anticoagulantmedicationraisesthepossibilityofbleedingproblems,highlightingtheneedforamore effectiveandsafemethodfortreatingthrombusdevelopment. MultiplestudieshavelookedatthethrombogenicityofMHVsandthefunctionofplateletsin thrombosisdevelopment.Forinstance,Hedayatetal.(2017)discoveredthatmechanicalvalvesactivate moreplateletsthanbioprostheticvalvesthroughoutsystole.likewise,Dangasetal.(2016)foundthat prostheticheartvalvethrombosisisacommonconsequenceofMHVimplantation,andplateletactivation servesanimportantroleinthisprocedure. Thecreationofnon-invasivetechnologiestoidentifythrombusformationisacurrentresearch topic.Forinstance,Coenetal.(2021)examinedthefunctionofplateletsinseverecardiovascularevents andemphasizedthepossibilityofplatelet-basedindicatorsforidentifyingthrombusdevelopment. Additionally,JolugboandAriens(2021)lookedintothestructureofthrombiandtheeffectivenessof thrombolysisandthrombectomyinseverestrokes. Somestudiesinvestigatedattheapplicationofprotaminesulfatetoreverseanticoagulationand preventthrombusdevelopment.Forinstance,Ainle(2009)discoveredthatprotaminesulfateinhibits factorVactivation,whichreducesthrombinproduction.Furthermore,Grzymala-Lubanskietal.(2014) foundthatprotaminesulfatecanreverseanticoagulationandavoidthrombusdevelopmentamongpeople withmechanicalheartvalves. 5 Furthertothisresearch,thereisproofthatthethrombogenicityofMHVsisimpactedbya numberofvariables,whichincludethevalve'ssurfacecharacteristics,bloodflowmotion,and theexistenceofadditionalriskfactorslikeatrialfibrillation(Kuliketal.,2016).Asaresult,a completestrategytoavoidthrombusformationamongindividualswithMHVsnecessitatesafull understandingofthesedeterminantsaswellasthecreationofcustomizedtherapeutictreatments. Thecurrentabsenceofconsistenttechniquesforearlyidentificationofbloodclotforms surroundingmechanicalheartvalvesemphasizesthecrucialimportanceofcreatingdependable diagnosticinstruments.Earlyidentificationhasasubstantialinfluenceonthecareofpatients, perhapspreventingadverseeventsandimprovingtherapyefficacy. Furthermore,thelackof noninvasivediagnosticproceduresexacerbatestheproblem,emphasizingtheneedfornovel treatments.CharacterizingbloodclotformssurroundingMHVshasthepotentialtofurther individualizetreatmentoptionsandimprovetherapeuticresults,meetingtheunmetrequirements ofpatientswithMHVs. BiomedicalMotivation Currently, theimplantationofamechanicalheartvalveispresentedwithlimitationsthat requirethepatienttobeonalife-longdosageofanticoagulantstoreducethelikelihoodof thrombusformation.Thislimitationpresentsothercomplications,suchasexcessivebleeding, butwithouttheuseofanticoagulants,thereisalsoahighriskofbloodclottingandfuturefatal complications.Thereisnocurrentsolutiontothisspecificissueduetoalackofknowledgeof whenclottingbeginstoformandthecharacteristicsofthebloodcomponentsbehindtheclotting. TheTGTwillbeabletosimulatethephysiologicalconditionsofhowamechanicalheartvalve wouldworkinsidethebody, providinguswithatooltotesttheclottinggrowthovertime. Alongsidethistool,thehematologyanalyzercanbeusedtomeasurethedifferentbloodcell 6 characteristicsintheclottedblood.Theinformationfoundcanhelpfutureresearchersobtaina betterunderstandingofwhatoccursduringtheformationofclotsonthesurfaceofthe mechanicalheartvalveandhowlongittakesfortheclottingtocommence. StatementofNeed ThelackofaconsistentmethodtodetectbloodclotformationsinMechanicalHeart Valvesatthemomentofformationhighlightstheneedforidentifyingcharacteristicsthatcanaid indiagnosingandtreatingthediseaseassoonaspossible.Inaddition,thereistheabsenceof consistentpinpointersinitsdiagnosisanddiseasemanagement,leadingtoprolongedpatient sufferingordeath.Assuch,itisacriticalnecessitytodevelopareliableandnoninvasivemethod thatcanaidinearlydiagnosisandcharacterization,facilitatethemonitoringofthrombus formation,andimprovetreatmentoutcomeswithoutneedingtocutthepatientopen. CharacterizingtheformationofbloodclotsinMechanicalHeartValveswouldprovidecritical insightsintotheunderlyingaspectsoftheailment,allowingformorepersonalizedtreatment approachesandacceleratingthedevelopmentofeffectivetherapies.Thereisanurgentnecessity toclosetheknowledgegapinmechanicalheartvalvesandcharacterizebloodclotformationsat theirsitestoimprovethelivesofpatientssufferingfromthesedebilitatingconditionsthatwould requireheartvalvereplacementsurgerytobeginwith. MaterialsandMethods I. ThrombogenicityTester(TGT)Preparation BeforeevenconsideringusingtheTGT, severalTygonloopsconsistingofclear PVCtubes,housings,andhoseclampswereputtogether, oneforeachtrial,simulatinga humanbloodvessel.However, missingorincompatiblepartspreventedaccomplishing thistaskinitially, requiringtheutilizationofa3-Dprinterinordertoprintthemissing 7 parts.Regardless,a27mmSt.JudeRegentmechanicalheartvalveshownbelowin Figure1, isplacedinsidetheloopwhereitwillopenandclose,muchlikethatofa biologicalheartvalve,whentheTGTbeginsrunning.ThemovementoftheTGT consistedofamotorcomponentandtheArduinoUnothatreceivedthecodefromthe Arduinosoftware.Inaddition,aBluetoothmicrophonewasalsoplacednearthesiteof theMHVtocapturetheclickingsoundsmadeduringthesimulationinsidetheincubator. Figure1.StJudeRegent27mmMHVBeforeClotting II. BloodPreparation ThebloodusedfortheexperimentswasacquiredfromLampireBiological Laboratoriesandretrievedfromhealthypigs.Thebloodwasshippedinfourseparate 500mLbottleswithNa-Heparintopreventclottingpriortoarrival;animageofthebottle canbeseenintheappendix.Allofthebloodwasmixedintoonecontainertoobtain consistentsamplesofthesamebloodthroughouteachrun.Priortofillingtheloopswith blood,between100-400mLof0.9%NaClsolutionwasaddedtothecontainerofblood tomaintainahematocritthatwasbetween37%-42%.Theamountof 0.9%NaCl solutionvarieddependingonhowhightheinitialhematocritwaswhenreceived.Once properhematocritwasobtained,tenbaselinevaluesweretakenofthebloodtodetermine thecellcount.Priortotheadditionoftheprotaminebuffersolution,theflow 8 measurementoftheTygonloopwastakentoensurethatitwasentirelyfilledupwith minimalairbubbles.TheTygonloopwasfilledwitharound400mLofbloodneededto filltheloopandprotaminebuffersolutionwasaddedaftertoreversetheanticoagulants containedintheblood.Animageofthefillingprocesscanbefoundintheappendix. III. TGTTesting ThefilledTygonloopwasthenplacedontotheTGTsystemshowninFigure2, andintotheincubatorheldat37 \u2103tosimulatethetemperatureofbloodinsideofa human.Thesimulationranfor20-40minutes,dependingonthegraphsproducedbythe flowmeterandwhethertheyshowedthatclottinghadoccurredornot.Ifnoclottingwas noticeable,thesimulationwouldrunforanother10minuteswithamaximumtimeof40 minutes.Onceclotted,theTygonloopwasquicklydisassembledandtheclottedblood waspouredintoabeakerforfurtheranalysis. 9 Figure2.TGTSetupWithFilledTygonLoop IV. HeskaDataCollecting RegardingtheutilizationoftheHeskaHematologyAnalyzer, about3mLof clottedbloodwasdistributedinto10smalltesttubesforeachrun,andfourtofiveruns wereperformedforeachexperimentday. Animageofthesampletesttubecanbefound intheappendix.Eachtesttubewasplacedundertheprobeneedlewiththeaspiratekey beingpressedtoobtainthedifferentbloodcomponentmeasurements.Anexampleofthe differentcomponentsthatweremeasuredisshowninFigure3, however,wewereonly focusedonthewhitebloodcellcount,redbloodcellcount,andplateletcount.Theblood componentvaluesobtainedfromtheclottedbloodarethencomparedtothebaseline bloodvaluestakenbeforethesimulation. 10 Figure3.ResultsofClottedBloodonHeskaElementHT5Screen V. SoundWaveMeasuring TheincorporationofaBluetoothmicrophoneinourTGTsimulationwastotestour hypothesisofwhetherornottheclickingnoiseoftheMHVclosingcouldbecaptured.We attachedthemicrophonetovalve5intest5fortheentire30-minutedurationtolistentothe clickingoftheMHVanddetermineiftherewereanychanges.Inaddition,sincewewereonly abletosuccessfullyobtainsoundwavesforonebloodexperiment,wealsoperformedseveral microphonetestswiththeTygonloopfilledwithwater.Eachwatertestranfor10minutesandit wasrepeatedthreedifferenttimesfortwoMHVsinseparateTGTs.Thiswasdonetoobtain moredatarelatingtotheamplitudeoftheclickingnoise.Furtheranalysiswasperformedin AdobeAuditioneliminatingrepeatingsoundsthatwereirrelevantorsoundsthatobstructedthe 11 clickingnoiseoftheMHV. Results Throughoutthecourse,fivesetsofexperimentswereperformedwithfourtofiverunsin eachexperimenttoobtaincellcountdatapertainingtotheplateletcount,whitebloodcellcount, andredbloodcellcountfromtheclottedbloodgainedthroughtheTGT.Thegathereddatafrom thehematologyanalyzerdisplayedaconsistenttrendbetweenthedifferentbloodcomponents. Thevaluesofthedifferentbloodcomponentsvariedduetodifferentfactorsalteringeachsetof experiments,creatingsomeoutliers.However, thedecreaseinplateletcountandwhitebloodcell countwasconsistentandtheredbloodcellcountincreasedasclottingoccurred. I. ClotFormation DuringtheTGTsimulation,thedurationofclottingconsistedoftakingbetween20and 40minutes.Sometestswerestoppedtooearlybecausewepredictedthatthevalvehadclotted sincetheflowdiagramwasabitnoisy,buttherewerelittletonoclotsfoundontheMHV. ImagesofthisMHVareshownbelowinFigure4, wherewestoppedthesimulationonvalve oneafter20minutes,leadingtonomajorclotsbeingvisibleontheMHV. Onthefrontsideof theMHV, thereareclotsthathadbeenformingalreadyinthebeakerofbloodduetotheblood notbeingmovedaroundasitisintheTGTsimulation.Weaddedtheclotfromthebeaker becausewehadthoughtthatclottinghadoccurredanditslippedthroughtheMHV, however,that wasnotthecase.Inanothertest,specifically,sessiontworunonevalveone,wewere successfullyabletoobtainmassiveclottingwithaclotmassof11657.8mggivenasimulation thatlasted40minutes.Figure5showsbothsidesoftheclottedMHV, andcomparingthe durationofthisMHVandthepreviousone,thereisaclearunderstandingthatclottingcan increasedrasticallywithina20-minuteincrease.Table1depictstheoverallmassesforeach 12 valueduringsession2andvalve1wasthehighest,duetoitbeinginthesimulationlonger. Figure4.BottomandTopViewofMHV1FromTestSession4 Figure5.BottomandTopViewofMHV1FromTestSession2 ValveNumber OriginalMass ClotMass 1 3347mg 11657.8mg 2 3929mg 7214mg 3 3984mg 10105.8mg 4 3980.3mg 8695.3mgTable1.OriginalandClottedMassofMHV\u2019sThroughSession2 13 II. CellCountData Twenty-twototalrunswereperformedthroughouttheentiretyofalloftheexperiments, givingusvaluabledatapertainingtowhitebloodcellcount,redbloodcellcount,andplatelet count.Figures6,7,and8portraythedataobtainedthroughtheexperimentsfromeachspecific clottedvalveandcompareittothebaselinevaluetakenfirst.Thegraphsprovethatasclotting occurred,whitebloodcellcountsandplateletcountsdecreaseddrasticallyandredbloodcell countsincreasedslightly. Therewasacleargreatervariabilityinvalve2forwhitebloodcell countandplateletcountduetosomeofthevalve2runsbeingstoppedbeforeclottinghad occurred,resultinginhighervaluesforthosebloodcomponents.Despitevalve2containinga higherrange,valves1,3,and4showasmallerandlowerrangeofdataforplateletcount, consistentlyremainingaround20 comparedtothemeanplateletcountof13510 3 /\u00b5\ud835\udc3f 10 3 /\u00b5\ud835\udc3f forthebaseline.Forthewhitebloodcellcount,therewasalsoacleardecreasebetweenall valves,butthemeanvaluesvariedforallofthem. 14 Figure6.BoxPlotofWhiteBloodCellCountsBeforeClottingandClotted Figure7.BoxPlotofRedBloodCellCountsBeforeClottingandClotted Figure8.BoxPlotofPlateletCountBeforeClottingandClotted 15 III. SoundWaveData TheuseofaBluetoothmicrophoneinourTGTsimulationwastodetermineifthe clickingoftheMHVcouldbecapturedandanalyzed.Consideringthatwewereonlyableto successfullycaptureclearsoundsofonesimulationwithblood,wealsoperformedseveraltests withwater.Inthewatersimulations,forevery30seconds,thepeakamplitudeofeverysound wavewasrecorded.Figure9showstheconsistentsoundwavesoftheMHVclickinginsideof thewaterTygonloop.Withallthedataobtainedfromthewatertest,wewereabletoplotthe resultsobtainedshowninFigure10.Theaverageamplitudethroughoutthewatersimulationwas 13.72dBandthegraphshowssomevariabilityandoutliers,butthatcouldbeduetotherestill beingexcessivenoiseinthebackground. LookingattheTGTsimulationwithblood,thesameprocesswasperformedtoobtainthe soundwavedatawiththistestcontainingdataupto16minutes.Theoriginalsimulationlasted 30minutes,however,wewereunabletoheartheclickingnoiseafterduetosomenoise obstructingandoverpoweringtheclickingnoiseoftheMHV. Thegraphofthissimulationis showninFigure11.depictingthewidervariabilityofamplitudeswhencomparedtothewater simulation,whichcouldbeduetoexcessivenoisestillbeingpresent.Despitethis,theoverall meanfortheamplitudeinthisrunwas23.66dB,furtherprovingthattheclickingoftheMHV closingcanbeheardthroughthistestandfuturetestsneedtobeperformedtodetermineifthere isachangeinsoundwhenclottingisforming.Valuesoftheamplitudeinbothtestscanbefound inAppendix9. 16 Figure9.SoundWavesofTGT10MinuteWaterSimulation Figure10.BoxPlotofSoundWavesinWaterSimulation 17 Figure11.BoxPlotofSoundWavesinBloodSimulation Discussion Theresultsofexperimentsdesignedtoassesstheevolutionofbloodclotsnear mechanicalheartvalvesprovidesignificantinsightsintothemechanismofthrombus developmentanditsrelationshipwithmodificationsinbloodcompositionandsoundwave patterns.Severalmajorresultshavearisenfromthoroughtestinganddataanalysis,providing insightintoboththebiologicalprocessesimplicatedandthepossibilityfornon-invasive monitoringstrategies. Initially, theexperimentsconsistentlyshowedchangesinbloodcompositionasclotting happened.Theevaluationofwhitebloodcellcount,redbloodcellcount,andplateletcount demonstrateddistincttendenciesrelatedtoclotformation.Whitebloodcellandplateletcounts fellsharply, whereasredbloodcellcountsincreasedsomewhat.Thisconsistencyoverseveral runsconfirmsthetrustworthinessofthecorrelationsseenandlaysthefoundationforfuture researchintotheprocessesbehindthesechanges. Furthermore,theexperimentsprovidedvitalinformationaboutthedynamicsofclot 18 formation.Clottingtimesvariedfrom20to40minutes,withvariationsdependingonthe experimentalcircumstances.Someexperimentswereprematurelyterminatedowingtonoise disruptionorothercircumstances,emphasizingthesignificanceofthoroughmonitoringanddata analysis.However, withinthisrange,considerabledisparitiesinclotmassweredetected,with certainvalvesformingenormousclotsinaveryshortperiodoftime.Thesefindingsemphasize theintricateinteractionofvariablesinfluencingthrombusdevelopmentandtheneedformore studytounderstandtheunderlyingprocesses. Inadditiontobloodcompositionchanges,thetestsinvestigatedthepossibilityofsound waveanalysisasanon-invasivemonitoringtool.ByconnectingaBluetoothmicrophonetothe thrombogenicitytester,wewereabletorecordtheclickingnoisesmadebytheartificialheart valveinsidetheincubator. Investigationofthesesoundwavesfoundcomparablepatternsinboth water(fig.10)andbloodmodels(fig.11),withdistinctpeaksmatchingvalveclosures. Importantly, themagnitudeofthesepeakschangedwiththepresenceofbloodclots,indicating thatsoundwaveanalysismaybeusedtoidentifythrombusdevelopment. Conclusions Bloodclotformationsaroundmechanicalheartvalvescontinuetoremainsignificant drawbackswhenitcomestoheartvalvereplacements,however,asufficientformoftestingwill eventuallyaidinitsearlydetection.Currently, therearenomethodscapableofconsistently achievingthisduetovariousfactors,evenlesssofornon-invasivemethods.Firstly,itwas criticaltounderstandthecorrelationbetweenbloodclotformationandbloodcomposition. Withinthefirstfewsessionsofexperimentation,itwasconcludedthatthebloodcomposition, thatistheredbloodcell,whitebloodcell,andplateletcounts,haveindeedchangedaspredicted withthelattertwodecreasingandtheformerincreasing.Intheoveralltesting,thereexistedonly 19", "/\u00b5\ud835\udc3f forthebaseline.Forthewhitebloodcellcount,therewasalsoacleardecreasebetweenall valves,butthemeanvaluesvariedforallofthem. 14 Figure6.BoxPlotofWhiteBloodCellCountsBeforeClottingandClotted Figure7.BoxPlotofRedBloodCellCountsBeforeClottingandClotted Figure8.BoxPlotofPlateletCountBeforeClottingandClotted 15 III. SoundWaveData TheuseofaBluetoothmicrophoneinourTGTsimulationwastodetermineifthe clickingoftheMHVcouldbecapturedandanalyzed.Consideringthatwewereonlyableto successfullycaptureclearsoundsofonesimulationwithblood,wealsoperformedseveraltests withwater.Inthewatersimulations,forevery30seconds,thepeakamplitudeofeverysound wavewasrecorded.Figure9showstheconsistentsoundwavesoftheMHVclickinginsideof thewaterTygonloop.Withallthedataobtainedfromthewatertest,wewereabletoplotthe resultsobtainedshowninFigure10.Theaverageamplitudethroughoutthewatersimulationwas 13.72dBandthegraphshowssomevariabilityandoutliers,butthatcouldbeduetotherestill beingexcessivenoiseinthebackground. LookingattheTGTsimulationwithblood,thesameprocesswasperformedtoobtainthe soundwavedatawiththistestcontainingdataupto16minutes.Theoriginalsimulationlasted 30minutes,however,wewereunabletoheartheclickingnoiseafterduetosomenoise obstructingandoverpoweringtheclickingnoiseoftheMHV. Thegraphofthissimulationis showninFigure11.depictingthewidervariabilityofamplitudeswhencomparedtothewater simulation,whichcouldbeduetoexcessivenoisestillbeingpresent.Despitethis,theoverall meanfortheamplitudeinthisrunwas23.66dB,furtherprovingthattheclickingoftheMHV closingcanbeheardthroughthistestandfuturetestsneedtobeperformedtodetermineifthere isachangeinsoundwhenclottingisforming.Valuesoftheamplitudeinbothtestscanbefound inAppendix9. 16 Figure9.SoundWavesofTGT10MinuteWaterSimulation Figure10.BoxPlotofSoundWavesinWaterSimulation 17 Figure11.BoxPlotofSoundWavesinBloodSimulation Discussion Theresultsofexperimentsdesignedtoassesstheevolutionofbloodclotsnear mechanicalheartvalvesprovidesignificantinsightsintothemechanismofthrombus developmentanditsrelationshipwithmodificationsinbloodcompositionandsoundwave patterns.Severalmajorresultshavearisenfromthoroughtestinganddataanalysis,providing insightintoboththebiologicalprocessesimplicatedandthepossibilityfornon-invasive monitoringstrategies. Initially, theexperimentsconsistentlyshowedchangesinbloodcompositionasclotting happened.Theevaluationofwhitebloodcellcount,redbloodcellcount,andplateletcount demonstrateddistincttendenciesrelatedtoclotformation.Whitebloodcellandplateletcounts fellsharply, whereasredbloodcellcountsincreasedsomewhat.Thisconsistencyoverseveral runsconfirmsthetrustworthinessofthecorrelationsseenandlaysthefoundationforfuture researchintotheprocessesbehindthesechanges. Furthermore,theexperimentsprovidedvitalinformationaboutthedynamicsofclot 18 formation.Clottingtimesvariedfrom20to40minutes,withvariationsdependingonthe experimentalcircumstances.Someexperimentswereprematurelyterminatedowingtonoise disruptionorothercircumstances,emphasizingthesignificanceofthoroughmonitoringanddata analysis.However, withinthisrange,considerabledisparitiesinclotmassweredetected,with certainvalvesformingenormousclotsinaveryshortperiodoftime.Thesefindingsemphasize theintricateinteractionofvariablesinfluencingthrombusdevelopmentandtheneedformore studytounderstandtheunderlyingprocesses. Inadditiontobloodcompositionchanges,thetestsinvestigatedthepossibilityofsound waveanalysisasanon-invasivemonitoringtool.ByconnectingaBluetoothmicrophonetothe thrombogenicitytester,wewereabletorecordtheclickingnoisesmadebytheartificialheart valveinsidetheincubator. Investigationofthesesoundwavesfoundcomparablepatternsinboth water(fig.10)andbloodmodels(fig.11),withdistinctpeaksmatchingvalveclosures. Importantly, themagnitudeofthesepeakschangedwiththepresenceofbloodclots,indicating thatsoundwaveanalysismaybeusedtoidentifythrombusdevelopment. Conclusions Bloodclotformationsaroundmechanicalheartvalvescontinuetoremainsignificant drawbackswhenitcomestoheartvalvereplacements,however,asufficientformoftestingwill eventuallyaidinitsearlydetection.Currently, therearenomethodscapableofconsistently achievingthisduetovariousfactors,evenlesssofornon-invasivemethods.Firstly,itwas criticaltounderstandthecorrelationbetweenbloodclotformationandbloodcomposition. Withinthefirstfewsessionsofexperimentation,itwasconcludedthatthebloodcomposition, thatistheredbloodcell,whitebloodcell,andplateletcounts,haveindeedchangedaspredicted withthelattertwodecreasingandtheformerincreasing.Intheoveralltesting,thereexistedonly 19 oneoutliertrial,thoughthatiscertainlyduetohumanerror,otherwise,theresultshavebeen consistent,reinforcingthecorrelationbetweenbloodcompositionandbloodclotformation. However, thiswasmerelythefirststepincharacterizingsaidformations. Implementationofthemicrophoneexternallycontainedafewtheoriesinregardto comingupwithanotherwayofcharacterizingthegrowthofbloodclots,andapossible non-invasivemethodatthat.Thedatagatheredduringthissessionusingsaidmicrophonehas confirmedthepresenceoftheclickingsoundthatoccurswhenthethrombogenicitytesteris runningwiththeMHVinside,preciselywhenthevalveopensandcloses.Inaddition,several programsandsoftwareexistinordertoisolatethebackgroundnoisedependingonthe microphoneutilized,hencetheuseofaBluetoothmicrophoneduringthistestingsession.During theoneinstancewherethemicrophoneisutilizedtotrackthesoundofavalvecontainingblood ratherthanwater, itcanbeseenthatthereisageneralbaselinetowheretheminimumand maximumamplitudesoftheclickingsoundsareconsistent,withfewoutliersduetofailureto isolateallofthebackgroundnoise.Asaresult,thetheorythattheamplitudeswillbegin decreasingasabloodclotstartstoformisviable,requiringfurthertestinginordertoconfirmit. FutureWork Intermsofthebloodcellcountmeasurements,withvariousdataalreadygatheredfrom previousgroups,thereisaclearunderstandingthatwhenclottingoccurs,whitebloodcellcount andplateletcountdecrease.Inourtesting,clotsthatwereproducedwereeitherextremely massiveornotvisibleatall,whichwasnotidealintryingtodeterminetheearlystagesof clotting.Togainabetterunderstandinganddataonearlyclotting,futuretestswiththeBluetooth microphoneshouldbeconducted.WehavedeterminedthattheclickingnoiseoftheMHV closinginabloodsimulationispresentandcanclearlybeheardwithamicrophone,thus 20 conductingmoresimulationswillgiveusvaluableinformationpertainingtothesoundscreated whentheMHVisclosing.Aspecificaimforfuturetestingwouldbewhetherornotthesounds becomedamperasclottingisbeginningtoform.Furtheranalysisofthesoundwavescanalsobe performedtofullyeliminateexcessnoiseandcomparingthebloodcountmeasurementstothe soundwavescanpotentiallyhelpstopthesimulationatanearlystage. Safety I. SafetyIssueswithBlood Handlingbloodinthelaboratoryposesinherentrisksduetotheriskofexposure tobloodbornepathogens.AppropriatePPEisnecessarywhilepresentinthelaboratory, namely, labcoatsandglovesmadeofanappropriatematerial(e.g.,nitrile)aswellas safetygogglesinordertopreventdirectcontactwithbloodandlimitexposureto bloodbornepathogens,ensuringthesafetyoftheoccupantsinthelongrun.Mix-upsand thelikearedetrimentalwhenhandlingblood,requiringproperandaccuratelabelingof thebloodsamplesandcomponentstopreventthis.Approvedprotocolsalsocallfor blood-contaminatedmaterialstobedisposedof.Appropriatedisinfectants,suchas bleach,willdecontaminateblood-contaminatedmaterialsbeforetheyaredisposedof. Naturally, containers,beakers,andtubesthatwereutilizedduringthisexperimentare placedinbiohazardcontainers,anddisposedofseparately. Inthecaseofanemergency regardingbloodcomingintocontactwiththehumanbody, thelaboratorycontains eyewashstationsandsafetyshowers. II. SafetyIssueswithSharpTools Theutilizationofsharptoolssuchasscrewdriversandscissorsinthelaboratory posessubstantialrisksofinjuriesshouldtheybemishandled.Properconductisa 21 requirementwhilepresentinthelaboratory, usingsaidtoolsasthey\u2019reintendedaswellas safetygogglestoreducetheriskofinjurytotheeyes.Intheeventofaninjurycausedby sharptools,theaffectedareamustbedisinfectedandbandagedimmediatelywiththehelp ofasupervisortopreventinfections. Acknowledgments AnthonyHo ExecutiveSummary,StatementofNeed,MaterialsandMethods:TGTPreparation,Conclusion,Safety Michelle Herrera BiomedicalMotivation,Results,MaterialsandMethods:BloodPreparation,TGTTesting,HeskaAnalysis,SoundWaveAnalysis,FutureWork,CostAnalysis Harsimran Kaur LiteratureReview,Discussion 22 References 1.Ainle,F. (2009).\u201cProtaminesulfatedown-regulatesthrombingenerationbyinhibiting factorVactivation.\u201dPubMedCentral.https://pubmed.ncbi.nlm.nih.gov/19531655/. 2.Jolugbo,P.,&Ariens,R.(2021),\u201cThrombuscompositionandefficacyofthrombolysis andthrombectomyinacuteischaemicstroke.\u201dPubMedCentral. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7610448/ 3.UniversityofBern,\"Reducingtheriskofbloodclotsinartificialheartvalves.\" ScienceDaily. ScienceDaily, January14,2020. <www.sciencedaily.com/releases/2020/01/200114101722.htm>. 4.Coenen,D.,(2021).:Themultifacetedcontributionofplateletsintheemergenceand aftermathofacutecardiovascularevents.\u201dScienceDirect. https://www.sciencedirect.com/science/article/pii/S0021915020315914. 5.Dangas,G.D.,Weitz,J.I.,Giustino,G.,Makkar,R.,&Mehran,R.(2016).ProstheticHeart ValveThrombosis.JournaloftheAmericanCollegeofCardiology,68(24),2670\u20132689. https://doi.org/10.1016/j.jacc.2016.09.958 6.Hedayat,M.,Asgharzadeh,H.,&Borazjani,I.(2017).Plateletactivationofmechanicalversus bioprostheticheartvalvesduringsystole.Journalofbiomechanics,56,111\u2013116. https://doi.org/10.1016/j.jbiomech.2017.03.002 7. HarrisC,CroceB,CaoC.Tissueandmechanicalheartvalves.AnnCardiothoracSurg.2015 Jul;4(4):399.doi:10.3978/j.issn.2225-319X.2015.07.01.PMID:26309855;PMCID: PMC4526499 23 8.Grzymala-Lubanski,B.,Labaf,A.,Englund,E.,Svensson,P. J.,&Sj\u00e4lander,A.(2014). Mechanicalheartvalveprosthesisandwarfarin\u2013treatmentqualityandPrognosis.Thrombosis Research, 133(5),795\u2013798.https://doi.org/10.1016/j.thromres.2014.02.031 Ainle,F. (2009).Protaminesulfatedown-regulatesthrombingenerationbyinhibitingfactorVactivation.PubMedCentral.https://pubmed.ncbi.nlm.nih.gov/19531655/ Coen,D.(2021).Themultifacetedcontributionofplateletsintheemergenceandaftermathofacutecardiovascularevents.ScienceDirect.https://www.sciencedirect.com/science/article/pii/S0021915020315914 Dangas,G.D.,Weitz,J.I.,Giustino,G.,Makkar,R.,&Mehran,R.(2016).ProstheticHeartValveThrombosis.JournaloftheAmericanCollegeofCardiology, 68(24), 2670\u20132689.https://doi.org/10.1016/j.jacc.2016.09.958 Grzymala-Lubanski,B.,Labaf,A.,Englund,E.,Svensson,P. J.,&Sj\u00e4lander, A.(2014).Mechanicalheartvalveprosthesisandwarfarin\u2013treatmentqualityandPrognosis.ThrombosisResearch, 133(5),795\u2013798.https://doi.org/10.1016/j.thromres.2014.02.031 Harris,C.,Croce,B.,&Cao,C.(2015).Tissueandmechanicalheartvalves.AnnalsofCardiothoracicSurgery, 4(4),399.doi:10.3978/j.issn.2225-319X.2015.07.01 Hedayat,M.,Asgharzadeh,H.,&Borazjani,I.(2017).Plateletactivationofmechanicalversusbioprostheticheartvalvesduringsystole.JournalofBiomechanics, 56,111\u2013116.https://doi.org/10.1016/j.jbiomech.2017.03.002 Jolugbo,P.,&Ariens,R.(2021).Thrombuscompositionandefficacyofthrombolysisandthrombectomyinacuteischemicstroke.PubMedCentral.https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7610448/ 24 Kawai,T.,Kawamura,Y.,Kawamura,T.,&Kawai,S.(2018).Directoralanticoagulantsformechanicalheartvalves:Asystematicreviewandmeta-analysis.JournalofCardiology,71(3),215\u2013222.https://doi.org/10.1016/j.jacc.2016.09.958 Kulik,A.,Kulik,M.,&Kulik,T. (2016).Mechanicalheartvalves:Currentstatusandfuturedirections.JournalofThoracicDisease,8(Suppl11),S1181\u2013S1188.https://doi.org/10.21037/jtd.2016.09.101 25 CostAnalysis Quantity Price TotalCost Chemicals ProtamineSulfate-4,400mgPBSbuffersolution-44mL0.9%NaClsolution~1000mLBleach-1gallon $100$0.49permL$29per6000mL$10 $100$21.91$4.83$10 LabEquipment HeskaElementHT5ScaleIncubator ~$7,000~$600~1,200 $7,000$600$1,200 Materials LabSupplies 5mLSyringes-4TapeHoseClamps-16ValveHousingsPVCClearTubes-4setsBluetoothMicrophoneArduinoUnoMotorDriverSt.JudeMechanicalHeartValve-4Screws+Bolts-16setsPaperboats-8Stirringrod-1250mLBeakers-4 $12for100$3$2.48N/A$120.77forentireroll$170$32$6.39$250$1.38for5$15for100$3.65$5for1 $12$3$39.68 $120.77$170$32$6.39$1,000$5.52$15$3.65$20 Samples PorcineBlood-5000mL $304/1000mL $1,520 Software Arduino N/A 0$ Facilities ExperimentsperformedinE233J Total~11,885 26 Appendix Appendix1.LampireBiologicalLaboratoriesPorcineBloodBottle Appendix2.FillingupTygonLoopwithBlood 27 Appendix3.SampleTestTubesForCollectingBlood Appendix4.Compilationofthetechnicalmemorandums: Technical Memorandums 2024 Appendix5.ReferencetooriginalBME198Aproposal: Final Proposal.pdf Appendix6.ClottedMHVssessions1through5: Clotted Valves Session 1 - 5 Appendix7.MassofMHVs:Mass of Clotted and Orignal MHV.pdf Appendix8.CompilationofSoundWaveFiles:TGT Sound Waves Appendix9Soundwaveamplitude(Incrementsof30sec)ofonewatertestandonebloodexperiment:Amplitude Wave Analysis of TGT Appendix10:HeskaData Experiments 1-3 Heska.xlsx 28", "Streamlining Ja\ufb00e Assay Design and Preserving Sensitivity with Atomized Picric Acid BME 298 Project Defense Charles W. Davidson College of Engineering CardioLab Adwait Pathak under Dr. Alessandro Bello\ufb01ore Reading Committee: Dr. Abdulmelik Mohammed and Dr. Alessandro Bello\ufb01ore 1 Nationwide Nephrotic Health and Susceptible Demographics \u25cf Chronic Kidney Disease (CKD) is a global health problem with increasing prevalence, often leading to kidney failure if undiagnosed. \u25cf CKD is a progressive loss of kidney function over time: It is categorized into stages based on glomerular \ufb01ltration rate (GFR), with stage 5 being kidney failure. \u25cf The global prevalence of CKD is rising, largely due to factors such as diabetes and hypertension. \u25cf Early detection is crucial for timely intervention: creatinine measurement \u25cb Creatinine is a cyclic compound produced from the breakdown of creatine in muscle metabolism. \u25cb Its concentration in blood is used to assess kidney function and diagnose Chronic Kidney Disease Creatinine (pngegg.com) Senior Citizens are at heightened risk for CKD (CDC.gov) Existing Lateral Flow Assay (LFA): The Ja\ufb00e Reaction \u25cf Picric Acid (2,4,6-trinitrophenol) forms colored complexes with creatinine, making it ideal for colorimetric assays. \u25cf The Ja\ufb00e Reaction uses picric acid (yellow) to form an orange-colored complex with creatinine. \u25cb The magnitude of change from yellow to orange is concentration dependent \u25cb The complex gives o\ufb00 an orange color used to quantify creatinine levels in a patient\u2019s blood. \u25cb Alternatives to the Ja\ufb00e Assay rely on enzymatic reactions: not cost-e\ufb00ective The Ja\ufb00e Reaction (pngegg.com)Picric Acid Janovsky ComplexCreatinine \u25cf Sample Application: A liquid sample is applied to the sample pad, which directs the \ufb02ow of the sample through the assay strip via capillary action. \u25cf Reaction Zones: The sample interacts with labeled detection molecules in the conjugate pad. These molecules are designed to bind speci\ufb01cally to the target analyte. \u25cf Result Interpretation: The result is interpreted visually, indicating the presence or absence of the target analyte. \u03bcPAD Lateral Flow Assay for Chronic Kidney Disease The Ja\ufb00e Reaction (pngegg.com) sample \ufb01lter paper + thermal ink membrane paper backing layer \u25cf \u03bcPADs are low-cost, portable devices designed for point-of-care diagnostics. \u25cf They can detect biomarkers, including creatinine, through colorimetric reactions, o\ufb00ering a promising alternative to traditional lab-based assays. \u25cf Challenges in \u03bcPAD Sensitivity \u25cb detecting low concentrations of biomarkers such as creatinine is a challenge \u25cb Current limitations in \u03bcPADs necessitate improving sensitivity without sacri\ufb01cing accessibility. Goal: Increase the sensitivity of the device", "Chronic Kidney Disease The Ja\ufb00e Reaction (pngegg.com) sample \ufb01lter paper + thermal ink membrane paper backing layer \u25cf \u03bcPADs are low-cost, portable devices designed for point-of-care diagnostics. \u25cf They can detect biomarkers, including creatinine, through colorimetric reactions, o\ufb00ering a promising alternative to traditional lab-based assays. \u25cf Challenges in \u03bcPAD Sensitivity \u25cb detecting low concentrations of biomarkers such as creatinine is a challenge \u25cb Current limitations in \u03bcPADs necessitate improving sensitivity without sacri\ufb01cing accessibility. Goal: Increase the sensitivity of the device by increasing colorimetric change at lower creatinine levels \u25cf A 2018 study demonstrates stable chromogenic signals for colorimetric detection. \u25cb Optimal reaction time: 5 minutes, temperature range: 35-40\u00b0C. \u25cf Deviations in temperature or time can destabilize the results, a\ufb00ecting accuracy. \u25cb Elevated temperatures disrupt colorimetric response. \u25cb Extending reaction time did not signi\ufb01cantly improve accuracy but introduced variability. \u25cb Precise control of temperature, time, and picric acid concentration is crucial for reliable creatinine measurement. Tseng et al. (2018) Reaction Site \u25cf Reagent Concentrations and Kinetics \u25cb Picric acid and creatinine is not a highly speci\ufb01c reaction. \u25cb Colorimetric changes are not always reliable, especially at low creatinine concentrations. \u25cf In this early study, we can see picric acid concentration\u2019s e\ufb00ect on colorimetric change plateaus around 3 - 4 * 10-3 M \u25cf Conversely, creatinine concentration has a large e\ufb00ect on colorimetric change. \u25cf As a result, further increasing the picric acid concentration is not ideal for measuring the lowest creatinine concentration (Bonsnes et al., 1945) Proposed Improvement and Predicate Study Background from Literature Deng et al. (2018) developed a paper-based colorimetric sensor using a novel spray painting method to de\ufb01ne hydrophobic channels and apply reagents for iron detection. Their work established: \u25cf Spray-based application allows uniform reagent deposition Cost-e\ufb00ective, scalable fabrication for \ufb01eld analysis \u25cf Quantitative colorimetric detection with high sensitivity Connection to This Study My approach mirrors and extends this strategy: \u25cf Spray-based reagent delivery was validated for picric acid application on uPADs \u25cf I con\ufb01rmed mass equivalence between pipetting and spraying, ensuring consistency \u25cf Atomization enables even reagent distribution and preservation of signal \ufb01delity \u25cf Like Deng et al., I used RGB-based colorimetric analysis to quantify analyte response \u25cf Unlike iron detection, this device targets creatinine using the picrate reaction, adapted for improved \ufb01eld-deployable CKD screening Deng et al. (2018) also used a colorimetric assay with sprayed reagents on a paper based analytical device. Challenges Faced and Proposed", "acid application on uPADs \u25cf I con\ufb01rmed mass equivalence between pipetting and spraying, ensuring consistency \u25cf Atomization enables even reagent distribution and preservation of signal \ufb01delity \u25cf Like Deng et al., I used RGB-based colorimetric analysis to quantify analyte response \u25cf Unlike iron detection, this device targets creatinine using the picrate reaction, adapted for improved \ufb01eld-deployable CKD screening Deng et al. (2018) also used a colorimetric assay with sprayed reagents on a paper based analytical device. Challenges Faced and Proposed Improvements: Spraying vs. Pipetting Pipetted samples (left) and sprayed samples (right) Testing and Data Collection Device Validation Work\ufb02ow: \u25cf Creatinine concentration: 1 mg/dL (lower end of physiological creatinine levels) \u25cf Application volume of picric acid: 15 \u00b5L and equivalent spray per zone Colorimetric Analysis: \u25cf Devices imaged under controlled lighting with \ufb01xed-height phone stand and ring light \u25cf ROI (Region of Interest): central 50\u00d750 pixel area cropped from each detection zone \u25cf \u11ebRGB values extracted using Python-based script and validated against ImageJ Controls \u25cf My initial objective was to verify device response using binary controls. The positive control contained a creatinine concentration used for testing. The negative control used a blank (distilled H2O) and the results con\ufb01rmed a reproducible and detectable colorimetric change. \u25cf These \ufb01ndings support the feasibility of detection, and under the right conditions contamination can be avoided. \u25cf Since I was only testing one concentration of creatinine, I used binary controls rather than a calibration curve. Future trials may consider multiple concentrations, for which a calibration curve would be e\ufb00ective. \u25cf Since I am not trying to quantify unknown concentrations, a calibration curve was deemed ine\ufb00ective. Pipetted Controls R G B (+) Before 206 175 76 (+) After 212 174 68 (-) Before 218 219 215 (-) After 222 220 217 Sprayed (+) Before 210 172 108 (+) After 214 169 64 (-) Before 212 208 213 (-) After 216 206 214 Top and bottom sides of the same sample were used for positive and negative controls. 25 uL of 1 mg/dL creatinine was administered, and photos were taken 10 minutes after and before. (Left, pipetted samples; right, sprayed samples) Quartile Plots:\u11ebG and\u11ebR Values at 5 and 10 minutes Mean Plots T-test Test Null hypothesis H\u2080: \u03bc\u2081 - \u00b5\u2082 = 0 Alternative hypothesis H\u2081: \u03bc\u2081 - \u00b5\u2082 \u2260 0 T-Value DF P-Value 9.29 314 0 Test Null hypothesis H\u2080: \u03bc\u2081 - \u00b5\u2082 = 0 Alternative hypothesis H\u2081:", "the same sample were used for positive and negative controls. 25 uL of 1 mg/dL creatinine was administered, and photos were taken 10 minutes after and before. (Left, pipetted samples; right, sprayed samples) Quartile Plots:\u11ebG and\u11ebR Values at 5 and 10 minutes Mean Plots T-test Test Null hypothesis H\u2080: \u03bc\u2081 - \u00b5\u2082 = 0 Alternative hypothesis H\u2081: \u03bc\u2081 - \u00b5\u2082 \u2260 0 T-Value DF P-Value 9.29 314 0 Test Null hypothesis H\u2080: \u03bc\u2081 - \u00b5\u2082 = 0 Alternative hypothesis H\u2081: \u03bc\u2081 - \u00b5\u2082 \u2260 0 T-Value DF P-Value 117.9 226 0 Splitting by Channel and Method Levene\u2019s Test for Equal Variances Descriptive Statistics Descriptive Statistics Sample N Median Sample N Median Time 324 7.5 Time 324 7.5 Value 324 -77.2172 Value 324 8.4946 Estimation for Difference 95.00% Estimation for Difference 95.00% Difference CI for Difference Difference CI for Difference 84.7172 (6.3792, 181.108) -0.532982 (-1.9324, 0.5648) Test Test Null hypothesis H\u2080: \u03b7\u2081 - \u03b7\u2082 = 0 Null hypothesis H\u2080: \u03b7\u2081 - \u03b7\u2082 = 0 Alternative hypothesis H\u2081: \u03b7\u2081 - \u03b7\u2082 \u2260 0 Alternative hypothesis H\u2081: \u03b7\u2081 - \u03b7\u2082 \u2260 0 Method W-Value P-Value Method W-Value P-Value Not adjusted for ties 119718 0 Not adjusted for ties 102708 0.308 Adjusted for ties 119718 0 Adjusted for ties 102708 0.3 Kruskal-Wallis Discussion \u25cf Pipetting outperformed spraying in generating a strong orange chromatic shift, primarily due to large increases in red channel (\u0394R) intensity \u25cf However, pipetted samples showed greater variability, especially at 5 minutes \u25cf Spray method showed lower \u0394G variance and consistent green suppression, indicating better reproducibility \u25cf Red channel intensity dropped in spray samples, suggesting an incomplete Ja\ufb00e reaction, could be due to suboptimal picric acid concentration or reaction time \u25cf \u0394R - \u0394G composite metric shows pipet > spray in total chromatic shift toward orange \u25cf Spray samples still produced reliable negative green shifts, suggesting the reaction proceeded di\ufb00erently than in pipetted samples \u25cf Findings imply that spray is mechanically valid, but pipetting is more reliable regarding chromogenic shift. Future Directions \u25cf Lowering picric acid concentration to reduce oversaturation and potential contaminative e\ufb00ects confounding sprayed samples \u25cf Testing on modi\ufb01ed or treated, more absorbent papers to control \ufb02uid wicking and reaction zone localization \u25cf Using a narrower channel to ensure reagent distribution without \ufb02ooding the analyte pad \u25cf Benchmarking against standard clinical assays and commercial LFAs \u25cf Validating the assay in synthetic or real blood samples \u25cf Adapting Python code", "valid, but pipetting is more reliable regarding chromogenic shift. Future Directions \u25cf Lowering picric acid concentration to reduce oversaturation and potential contaminative e\ufb00ects confounding sprayed samples \u25cf Testing on modi\ufb01ed or treated, more absorbent papers to control \ufb02uid wicking and reaction zone localization \u25cf Using a narrower channel to ensure reagent distribution without \ufb02ooding the analyte pad \u25cf Benchmarking against standard clinical assays and commercial LFAs \u25cf Validating the assay in synthetic or real blood samples \u25cf Adapting Python code for smartphone compatibility with automated \u0394R - \u0394G quanti\ufb01cation \u25cf Printer work\ufb02ow has now been optimized, mass production could be feasible upon having membrane paper lie reliably and reproducibly \ufb02at against the \ufb01lter paper. \u25cf Pipetting is more likely to cause a chromogenic shift than spraying. Questions 1.FASTSTATS - kidney disease. Centers for Disease Control and Prevention. April 28, 2024. Accessed August 8, 2024. 2.Stevens, Lesley A.; Levey, Andrew S.. Measured GFR as a Con\ufb01rmatory Test for Estimated GFR. Journal of the American Society of Nephrology 20(11):p 2305-2313, November 2009. DOI: 10.1681/ASN.2009020171 3.O\u2019Seaghdha CM, Lyass A, Massaro JM, et al. A risk score for chronic kidney disease in the general population. The American journal of medicine. March 1, 2012. Accessed August 8, 2024. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3285426/. 4.Y etisen AK, Akram MS, Lowe CR. Paper-based micro\ufb02uidic point-of-care diagnostic devices. Lab Chip. 2013 Jun 21;13(12):2210-51. doi: 10.1039/c3lc50169h. 5.Tseng C-C, Yang R-J, Ju W-J, Fu L-M. Micro\ufb02uidic paper-based platform for whole blood creatinine detection. April 30, 2018. Accessed August 8, 2024. https://www.sciencedirect.com/science/article/abs/pii/S1385894718307642. 6.Jafry AT, Lim H, Kang SI, Suk JW, Lee J. A comparative study of paper-based micro\ufb02uidic devices with respect to channel geometry. December 29, 2015. Accessed August 8, 2024. https://www.sciencedirect.com/science/article/abs/pii/S0927775715304313. 7.Strong EB, Knutsen C, Wells JT, et al. Wax-printed \ufb02uidic time delays for automating multi-step assays in paper-based micro\ufb02uidic devices (micropads). MDPI. March 19, 2019. Accessed August 8, 2024. https://www.mdpi.com/2411-5134/4/1/20. 8.J V. Reaction of alkaline sodium picrate with creatinine: I. Kinetics and mechanism of formation of the mono-creatinine picric acid complex. Clinical chemistry. October 22, 1976. Accessed August 8, 2024. https://pubmed.ncbi.nlm.nih.gov/10095/. 9.Fu L-M, Tseng C-C, Ju W-J, Yang R-J. Rapid Paper-Based System for Human Serum Creatinine Detection. Inventions. 2018; 3(2):34. https://doi.org/10.3390/inventions3020034 10.Dal Dosso F, Decrop D, P\u00e9rez-Ruiz E, et al. Creasensor: SIMPLE technology for creatinine detection in plasma. Anal Chim Acta. 2018;1000:191-198. doi:10.1016/j.aca.2017.11.026. 11.K\u0131l\u0131\u00e7, V., Horzum, N., & Ertugrul Solmaz, M. (2020). From Sophisticated Analysis to Colorimetric Determination: Smartphone Spectrometers and Colorimetry. IntechOpen.", "of formation of the mono-creatinine picric acid complex. Clinical chemistry. October 22, 1976. Accessed August 8, 2024. https://pubmed.ncbi.nlm.nih.gov/10095/. 9.Fu L-M, Tseng C-C, Ju W-J, Yang R-J. Rapid Paper-Based System for Human Serum Creatinine Detection. Inventions. 2018; 3(2):34. https://doi.org/10.3390/inventions3020034 10.Dal Dosso F, Decrop D, P\u00e9rez-Ruiz E, et al. Creasensor: SIMPLE technology for creatinine detection in plasma. Anal Chim Acta. 2018;1000:191-198. doi:10.1016/j.aca.2017.11.026. 11.K\u0131l\u0131\u00e7, V., Horzum, N., & Ertugrul Solmaz, M. (2020). From Sophisticated Analysis to Colorimetric Determination: Smartphone Spectrometers and Colorimetry. IntechOpen. doi: 10.5772/intechopen.82227 12.Habdias A. Silva-Neto, Iana V.S. Arantes, Andr\u00e9 L. Ferreira, Guida H.M. do Nascimento, Gabriel N. Meloni, William R. de Araujo, Thiago R.L.C. Paix\u00e3o, Wendell K.T. Coltro, 13.Recent advances on paper-based micro\ufb02uidic devices for bioanalysis, TrAC Trends in Analytical Chemistry, Volume 158, 2023, 116893, ISSN 0165-9936, https://doi.org/10.1016/j.trac.2022.116893. 14.Liu, Jingji & Kong, Xiaopeng & Wang, Hongliang & Zhang, Yajun & Fan, Yiqiang. (2019). Roll-to-roll wax transfer for rapid and batch fabrication of paper-based micro\ufb02uidics. Micro\ufb02uidics and Nano\ufb02uidics. 24. 10.1007/s10404-019-2310-2. 15.Beyond Wax Printing: Fabrication of Paper-Based Micro\ufb02uidic Devices Using a Thermal Transfer Printer. Ryan A. Ruiz, Jorge L. Gonzalez, Miguel Vazquez-Alvarado, Nathaniel W. Martinez, and Andres W. Martinez. Analytical Chemistry 2022 94 (25), 8833-8837 DOI: 10.1021/acs.analchem.2c01534 16.Chin-Chung Tseng, Song-Yu Lu, Szu-Jui Chen, Ju-Ming Wang, Lung-Ming Fu, Yi-Hong Wu, Micro\ufb02uidic aptasensor POC device for determination of whole blood potassium, Analytica Chimica Acta, Volume 1203, 2022, 339722, ISSN 0003-2670, https://doi.org/10.1016/j.aca.2022.339722. 17.Tambaru, D., R. H. Rupilu, F. Nitti, I. Gauru, and Suwari. Development of paper-based sensor coupled with smartphone detector for simple creatinine determination. 2017.doi:10.1063/1.4978168 18.Arnold, C., Bissonnette, L., Choi, D. H., Fang, C., Fu, E., Govorov, A. O., Hauck, T. S., He, Y., Kit-Anan, W., Lee, W. G., \u2026 Hart, R. W. (2013a, November 19). Advances in paper-based point-of-care diagnostics. Biosensors and Bioelectronics. https://www.sciencedirect.com/science/article/pii/S095656631300777X 19.Jie Hu, ShuQi Wang, Lin Wang, Fei Li, Belinda Pingguan-Murphy, Tian Jian Lu, Feng Xu, Advances in paper-based point-of-care diagnostics, Biosensors and Bioelectronics, Volume 54, 2014, Pages 585-597, ISSN 0956-5663, https://doi.org/10.1016/j.bios.2013.10.075. 20.Srisawasdi, P., Chaichanajarernkul, U., Teerakanjana, N., Vanavanan, S. and Kroll, M.H. (2010), Exogenous interferences with Ja\ufb00e creatinine assays: addition of sodium dodecyl sulfate to reagent eliminates bilirubin and total protein interference with Ja\ufb00e methods. J. Clin. Lab. Anal., 24: 123-133. https://doi.org/10.1002/jcla.20350 21.Harry Quon, Craig E. Grossman, Rebecca L. King, Mary Putt, Keri Donaldson, Larry Kricka, Jarod Finlay, Timothy Zhu, Andrea Dimofte, Kelly Malloy, Keith A. Cengel, Theresa M. Busch, Interference with the Ja\ufb00\u00e9 method for creatinine following 5-aminolevulinic", "0956-5663, https://doi.org/10.1016/j.bios.2013.10.075. 20.Srisawasdi, P., Chaichanajarernkul, U., Teerakanjana, N., Vanavanan, S. and Kroll, M.H. (2010), Exogenous interferences with Ja\ufb00e creatinine assays: addition of sodium dodecyl sulfate to reagent eliminates bilirubin and total protein interference with Ja\ufb00e methods. J. Clin. Lab. Anal., 24: 123-133. https://doi.org/10.1002/jcla.20350 21.Harry Quon, Craig E. Grossman, Rebecca L. King, Mary Putt, Keri Donaldson, Larry Kricka, Jarod Finlay, Timothy Zhu, Andrea Dimofte, Kelly Malloy, Keith A. Cengel, Theresa M. Busch, Interference with the Ja\ufb00\u00e9 method for creatinine following 5-aminolevulinic acid administration, Photodiagnosis and Photodynamic Therapy, Volume 7, Issue 4, 2010, Pages 268-274, ISSN 1572-1000, https://doi.org/10.1016/j.pdpdt.2010.07.008. (https://www.sciencedirect.com/science/article/pii/S1572100010000918) 22.Roy W. Bonsnes, Hertha H. Taussky, ON THE COLORIMETRIC DETERMINATION OF CREATININE BY THE JAFFE REACTION, Journal of Biological Chemistry, Volume 158, Issue 3, 1945, Pages 581-591, ISSN 0021-9258, https://doi.org/10.1016/S0021-9258(19)51334-5. (https://www.sciencedirect.com/science/article/pii/S0021925819513345) 23.He Y, Zhang S, Zhang X, Baloda M, Gurung AS, Xu H, Zhang X, Liu G. Ultrasensitive nucleic acid biosensor based on enzyme-gold nanoparticle dual label and lateral \ufb02ow strip biosensor. Biosens Bioelectron. 2011 Jan 15;26(5):2018-24. doi: 10.1016/j.bios.2010.08.079. Epub 2010 Sep 9. PMID: 20875950. References", "Automation for Large Throughput Fabrication and Processing of SJSU\u2019s CardioLab\u2019s Paper-Based Microfluidic Devices Almary Bernal August 12, 2025 Dr. Alessandro Bellofiore Reading Committee: Dr. Abdulmelik Mohammed \u00a9 2025 Almary Bernal All rights reserved. Abstract Paper-based microfluidic devices evolved into a state-of-the-art technology in and out of the biomedical field due to its affordability, easy accessibility, and notable properties. As a result, they have become the ideal device for point-of-care sites and, especially, research settings conducting diagnostics or reactivity tests. Although these chips have substantial beneficial characteristics, they still have their drawbacks. The most prominent disadvantages of paper-based microfluidic devices is its lengthy fabrication process- specifically if simple, affordable equipment is used- and throughput (number of chips made and processed per hour). With this being said, this study introduces the implementation of automation techniques for the large throughput in fabrication and processing methods for paper-based microfluidic devices. While there are a multitude of chip structures, dimensions, and fabrication methods, this research study focuses strictly on CardioLab\u2019s predetermined paper-based microfluidic device designs and methodologies. Because of this, the specific aims presented in this paper are imposed to mitigate CardioLab\u2019s procedure\u2019s weaknesses and areas that cause discrepancies. This paper presents the implementation of a streamlined fabrication process that results in a high-yield chip production under a timely manner and an automated syringe pumping technique set to dispense sample solution onto multiple chips in a systematic manner. To do so, a craft cutter, spray adhesive, syringe pump, and XY positioning system were all instituted to CardioLab\u2019s paper-based microfluidic device manufacturing methodologies. Table of Contents 1. Introduction 4 1.1 Microfluidic Devices 4 1.2 CardioLab\u2019s paper-based microfluidic chips 5 1.3 Problem/ Need 7 2. Literature Review 8 2.2 Microfluidic Devices 9 2.3 Paper-based Microfluidic Devices 9 3. Research Hypothesis and Objectives 13 3.1 Specific Aim #1: Fabrication Automation 14 3.2 Specific Aim #2: Processing Automation 14 4. Materials and Methods 15 4.1 Materials 16 4.2 Specific Aim #1: Methodologies 20 4.2.1 Fabrication Process 20 4.2.1.1 Data Analysis 25 4.2.2 Printing Assessment 26 4.2.2.1 Data Analysis 26 4.3 Specific Aim #2: Methodologies 28 4.3.1 Data Analysis 32 4.4 Safety 34 5. Results & Discussion 35 5.1 Specific Aim #1: Printing Assessment 35 5.2 Specific Aim #1: Fabrication Process 36 5.3 Specific Aim #2: Dispensing System 38 5.4 Specific Aim #2: Accuracy 40 6. Limitations & Future Work 41 7. Conclusion 42 References 44 Appendix A 46 Appendix", "Aim #1: Methodologies 20 4.2.1 Fabrication Process 20 4.2.1.1 Data Analysis 25 4.2.2 Printing Assessment 26 4.2.2.1 Data Analysis 26 4.3 Specific Aim #2: Methodologies 28 4.3.1 Data Analysis 32 4.4 Safety 34 5. Results & Discussion 35 5.1 Specific Aim #1: Printing Assessment 35 5.2 Specific Aim #1: Fabrication Process 36 5.3 Specific Aim #2: Dispensing System 38 5.4 Specific Aim #2: Accuracy 40 6. Limitations & Future Work 41 7. Conclusion 42 References 44 Appendix A 46 Appendix B 47 Appendix C 48 1 List of Figures Figure 1. Ivana Kovac\u2019s predetermined chip dimensions and structure. 7 Figure 2. Flowchart depicting the current fabrication process. 20 Figure 3. Flowchart outlining this project\u2019s automated fabrication process. 22 Figure 4. Four-layered \u00b5PAD structure used in this study. 23 Figure 5. Whole Dispensing System Set-Up 28 2 List of Tables Table 1. Fabrication methods, usage, materials, and equipment for paper-based microfluidic devices. 11 Table 2. List of materials used and their justification/use for Specific Aim #1. 16-17 Table 3. List of the materials used and their justification/ use Specific Aim #2. 17-18 Table 4. List of the materials used and their justification/ use Specific Aim #2. 18-19 Table 5. NEMA 17 Motor Wiring with A4988. 32 Table 6. A4988 Pinout. 32 Table 7. Consistency Results for the \u00b5PAD design. 35-36 Table 8. Results for the fabrication process. 36 Table 9. Results for the Dispensing System. 39 3 1. Introduction Advancements to modern medicine through medical devices and drugs allow healthcare providers to facilitate the analysis of the population\u2019s overall well-being. Blood diagnostics is one of the most popular forms of screening tests that allow an examination of a patient\u2019s internal well-being to be performed. Different properties of the blood molecules are investigated when these tests are processed. There are various forms of analysis procedures that are commonly used in the clinical and research setting. The difference between the two environments is that normally there is a particular aim in the research setting while a clinical setting explores all possibilities. What is meant by this? Research labs focus on their precise topic (i.e. detection of cancerous cells) while in clinical settings, phlebotomists analyze the blood sample for varying substances. While both settings have their own requirements for sample analysis tests, they both have the urgent need of having a tool be accessible, fast, and easy-to-use. Tests like saliva or nasal-swab tests that", "that normally there is a particular aim in the research setting while a clinical setting explores all possibilities. What is meant by this? Research labs focus on their precise topic (i.e. detection of cancerous cells) while in clinical settings, phlebotomists analyze the blood sample for varying substances. While both settings have their own requirements for sample analysis tests, they both have the urgent need of having a tool be accessible, fast, and easy-to-use. Tests like saliva or nasal-swab tests that present the patient\u2019s diagnosis for certain diseases are primarily used for flu or COVID-19 testing 5 . Much like these tools, lab-on-chip devices can also analyze a patient\u2019s well-being. Microfluidic chips have become a staple, state-of-the-art device in the biomedical field. 1.1 Microfluidic Devices Microfluidic chips are micro devices that are commonly used for detection, mixing, and separation of small volumes of reagents 2 . These devices have multiple components that serve for intended purposes. Microchannels are the essential factor for each device as this is what determines the function of the chip. Because of their growing use in the biomedical setting, different types of microfluidic devices have been developed. From cross-junction to sieve type to bottleneck design, these devices are actively being designed to target different applications 12 . 4 Paper-based microfluidic devices (\u00b5PADs) have established themselves as inexpensive and effective while also being easy to fabricate. These specific chips are becoming increasingly useful for clinical diagnostics. Due to their associated characteristics, they are actively being used in research laboratories. Though, time efficiency is a concern for these chips as a larger amount of samples theoretically takes a significant amount of time to process. . Now, paper-based microfluidic chips are used in environmental and biomedical diagnostics 9 . Research has been done for the use of multi-channel chips that can detect different pathogens and heavy metals in substances and reagents 16 . Unfortunately, minimal research has been extended for distinct use of multi-channel paper-based microfluidic chips that can detect certain elements in blood samples in a more efficient, fast-paced, and cost-effective manner. Tedious is the only way to describe the fabrication process of a single paper-based microfluidic chip. While these chips are highly desirable over other forms of microfluidic devices, the procedure to create and process these is lengthy and precise. Currently, in research labs and clinical settings, blood diagnostics can take hours for a large population of samples.", "distinct use of multi-channel paper-based microfluidic chips that can detect certain elements in blood samples in a more efficient, fast-paced, and cost-effective manner. Tedious is the only way to describe the fabrication process of a single paper-based microfluidic chip. While these chips are highly desirable over other forms of microfluidic devices, the procedure to create and process these is lengthy and precise. Currently, in research labs and clinical settings, blood diagnostics can take hours for a large population of samples. This not only affects the timeline of a research project but also the efficiency, accuracy, and precision of the gathered data by both the chip and the research associate. 1.2 CardioLab\u2019s paper-based microfluidic chips As previously mentioned, paper-based microfluidic devices are the desired device for reactivity and diagnostics in research settings. San Jose State University is the home of CardioLab\u2019s CKD research projects that study Jaffe\u2019s method for the detection of chronic kidney disease through creatinine and picric acid/NaOH solutions. Chronic kidney disease(CKD) is the prolonged alteration of a patient\u2019s kidney\u2019s structure and/or function 15 . This disease is typically screened with the use of the glomerular filtration rate (GFR) through its biomarkers: creatinine 5 and cystatin C 15 . Creatinine detection can be obtained through Jaffe\u2019s reaction following the colorimetric method where in the case of a creatinine and picric acid/ NaOH reaction, the solution will turn an orange-red color 14 . This reaction is studied in CardioLab through the use of paper-based microfluidic devices. The CKD projects manually construct each chip used for the creatinine analysis, the chip\u2019s predetermined dimensions and setup can be seen in Figure 1. CardioLab\u2019s CKD projects are part of a collaboration with Santa Clara University for Machine Learning in Motion (MLM). The aim of this collaboration is to train the MLM to process and analyze images of \u00b5PADs after the blood sample has been dispensed onto the chips\u2019 well. In order to train the MLM, the CKD projects need to analyze hundreds of \u00b5PADs under different parameters, therefore upwards to thousands of chips need to be made for all the CKD groups. Although CardioLab follows a simple fabrication procedure, the process to fabricate at least 20 chips is extremely lengthy and has concurrent variation and inconsistencies throughout. Because each chip is manually cut and assembled by the student researcher, a number of errors are introduced. 6 Figure 1. Ivana Kovac\u2019s predetermined chip dimensions", "the MLM, the CKD projects need to analyze hundreds of \u00b5PADs under different parameters, therefore upwards to thousands of chips need to be made for all the CKD groups. Although CardioLab follows a simple fabrication procedure, the process to fabricate at least 20 chips is extremely lengthy and has concurrent variation and inconsistencies throughout. Because each chip is manually cut and assembled by the student researcher, a number of errors are introduced. 6 Figure 1. Ivana Kovac\u2019s predetermined chip dimensions and structure. 7 1.3 Problem/ Need There is a need for automation methods that produce a large-throughput fabrication of paper-based microfluidic chips in order to make these devices more accessible and easy to use, especially in a research setting. Under the direction of SJSU\u2019s Dr. Bellofiore\u2019s CardioLab, paper-based microfluidic devices are being fabricated- from scratch- for a number of studies to examine and quantify the efficiency of these chips. It is through this extensive process that we can personally see the drawbacks and points of weakness for these chips: the printing inconsistencies, the cutting of the filter and membrane paper, the attachment process of the filter 7 and membrane paper, and the solution dispensing methodology. The ongoing methods used are successful as paper-based microfluidic chips are still created and used for research studies. Yet, they do have the points of improvements mentioned earlier. There is a need for these points to be addressed and refined in order to build a time efficient, productive research lab as it will facilitate a fruitful fabrication process. In this paper, I subjected the aforementioned factors for an automation enhancement. Various automation methods were studied and compared in order to establish the most effective technique when contrasted with the manual process. The automations presented are also key to address the high-yield production need for the MLM training. If we reduce overall fabrication and dispensing duration, more chips can be produced and therefore processed, again resulting in a highly efficient lab making ample progress on the MLM end goal. The findings of this research verify the hypothesis that the proposed automation techniques increase time efficiency and throughput of CardioLab\u2019s paper-based microfluidic device production while also taking accuracy into account. 2. Literature Review Point-of-care diagnostic devices have become an established need for clinical settings around the world where a limited amount of resources are available. Unfortunately, these locations are the ones with the most dire need for", "in a highly efficient lab making ample progress on the MLM end goal. The findings of this research verify the hypothesis that the proposed automation techniques increase time efficiency and throughput of CardioLab\u2019s paper-based microfluidic device production while also taking accuracy into account. 2. Literature Review Point-of-care diagnostic devices have become an established need for clinical settings around the world where a limited amount of resources are available. Unfortunately, these locations are the ones with the most dire need for medical attention. Sometimes, said point-of-care (POC) settings are in remote areas where expensive medical equipment is not part of their main priority. Due to this, WHO\u2019s \u201cASSURED (i.e. affordable, sensitive, specific, user-friendly, rapid and robust, equipment-free and deliverable to users)\u201d (Kumar, 2019) 9 guidelines become prominent. Settings with a certain amount of supplies need efficient and effective on-the-go type devices that can fit a laboratory\u2019s work into a singular device. From this 8 need, microfluidic devices became about and have since grown into becoming a great asset to the biomedical engineering field. 2.2 Microfluidic Devices Microfluidic chips have established themselves as state-of-the-art devices that have elevated clinical performance. Not only do they prove to be efficient in a clinical setting, such as in a point-of-care setting, but also in the research level. Due to their characteristics- low-cost, easy to use, transportable, and quick- they\u2019ve been proven to be beneficial as lab-on-chip. From being first introduced in 1979 1 , microfluidic devices have expanded to take various different applications, forms, designs, and fabrication methods. Initially, microfluidic chips were designed with the use of common polymers as well as other popular but expensive materials such as silicon and glass 1 . Now, these chips have become more accessible through the use of lower cost materials and with a simpler fabrication process. Microfluidic devices are known for their advancements in being used for synthesis, in-vitro models, and organ-on-chip 5 . Typically, in clinical settings, microfluidic chips designed for synthesis are the most common ones taking their place as diagnostic tools. Though, as stated previously, material does differ for these chips, many materials popularly used are expensive and on top of that, they need expensive machinery in order to fabricate the material into the needed microfluidic chip. Therefore, most of the time, these chips are not desirable for those environments where POC is required with limited resources. 2.3 Paper-based Microfluidic Devices From this comes", "settings, microfluidic chips designed for synthesis are the most common ones taking their place as diagnostic tools. Though, as stated previously, material does differ for these chips, many materials popularly used are expensive and on top of that, they need expensive machinery in order to fabricate the material into the needed microfluidic chip. Therefore, most of the time, these chips are not desirable for those environments where POC is required with limited resources. 2.3 Paper-based Microfluidic Devices From this comes the paper-based microfluidic analytical devices. These chips carry the same characteristics as regular microfluidic devices though they are extremely more efficient for 9 their use. They are more affordable, one-time use, and fast to fabricate chips that can be used in POC settings for several applications such as clinical diagnostics, environmental monitoring, and food analysis 1 . Not only that, but they are also safe to use and for the environment as they are biodegradable and easy to recycle 8 . Again, due to their lab-on-chip characteristics with the added benefit of being significantly cheaper than the other microfluidic chips or even other more expensive machinery, paper-based microfluidic devices are preferable. With the use of paper as the main material, the chip becomes porous, has a high surface-to-volume ratio, and uses minimal volume 8 . Now because of the characteristics that come with paper, and later tamed paper, the device does not need another force aside from capillary force in order to promote fluid flow 9 . This works when a small volume of liquid is pipetted to the porous and hydrophilic section of the chip, then the adhesion forces- those present between solid and liquid interactions- help maneuver the liquid across the rest of the hydrophilic parts of the paper 9 . Paper-based microfluidic devices are exceptionally versatile from fabrication methods to application usage. Popularly, paper-based microfluidic chips are generally made through the use of filter paper, the most common being the Whatman Filter Paper 8 . The material is then used to create the channels where the sample volume will travel to and from. This can be done in two ways, either through the treatment of the paper in order to create a hydrophobic region 6 or through mechanical cutting 9 . Now there are a vast number of fabrication techniques to create these barriers and microchannels when treating the paper by creating a hydrophobic region. Each method", "common being the Whatman Filter Paper 8 . The material is then used to create the channels where the sample volume will travel to and from. This can be done in two ways, either through the treatment of the paper in order to create a hydrophobic region 6 or through mechanical cutting 9 . Now there are a vast number of fabrication techniques to create these barriers and microchannels when treating the paper by creating a hydrophobic region. Each method carries their own unique merits and demerits as well as processing machines and techniques. Therefore, the manufacturing process of each microfluidic chip is entirely different for each application, clinic, or research setting. Table 1 1 shows all the different fabrication methods that are more commonly used along with their pertaining materials and equipment. 10 Table 1. Fabrication methods, usage, materials, and equipment for paper-based microfluidic devices. 1 Similarly, detection methods vary from test-to-tests, each again having their respective pros and cons. The most commonly used detection methods are photometry, fluorescence, and mini-potentiostat 8 . In recent years, researchers have utilized smartphones as a detection method for the use of these particular biosensors. This is done so through the determination of different visual reactions caught under a smartphone camera. From the use of smartphone cameras to 11 specific device applications, detection using smartphones becomes even more accessible and acceptable in point-of-care settings. Images captured through a smartphone lens are analyzed through computer softwares such as ImageJ 13 . This software allows for the detection of different colors, contact angle, and other properties of the gathered data. Past chips have ultimately paved the way for a combination of microfluidic devices to merge. For example, digital microfluidic devices can be made using paper based material, therefore cross-linking both properties and creating a greater chip. Paper-based digital microfluidic chips essentially transport individual droplets through the chip without the fabrication of microchannels 1 . Furthermore, another form of paper-based chips are those with continuous microfluidic flow (p-CMF) 1 . These two forms of lab-on-chips have become staple for POC settings due to their combination of characteristics and properties that prove to be beneficial and effective. Previous studies have shown how microfluidic chips can be used in various areas of research. From biomedical diagnostics to environmental monitoring, these lab-on-chips have been studied for the use of detecting nucleic acids, microorganisms cells, and other pathogens 6 .", "another form of paper-based chips are those with continuous microfluidic flow (p-CMF) 1 . These two forms of lab-on-chips have become staple for POC settings due to their combination of characteristics and properties that prove to be beneficial and effective. Previous studies have shown how microfluidic chips can be used in various areas of research. From biomedical diagnostics to environmental monitoring, these lab-on-chips have been studied for the use of detecting nucleic acids, microorganisms cells, and other pathogens 6 . Great advances have been made for these devices to be made more accessible especially at the research level. Researchers have created softwares like AutoPAD 3 that aid in the design of microfluidic chips as they automate a paper-based microfluidic design through computer software to be then printed onto the desired filter paper 2 . Although profound progress has been made, there still remain gray areas that need to be addressed and improved especially at the research level. With every fabrication decision taken for the production of a microfluidic chip, something is unfortunately put at risk. For example, in order to make a specific chip or to be more cost-efficient, a researcher might choose the laser toner printing fabrication method in 12 order to create their channels. This will result in a tedious process of individually cutting each printed channel and will eventually affect the research timeline. Because of these \u201cgive-and-take\u201d areas, automation needs to be studied for the fabrication and dispensing process of microfluidic devices. This research paper aims at closing the gap between microfluidic production drawbacks and efficient use of the lab-on-chips. The cutting and adhesion of filter and membrane paper was also analyzed as a way to create an economical path for the fabrication of the chip. Finally, the introduction of an automated solution dispensing technique was studied for processing efficiency and increase in throughput. 3. Research Hypothesis and Objectives As aforementioned, paper-based microfluidic chips are ideal devices for POC and research settings due to their affordability, characteristics/ properties, and fabrication process. Nonetheless, they are not a perfect device as they have multiple fabrications and use factors that can ultimately affect a study\u2019s deliverables and results or cause a laboratory backup. Because these are one-time use devices, multiple chips need to be made simultaneously in order to be efficient, stay within the research timeline, and process a myriad of diagnostics chips. In most cases, due to the accessibility", "for POC and research settings due to their affordability, characteristics/ properties, and fabrication process. Nonetheless, they are not a perfect device as they have multiple fabrications and use factors that can ultimately affect a study\u2019s deliverables and results or cause a laboratory backup. Because these are one-time use devices, multiple chips need to be made simultaneously in order to be efficient, stay within the research timeline, and process a myriad of diagnostics chips. In most cases, due to the accessibility of the materials needed for these chips, these devices are made by hand using simple equipment. Therefore, this creates a gray area that is very time consuming and can also cause variation ultimately leading to data inconsistency. Now, this project aims at increasing efficiency, effectiveness, and consistency by addressing the core need of automation for the large-throughput fabrication and processing methods for paper-based microfluidic devices, specifically for those made in SJSU\u2019s CardioLab. Ultimately, the goal for this project was to automate a method that processes multiple data 13 samples in a time efficient manner. As well as, close the gap and mitigate the gray areas in order to lessen discrepancies in data caused by the current methodologies. Although these could be considered broad goals and objectives, they can be divided into two specific aims. 3.1 Specific Aim #1: Fabrication Automation The first part of this project is the implementation of a streamlined fabrication process that results in a high-yield chip production. This specific aim targets the initial step in CardioLab\u2019s CKD projects by automating the manufacturing methods currently being used. The process of cutting and fixing each individual chip in the fabrication procedure is all done manually. Therefore, the goal was to increase time efficiency, throughput, and consistency in the chip making process by automating what was manual labor. For this reason, my null hypothesis is that the automation displayed has no effect on fabrication duration and throughput. Now, my alternative hypothesis is that the automation being studied does have an effect on the fabrication time and the throughput. Another goal for this aim was the assessment of the existing printing process. Due to its high output of inconsistent printed microchannels, a simple comparison of the current printer with a different printer helped analyze the printing variability. In order to measure success for all the goals of this specific aim, throughput (number of chips produced per trial), consistency (%), and", "alternative hypothesis is that the automation being studied does have an effect on the fabrication time and the throughput. Another goal for this aim was the assessment of the existing printing process. Due to its high output of inconsistent printed microchannels, a simple comparison of the current printer with a different printer helped analyze the printing variability. In order to measure success for all the goals of this specific aim, throughput (number of chips produced per trial), consistency (%), and time (min) will be measured and used as data. 3.2 Specific Aim #2: Processing Automation The second part of this project is the automation of the current pipetting technique in order to dispense sample solution, blood samples in CardioLab\u2019s case, onto multiple chips in a systematic manner. This aim addresses the second most time consuming step in the CKD projects as singular, manual pipetting is the existing method for the sample dispensing. 14 Therefore, again, the goal was to create and implement an automation method for the technique that is currently being done manually. By introducing a syringe pump and positioning system, not only will throughput increase but also processing time will decrease. The measurable criteria for assessing success for this aim are trial duration (min) and inaccuracy (mm) of the dispensed droplet to the chip\u2019s dispensing well. Due to the nature of this aim, there are two separate hypotheses being tested. The first is for the analysis of trial duration and throughput where my null hypothesis is that my automation system has no effect on the overall dispensing process\u2019 duration and throughput. While my alternative hypothesis is that the automated dispensing system does display a difference in trial duration and throughput. The second null hypothesis is for the inaccuracy of the dispensing of sample droplets onto the \u00b5PAD well where the automation studied showed no difference in inaccuracy between the two tested speeds. The alternative hypothesis to this is that my automation did have an effect on the accuracy of the droplet dispensing. 4. Materials and Methods The fabrication process for paper-based microfluidic devices varies greatly depending on the equipment, methods, and applications designated for the chip. While the general methodology of this project could be implemented to similar paper-based microfluidic device studies, this research strictly follows CardioLab\u2019s predetermined device design from previous student researchers. Consequently, the presented fabrication and processing automations are exclusively targeted to increase throughput, consistency,", "that my automation did have an effect on the accuracy of the droplet dispensing. 4. Materials and Methods The fabrication process for paper-based microfluidic devices varies greatly depending on the equipment, methods, and applications designated for the chip. While the general methodology of this project could be implemented to similar paper-based microfluidic device studies, this research strictly follows CardioLab\u2019s predetermined device design from previous student researchers. Consequently, the presented fabrication and processing automations are exclusively targeted to increase throughput, consistency, and time-efficiency for the paper-based microfluidic devices created in CardioLab\u2019s CKD sister projects. Because of this, the determined \u201cbottleneck\u201d steps in the chip manufacturing process are limited to CardioLab\u2019s current procedures. It is to be noted though, the automation methods from this research have the 15 potential to be implemented to other paper-based microfluidic chips if the dimensions and methods are akin to those presented in this project or where their differences are deemed negligible. In order to address the areas in need of improvement, this research introduces an automation for the fabrication and processing of paper-based microfluidic devices created in CardioLab\u2019s CKD projects. As previously stated, this approach is split into two specific aims which reflect the two biggest bottleneck areas in the process. The fabrication procedure is the first and longest step for the CKD projects as this is when the chips are made. In this step, the most time consuming area is the cutting and assembly of each individual chip. Therefore, this research implements an automated technique using a craft cutter and spray adhesive in order to increase productivity and consistency in a time efficient manner. This step also introduces the most inconsistency due to printing patterns. In order to analyze the printing variation, a comparison between the thermal printer currently being used and a competing thermal printer is imposed. Next, the sample dispensing step is the second most tedious, lengthy process as the sample solution is manually dispensed to each individual chip using a micropipette. As to process multiple chips at once, this project displays an automated process involving a syringe pump on an XY positioning system that works simultaneously. 4.1 Materials Table 2. List of materials used and their justification/use for Specific Aim #1. Product Justification White printer paper (8.5x11in) Primary material that composes the uPads. Itari Thermal Printer Printing uPad design onto paper; used for printer consistency comparison. 16 Phemomo Thermal Printer Printing uPad", "manually dispensed to each individual chip using a micropipette. As to process multiple chips at once, this project displays an automated process involving a syringe pump on an XY positioning system that works simultaneously. 4.1 Materials Table 2. List of materials used and their justification/use for Specific Aim #1. Product Justification White printer paper (8.5x11in) Primary material that composes the uPads. Itari Thermal Printer Printing uPad design onto paper; used for printer consistency comparison. 16 Phemomo Thermal Printer Printing uPad design onto paper; used for printer consistency comparison. CAMEO Craft Cutter Used to automate the cutting process for the fabrication of uPads. Make Market Adhesive Cutting Mat Essential for the cutting process; adhesive mat fastens the paper. Elmer\u2019s Multi-Purpose Spray Adhesive Primary adhesive that brings the uPads together. Scraper Tool Facilitates effective mat cleaning. Silhouette AutoBlade (Type B) Used on the craft cutter to cut desired design/ pattern. Silhouette Studio Software Needed for the use of the CAMEO craft cutter; cutting designs are created here. ImageJ Used for pixel count for consistency purposes on the Itari and Phemomo prints. IPhone Camera Captured uPad design prints for consistency purposes. Table 3. List of the materials used and their justification/ use for Specific Aim #2. Product Justification Main Components LONGWEI Power Supply Supplies the XY system with 12V. Chemyx Inc. Model Fusion 720 syringe pump Main syringe pump machine. VEVOR SBR20-800mm linear rail (x2) Used to manufacture the XY positioning system. SBR20UU slide blocks (x4) Used to manufacture the XY positioning system. NEMA 17 Stepper Motors (x2) Used to manufacture the XY positioning system. Arduino Nano Every The brain and board for the XY positioning system, 17 A4988 Stepper Motor Drivers (x2) Used to manufacture the XY positioning system. Arduino IDE 2.3.6 Software Used to code the XY positioning system\u2019s mobility. Python Software Used to turn on and sync the syringe pump with the XY positioning system. Pinion & rack gear (3D printed) Used to manufacture the XY positioning system. Tr8x8 Lead Screw & Brass Nut Used to manufacture the XY positioning system. 8x4 uPad tray (3D printed) The tray used to carry multiple \u00b5PADs. Lopez Valve Closed Enteral Tube Essential valve used for the dispensing of the sample solution. 3mL Syringe Used with the syringe pump. Syringe tubing with Luer connectors Connected the syringe to the Lopez valve in order to dispense solution. Table 4. List of the materials used and", "printed) Used to manufacture the XY positioning system. Tr8x8 Lead Screw & Brass Nut Used to manufacture the XY positioning system. 8x4 uPad tray (3D printed) The tray used to carry multiple \u00b5PADs. Lopez Valve Closed Enteral Tube Essential valve used for the dispensing of the sample solution. 3mL Syringe Used with the syringe pump. Syringe tubing with Luer connectors Connected the syringe to the Lopez valve in order to dispense solution. Table 4. List of the materials used and their justification/ use for Specific Aim #2. Product Justification Supporting Components 5mm to 8mm Shaft Coupler Used to manufacture the XY positioning system. 3mm to 5mm Shaft Coupler Used to manufacture the XY positioning system. Dupont wires Used to manufacture the XY positioning system. Banana plug to Alligator clips wires Used to connect the XY positioning system to the 12V power supply. 470\u00b5F Capacitors (x2) Used to manufacture the XY positioning system. Top hung adjustable Used to manufacture the XY positioning system. 18 roller bracket 16 in. drawer slides Used to manufacture the XY positioning system. Everbilt Black rubber and brass hooded ball swivel stem caster Used to manufacture the XY positioning system. Velcro strip Used to manufacture the XY positioning system. Weatherproof Electrical box Used to manufacture the XY positioning system. Corner brackets Used to manufacture the XY positioning system. Pillow block bearings (x2) Used to manufacture the XY positioning system. Retort Stand with guide tube clamp Used to hold the syringe tubing vertically. Two wooden slacks Used to station the XY positioning system. 19 4.2 Specific Aim #1: Methodologies 4.2.1 Fabrication Process Figure 2. Flowchart depicting the current fabrication process. The goal for this objective is to enhance and refine the existing fabrication process that is followed in CardioLab\u2019s CKD projects which is shown in Figure 2. Currently, the chips are made using the following method; the Itari portable printer is used to print the predetermined microchannel designs onto a Whatman 4 filter paper sheet (8.5x11in.) and a plasma membrane sheet. After printing is conducted, the microchannels are then each individually, manually cut using scissors or a Westcott cutter (guillotine paper cutter). Then, the microchannels are taken into the Kerr 666 furnace to be baked at 100 \u2103 for 30 minutes and 10 minutes for the filter paper and cellulose membrane respectively. While the microchannels have been taken out of the oven and left to cool, the student", "onto a Whatman 4 filter paper sheet (8.5x11in.) and a plasma membrane sheet. After printing is conducted, the microchannels are then each individually, manually cut using scissors or a Westcott cutter (guillotine paper cutter). Then, the microchannels are taken into the Kerr 666 furnace to be baked at 100 \u2103 for 30 minutes and 10 minutes for the filter paper and cellulose membrane respectively. While the microchannels have been taken out of the oven and left to cool, the student researcher prepares very small strips of double-sided tape. This adhesive will be used in the next assembling step. It is important that the strips of tape are small enough to only fit on the printed microchannel borders as it can cause discrepancies in data if the 20 tape hovers into the solution dispensing area. The last and final step is to assemble all the pieces together. Cardiolab\u2019s paper-based microfluidic design is a 3 layered chip, which is depicted in Figure 1, where the first layer is the filter paper, middle layer is the cellulose plasma membrane, and the bottom layer is a plastic backing layer. In order to fix these layers together, the prepared small strips of adhesive are applied to the back side of each microchannel border. Afterwards, the layers are stacked in a \u201csandwich\u201d effect. This step can lead to data inconsistencies due to the manual assembly of the chips. For this reason, the student researcher needs to be as precise and careful as possible when layering the papers. By following CardioLab\u2019s CKD\u2019s methods, this process takes one researcher about 1 hour and 56 minutes to complete a single round of chip fabrication, which results in an average of 75-82 chips. The time frame can be broken down as follows: 1 minute to print the \u00b5PAD design sheet, about 1 hour and 40 minutes to cut 410 tiny strips of double sided tape- 5 strips per chip in order to secure hydrophobicity and proper chip alignment- and to assemble the \u00b5PAD design sheets, and lastly about 15 minutes to cut out each individual chip. One can easily see that the bottleneck area for the current fabrication process is the adhesive cutting, assembly, and then the individual cuts. This time neglects the paper baking procedure because this step is a requirement regardless if automation is conducted or not. Due to high printing inconsistencies, the number of serviceable microchannels is", "order to secure hydrophobicity and proper chip alignment- and to assemble the \u00b5PAD design sheets, and lastly about 15 minutes to cut out each individual chip. One can easily see that the bottleneck area for the current fabrication process is the adhesive cutting, assembly, and then the individual cuts. This time neglects the paper baking procedure because this step is a requirement regardless if automation is conducted or not. Due to high printing inconsistencies, the number of serviceable microchannels is about 75-85% per filter paper. As a result, the project timeline is affected as the throughput of one test run is significantly low for the amount of time it takes to fabricate the chips. For this reason, automation is clearly needed for this step. 21 Figure 3. Flowchart outlining this project\u2019s automated fabrication process. In order to mitigate the low throughput, inconsistencies, and lengthy process, the simple automation process depicted in Figure 3 was performed. It should be noted that although the baking procedure is presented, the step was not included when gathering the time dependent data. Again, this is because this specific step is performed regardless if an automation process is conducted or not. It is also important to know that at the time of the initial experimental trials, plasma membrane sheets were scarce and already divided between other groups therefore printer paper was used for the fabrication of the \u00b5PADs made in this study. Though, due to the nature of this project and its focus centering on automation, printer paper was solely used throughout the study due to its accessibility and functionality. As previously mentioned, CardioLab\u2019s paper-based microfluidic chips are designed in a three layered structure shown in Figure 1. In this study though, a new \u00b5PAD structure is introduced where instead of three total layers, the chips are four layers as seen in Figure 4. Instead of having to cut tiny strips of double sided tape, 22 spray adhesive was used on printer paper which will be referred to as the adhesive sheet layer. Conveniently, all papers used in this study as well as the intended filter and membrane papers are all the same size, 8.5x11inches. Because of this, precision during the assembly process is straightforward. Therefore the \u201csandwich\u201d effect still takes place where an adhesive sheet lays between the top paper and the bottom paper. Figure 4. Four-layered \u00b5PAD structure used in this study. With this", "spray adhesive was used on printer paper which will be referred to as the adhesive sheet layer. Conveniently, all papers used in this study as well as the intended filter and membrane papers are all the same size, 8.5x11inches. Because of this, precision during the assembly process is straightforward. Therefore the \u201csandwich\u201d effect still takes place where an adhesive sheet lays between the top paper and the bottom paper. Figure 4. Four-layered \u00b5PAD structure used in this study. With this being said, the fabrication process for this study is as follows; a thermal printer was used to print rows of the predetermined microchannel design onto a sheet of printer paper which is the same size as both the filter and plasma membrane paper sheets, each sheet fits 82 \u00b5PAD designs. Once the printing process was completed, the Kerr 666 furnace was used at 100 \u2103 to bake one printed sheet for 30 minutes and the other for 10 minutes. After baking and cooling, the assembling and automation process began. Before though, two cutting files were made using the craft cutter\u2019s software: the adhesive sheet file and the final cut file. The adhesive sheet cut was made by uploading the exact microchannel design file used for printing onto the Silhouette studio software where each dispensing rectangular well, in said design file, was 23 highlighted in order to be cut. After all the wells were highlighted, the file was saved for future use. Similarly, the last cut file was made by again uploading the same microchannel design file used for printing onto the Silhouette studio software. Here though, instead of highlighting the wells, each singular \u00b5PAD was highlighted to be cut and separated from the sheet assembly. Again, once all \u00b5PADs were individually highlighted, the file was saved. The Silhouette studio software was digitally connected to the CAMEO 4 craft cutter via bluetooth. Now, having created these two files and synced the craft cutter to the software, the \u00b5PAD fabrication begins. First, a blank sheet of printer paper was mounted onto a sticky craft cutting mat which was then fed onto the CAMEO 4 craft cutter. With the Silhouette studio software open, the adhesive cut file was sent to the craft cutter initiating the first cut with the following settings: plain cardstock as the material, cut action, and with the use of the autoblade. Once the first cut was done, the newly", "craft cutter to the software, the \u00b5PAD fabrication begins. First, a blank sheet of printer paper was mounted onto a sticky craft cutting mat which was then fed onto the CAMEO 4 craft cutter. With the Silhouette studio software open, the adhesive cut file was sent to the craft cutter initiating the first cut with the following settings: plain cardstock as the material, cut action, and with the use of the autoblade. Once the first cut was done, the newly cut printer paper was peeled from the sticky craft mat where a scrapper was then used to remove the cut out wells from the mat. Elmer\u2019s Spray adhesive was then applied onto both sides of the cut printer paper following the directions on the bottle. Quickly afterwards, all three sheets of paper were gathered and aligned for assembly: top \u00b5PAD sheet of paper, adhesive sheet, and bottom \u00b5PAD sheet of paper. After the sheets were assembled, they were then mounted onto the sticky craft mat and then positioned at the craft cutter\u2019s feeder. Once more, using the Silhouette studio software, the final cut file was sent to the CAMEO 4 craft cutter with the same settings as previously mentioned. Once the final cut was finished, the cut sheet assembly was peeled from the sticky mat leaving behind 82 individual chips on the mat. Again, a scraper was used to dismount the \u00b5PADs from the sticky mat. Finally, attach the chip onto a plastic backing. This same cutting and assembly process was executed for 10 experimental trials, five of which were set at a cutting speed of 10 sec while the remaining five were set at a cutting speed of 15 sec. The 24 data gathered from each trial was trial duration in minutes. Each trial was the fabrication of one full sheet of printed \u00b5PAD designs. 4.2.1.1 Data Analysis Now, in order to measure success there needs to be measurable criteria and data to be collected. Although there are various variables during this process, the factors for this design experiment are the printing, adhesive assembling, and cutting. While these independent variables are seemingly hard to calculate and measure, their quantitative effectiveness can be computed. For the fabrication process presented, the craft cutting speed (Speed 10 cm/sec vs Speed 15 cm/sec) was studied. The CAMEO craft cutting machine\u2019s cutting speed is interchangeable, therefore changing the speeds allowed for an analysis of", "criteria and data to be collected. Although there are various variables during this process, the factors for this design experiment are the printing, adhesive assembling, and cutting. While these independent variables are seemingly hard to calculate and measure, their quantitative effectiveness can be computed. For the fabrication process presented, the craft cutting speed (Speed 10 cm/sec vs Speed 15 cm/sec) was studied. The CAMEO craft cutting machine\u2019s cutting speed is interchangeable, therefore changing the speeds allowed for an analysis of cutting time efficiency and whether cutting speed had notable difference in throughput. The dependent variable for this specific aim was the trial duration in minutes per full printed sheet of \u00b5PAD design in relation to the automation implementation. By this it is meant that each trial was based on one sheet of 82 printed \u00b5PAD designs, where the automation\u2019s fabrication time was collected. A systematic comparison was made for both of the cutting speed\u2019s trail duration where ten trials were run, five for each speed therefore the sample size is 5. The null hypothesis is that the automation displayed has no effect on fabrication duration and throughput. Now, my alternative hypothesis is that the automation being studied does have an effect on the fabrication time and the throughput. The resulting data was analyzed using a two-tailed, nonparametric, Wilcoxon signed-rank test. This test is most appropriate for my data not only due to its ability to analyse small sample sizes but also because it is to be used when analyzing a paired test under different parameters. In this case, the same exact fabrication procedure was followed for both speeds therefore defining it as a paired test. The only difference is that speed is studied at two rates. Overall, it helped 25 conclude whether to reject or fail-to-reject my null hypothesis for this specific aim. For this two-tailed, Wilcoxon signed-rank test the p-value was determined in order to assess the null hypothesis\u2019 rejection or failed rejection. After this analysis, a simple direct comparison was made between the manual fabrication duration and my automated fabrication duration. This allows for an easy contrast between the two methods\u2019 results that establish which procedure is more effective. Throughout the data analysis portion of this study, Google Sheets was used as it is readily available to all SJSU students. Any relevant data or software functions can be found in the appendix in the Sheets file. 4.2.2 Printing", "rejection or failed rejection. After this analysis, a simple direct comparison was made between the manual fabrication duration and my automated fabrication duration. This allows for an easy contrast between the two methods\u2019 results that establish which procedure is more effective. Throughout the data analysis portion of this study, Google Sheets was used as it is readily available to all SJSU students. Any relevant data or software functions can be found in the appendix in the Sheets file. 4.2.2 Printing Assessment As previously mentioned, a major source of data discrepancies comes from the printing process. Therefore, in order to assess this area of need, this project compared the existing thermal printer to a competing thermal printer. Currently, CardioLab uses the Itari portable thermal printer for the microchannel printing process. This printer has a low consistency percentage resulting in about 60-70% of the printed \u00b5PADs actually being functional for both the filter and plasma membrane papers. In order to analyze whether it is strictly printer performance, the Itari printer was compared to the Phomemo thermal printer, a competing printer with raving reviews. New printer ribbons were used on the Itari printer as before this study began, the original Itari printer\u2019s ribbons malfunctioned. The same printing process described earlier was used for both printers. A picture was taken for each of the printed sheets for data analysis. 4.2.2.1 Data Analysis The objective for this aim was simply to directly compare and contrast the effectiveness and consistency of the Itari and Phomemo thermal printers. In order to do so, the main and only 26 measurement gathered for data analysis was the consistency percentage which was calculated using Formula 1 in the appendix. For this objective, a total of 5 printing trial runs for each printer were conducted therefore the sample size is 5. Each trial\u2019s consistency was measured. Determination of functional chips was made by first calculating the black pixel intensity in each individual \u00b5PAD design. By knowing the black pixel intensity, we can determine the chip\u2019s potential to be functional for the intended hydrophobicity aspect. In other words, the closer a chip\u2019s black pixel intensity is to the functioning maximum pixel intensity of that trial, the higher potential it has to become an effective hydrophobic barrier after baking. Pixel intensity count was done with the use of ImageJ\u2019s Analyze>Measure function. Under the Analyze> Set Measurements function, Area and Limit to Threshold were", "individual \u00b5PAD design. By knowing the black pixel intensity, we can determine the chip\u2019s potential to be functional for the intended hydrophobicity aspect. In other words, the closer a chip\u2019s black pixel intensity is to the functioning maximum pixel intensity of that trial, the higher potential it has to become an effective hydrophobic barrier after baking. Pixel intensity count was done with the use of ImageJ\u2019s Analyze>Measure function. Under the Analyze> Set Measurements function, Area and Limit to Threshold were checked. Each trial results\u2019 picture was converted to 8-bit. Because the Limit to Threshold function only calculates white pixels, the picture was inverted therefore technically one was determining the white pixels present but in actuality it was the black \u00b5PAD design. Afterwards, Threshold adjustment in B&W took place where the threshold was set at a range of 195 being the minimum and 235 being the maximum for all 10 images. Finally, the black pixel intensity was found by using the rectangle tool over each of the \u00b5PADs- the rectangle was sampled from the first printed \u00b5PAD for each sheat- in one printed sheet- and noting the area after using the Analyze>Measure tool. The same rectangle was used for all \u00b5PAD designs in a trial; any white pixels outside of the rectangle are presumed to be ink bleeds. It was assumed that the \u00b5PAD design remained structurally the same for all trials. After all the pixel intensities were found, a pass/fail test was conducted with a maximum and minimum number of pixels that make up a functional \u00b5PAD as the determining factor for each individual chip of a trial. The max and min were determined by assessing the outliers when compared to the average pixel count. Consistency was now easily calculated. Then, a simple 27 comparison was done to establish the effective printer. The findings for this objective justify the continued use or implantation of the thermal printer and any inconsistencies in future prints. 4.3 Specific Aim #2: Methodologies The goal for this objective was to introduce an automation technique for sample dispensing that completely replaces the existing manual delivery. Currently, CardioLab follows a simple solution delivery process using a manual pipette. One chip is processed at a time by a student researcher which can lead to a lengthy procedure when accounting that each trial can have anywhere from 70-82 chips. This highlights the need for automation, which this project targets.", "in future prints. 4.3 Specific Aim #2: Methodologies The goal for this objective was to introduce an automation technique for sample dispensing that completely replaces the existing manual delivery. Currently, CardioLab follows a simple solution delivery process using a manual pipette. One chip is processed at a time by a student researcher which can lead to a lengthy procedure when accounting that each trial can have anywhere from 70-82 chips. This highlights the need for automation, which this project targets. Figure 5. Whole Dispensing System Set-Up. So as to improve time efficiency, productivity, and increase reaction time precision this research introduces the implementation of a syringe pump system stationed and connected to an XY positioning system. This system, when turned on and Python code run, delivers a 58\u00b5L sample onto each \u00b5PAD well in the \u00b5PAD tray in a systematic manner. This aim can also be separated into two parts: the syringe pump and the XY positioning system. 28 The dispensing mechanism used for this specific aim consists of the following: Chemyx Fusion 720 syringe pump, 3mL syringe, syringe tubing with a Luer connector, Lopez enteral valve, and a retort stand with a guide tube clamp. Every piece connects respectively forming the syringe pump mechanism where the tip of the Lopez valve is stationed right above the bottom right \u00b5PAD\u2019s dispensing well on the chip tray stage. For this study, the syringe was filled with our solution of colored water. The syringe pump is operated by two programming mechanisms, the pump\u2019s own settings and the computer programming software Python. Because the goal of this study was to dispense solution onto multiple \u00b5PADs in one trial, the syringe pump was set up to follow the Multi-Step Mode where the pump took multiple pumping steps. Once the Multi-Step Mode was reached, the syringe set up was conducted to be 8.66mm since the syringe used was a standard 3mL syringe. Now, there are 32 \u00b5PADs in the chip tray therefore the pump needs to take 32 steps which was reflected and programmed in the syringe pump. The smallest volume that the syringe pump can effectively and consistently deliver is 58 \u00b5L which is perfect as it falls within the blood droplet sample range (up to 60 \u00b5L). With this being said, the syringe pump was set to consistently deliver 0.058mL in the steps interface. Delay time is necessary for this system as it", "32 \u00b5PADs in the chip tray therefore the pump needs to take 32 steps which was reflected and programmed in the syringe pump. The smallest volume that the syringe pump can effectively and consistently deliver is 58 \u00b5L which is perfect as it falls within the blood droplet sample range (up to 60 \u00b5L). With this being said, the syringe pump was set to consistently deliver 0.058mL in the steps interface. Delay time is necessary for this system as it will be working simultaneously with the XY positioning system. The delay timing will reflect the \u00b5PAD tray. This is because once the pump dispenses the sample onto the well, the XY positioning system will move horizontally or vertically. The tray is made of 8 rows and 4 columns making 32 \u00b5PAD device slots, each slot will be filled with a chip. The whole system begins with the bottom right well then moves horizontally to the right and after two more wells, the tray moves in the vertical direction downwards. It then follows a snake-like pattern upwards until it reaches the top right well. Moving in the x-axis takes the XY positioning system about 10 seconds while moving in the y-axis takes about 16 seconds. With this being said, steps 29 two, three, and four all have a 10sec delay while step five has a 16sec delay. Following this pattern, steps six, seven, and eight, have a 10sec delay and step nine has a 16sec delay and so on. Aside from pumping speed- which will be discussed later- this completes the Chemyx pump set up. The second programming set up takes place in Python. This software is only used in order to sync both the syringe pump and XY positioning system to start at the same time. The Python code triggers both the pump and the Arduino board to start at the same time once the user runs the python file. The code also lets the user know it successfully sent the start command to both systems. This is done through USB connection of both systems to the central laptop. The full Python code can be accessed in the appendix. Now, the second part of the processing automation is the XY positioning system. This system contains a significant number of main and supporting components that can be located in Table 3 and Table 4. While all parts are equally important for", "The code also lets the user know it successfully sent the start command to both systems. This is done through USB connection of both systems to the central laptop. The full Python code can be accessed in the appendix. Now, the second part of the processing automation is the XY positioning system. This system contains a significant number of main and supporting components that can be located in Table 3 and Table 4. While all parts are equally important for the function of the mechanism, the following are the primary and essential components that are the backbone of the system: the power supply, linear rails with their slide blocks, stepper motors, Arduino board, motor drivers, Tr8x8 threaded rod, and the XY stage. This system only uses two linear rails both stationed horizontally and parallel to each other where both have a slide block carrying either the x-axis or y-axis stepper motor. The system uses a pinion and rack gear connected to the x-axis stepper motor, through a coupler, to move horizontally. The y-axis stepper motor is connected to the Tr8x8 threaded rod, also via coupler. This threaded rod connects both slide blocks which ultimately enables it to move the stage vertically. Now, the stage is made up of an electrical weatherproof box with one brass nut in the front and back of the box allowing the threaded rod to then be inserted into the box via brass nuts. This allows the box to act as the XY stage, as the stepper motor spins the threaded rod, the brass nuts rotate down the rod moving the stage 30 downward. Now in order for this to occur though, drawer sliders were added vertically to the sides of the stage, connecting the slide blocks at their base. Attached to the side of the electrical box is a top hung roller bracket that is slid into one of the drawer slides. This holds the stage in the x and y axis, where instead of rotating as the threaded rod is spun, it moves downward. Due to momentum and movement hiccups, a black rubber and brass hooded ball swivel stem caster was added to the bottom of the stage in order to lessen the friction and load on the slide blocks. The uPAD tray is mounted on top of the stage via velcro. Now, for the programming portion of this system. An Arduino Nano Every", "stage in the x and y axis, where instead of rotating as the threaded rod is spun, it moves downward. Due to momentum and movement hiccups, a black rubber and brass hooded ball swivel stem caster was added to the bottom of the stage in order to lessen the friction and load on the slide blocks. The uPAD tray is mounted on top of the stage via velcro. Now, for the programming portion of this system. An Arduino Nano Every board was programmed using the Arduino IDE software. This code is essentially the \u201cbrains\u201d of the XY positioning system. This code defines all the pins connected to the board and has the primary function of enabling the motors for movement at specified timing and with specified distance. The code has the chip tray dimensions defined therefore all the dispensing wells are already located via code. As mentioned earlier, the code programs the XY positioning system to move in a serpentine manner. Once the last sample is dispensed, the system returns back to the origin. The only interchangeable part of the code is the dwell time. This is the amount of time the XY positioning system stays still before moving. This still moment is when the syringe pump is dispensing the sample solution. The full Arduino code can be accessed in the appendix. Lastly, moving onto the \u201cveins\u201d segment of this system. Tables 5 and 6 depict all the wired connections present in the XY positioning system. It is to be noted, two 470\u00b5F capacitors were used on the stepper drivers and only used 12V from the power supply. For the NEMA 17 motor wiring, colored specifications were followed via the informational card in the NEMA 17 box. Also, due to my stepper motors, M1, M2, and M3 were all set to HIGH which is a 1/16 microstepping for smoother movement. 31 Table 5. NEMA17 Motor Wiring with A4988 Motor Wire Function A4988 Pin Black A+ 1A Green A- 1B Red B+ 2A Blue B- 2B Table 6. A4988 Pinout. A4988 Pin Connected to VMOT +12V from Power supply GND (next to VMOT) Power supply ground VDD +5V from Arduino GND (near VDD) Arduino GND STEP Arduino D3 for X/ D5 for Y DIR Arduino D2 for X/ D4 for Y ENABLE Arduino D8 for X/ D9 for Y SLEEP Arduino\u2019s 5V RESET Arduino\u2019s 5V M1 Pulled on HIGH (Arduino\u2019s 5V)", "Function A4988 Pin Black A+ 1A Green A- 1B Red B+ 2A Blue B- 2B Table 6. A4988 Pinout. A4988 Pin Connected to VMOT +12V from Power supply GND (next to VMOT) Power supply ground VDD +5V from Arduino GND (near VDD) Arduino GND STEP Arduino D3 for X/ D5 for Y DIR Arduino D2 for X/ D4 for Y ENABLE Arduino D8 for X/ D9 for Y SLEEP Arduino\u2019s 5V RESET Arduino\u2019s 5V M1 Pulled on HIGH (Arduino\u2019s 5V) M2 Pulled on HIGH (Arduino\u2019s 5V) M3 Pulled on HIGH (Arduino\u2019s 5V) 4.3.1 Data Analysis As discussed earlier, in order to prove that the proposed automation system is deemed effective and strategic, there needs to be a metric of success. This specific aim\u2019s data was 32 analyzed using the nonparametric analysis; a Wilcoxon signed rank test. The decision to use this test lied on the fact that this test is used to compare paired data, meaning the exact same procedure and sample was tested twice under different conditions which is the case for these data points. A two-tailed test was used therefore \u03b1 = 0.05 where the null hypothesis being tested is that my automation system has no effect on the overall dispensing process\u2019 duration and throughput. While my alternative hypothesis is that the automated dispensing system does display a difference in trial duration and throughput. Due to the nature of this aim, the main independent variables and factors are the pipetting speeds (mL/min) and its correlating XY positioning dwell time (sec). The syringe pump was programmed to dispense at two rates: 0.4 (mL/min) and 0.6 (mL/min). Similarly, the XY positioning system\u2019s dwell time was also adjusted accordingly at two times: 12 (sec) and 9 (sec) respectively. Although there are four interchangeable variables, there are only two experimental conditions: 0.4(mL/min)+12sec speed and 0.6(mL/min)+9sec speed. This is because the dwell time has to be changed when the dispensing speed is changed, they correlate. These factors were used to gather trial duration (min) data points to be used for aim analysis. In this study, 10 total experimental trials were performed which defines a sample size of 5. The data analysis for this specific aim was also separated, meaning two separate nonparametric, two-tailed Wilcoxon signed rank tests were performed for the data collected for this aim. The second Signed rank test was done now with inaccuracy (mm) as the response. This is done", "they correlate. These factors were used to gather trial duration (min) data points to be used for aim analysis. In this study, 10 total experimental trials were performed which defines a sample size of 5. The data analysis for this specific aim was also separated, meaning two separate nonparametric, two-tailed Wilcoxon signed rank tests were performed for the data collected for this aim. The second Signed rank test was done now with inaccuracy (mm) as the response. This is done by interpreting all dispensed solution results. After each trial run, a picture was taken of the \u00b5PAD tray with the dispensed solution in each \u00b5PAD well. ImageJ was used to analyse the results, first by setting the scale in order to calculate inaccuracy in millimeters instead of pixels. To do so, the height and width of the \u00b5PAD tray was measured in pixels simply using the line 33 feature. From this, the pixel aspect ratio was calculated by dividing the width by the height. Afterwards, while still having the height line on the tray, the Analyze> Set Scale feature was used. Here, you input the known height of the tray, the pixel aspect ratio, and the new units.Now, using the Analyze>Measure tool with the centroid feature checked, the distance between the centroid of the droplet and the \u00b5PAD well was calculated but can also be done using Formula 2 in the appendix. Following this, the average and standard deviation was calculated, using Formula 3 and Formula 4, which becomes the inaccuracy of each trial. The studied null hypothesis, for the inaccuracy of the dispensing of sample droplets onto the \u00b5PAD well, was the automation studied showed no difference in accuracy between the two tested speeds. The alternative hypothesis to this is that my automation did have an effect on the accuracy of the droplet dispensing. Because inaccuracy is the distance between the centroids of the droplet and the chip well, in actuality it showcases the accuracy of the dispensing system, the shorter the distance is the more accurate the system is. The same test parameters and justifications can be used for this signed rank test except for the data being gathered and compared, in this case it is inaccuracy (mm) instead of trial duration. 4.4 Safety Laboratory safety is to be kept at the utmost priority throughout any part of this project. Albeit, the contents for this specific project do", "in actuality it showcases the accuracy of the dispensing system, the shorter the distance is the more accurate the system is. The same test parameters and justifications can be used for this signed rank test except for the data being gathered and compared, in this case it is inaccuracy (mm) instead of trial duration. 4.4 Safety Laboratory safety is to be kept at the utmost priority throughout any part of this project. Albeit, the contents for this specific project do not directly manage hazardous chemicals, other projects in and out of CardioLab do and surround the working station for this project. For this reason, general personal protective equipment (PPE) is to be worn at all times while working in the laboratory. General PPE to be worn are the following: lab coat, gloves, and safety goggles. Gloves are very important for this project due to the handling of cellulose membrane paper, which can be affected by natural oils found on the skin. All BME Lab protocols are to be 34 followed at all times to ensure safety for everyone. During the beginning stages of automation testing, blood samples will not be used. Instead, the solution will be replaced with colored water. Oven temperatures are hazardously high therefore baking steps should be exercised with great caution. Spray adhesive has a strong odor, therefore masks should always be worn. 5. Results & Discussion 5.1 Specific Aim #1: Printing Assessment The results of the printing assessment displayed that a larger consistency percentage was attained through the use on the competing thermal printer. Phomemo\u2019s prints had a consistency of above 57% while Itari\u2019s prints had a consistency of above 48%. Due to this, the Phomemo thermal printer was used for the remainder of the study. Not only were the prints significantly better than Itari\u2019s but each sheet had the full 82 chip designs fully printed out for every trial. On the other hand, Itari only had two clear printed sheet designs out of the five trials. It should be noted that after continuous use without printing breaks both thermal printers overheat and cause ink dragging toward the bottom of the sheet where the chip is no longer functional due to unclear design. Looking at the bigger picture, both thermal printers have a notably low consistency percentage across all print numbers. Neither printer passes at least the 80% threshold. This can definitely have an effect on", "had two clear printed sheet designs out of the five trials. It should be noted that after continuous use without printing breaks both thermal printers overheat and cause ink dragging toward the bottom of the sheet where the chip is no longer functional due to unclear design. Looking at the bigger picture, both thermal printers have a notably low consistency percentage across all print numbers. Neither printer passes at least the 80% threshold. This can definitely have an effect on the functionality of the chips because it leads to variance in results. Because not all printed designs have the same black pixel intensity, some will be more/less hydrophobic than others which ultimately skews the results. Table 7. Consistency Results for the \u00b5PAD design printing. Print Number Itari Consistency (%) Phomemo Consistency (%) 1 69.51 68.29 2 51.22 68.29 35 3 65.85 75.61 4 47.56 62.20 5 54.88 57.32 5.2 Specific Aim #1: Fabrication Process Table 8. Results for the fabrication process. Trial # Cutting Speed (cm/sec) Trial Duration (mim) 1 10 14:37 2 10 13:48 3 10 13:42 4 10 14:03 5 10 13:39 6 15 13:35 7 15 12:38 8 15 12:30 9 15 12:47 10 15 12:34 Noting that CardioLab\u2019s current fabrication process takes around two hours to make 82 chips, one can easily do a direct comparison and determine by looking at the results table that my automated fabrication process significantly decreases the fabrication time. With the same throughput of 82 chips per trial, my automated system took less than 15 minutes. To put it into perspective, if I were to allocate two hours to just fabricate \u00b5PADs using my automated method, then 8 full runs can be made resulting in about 656 chips per researcher. As one goes down the 36 results table, one can see that I became acclimated to the fabrication process fairly quickly. In order to further directly compare the manual fabrication process to the automated fabrication process, the time frame for each step can be broken down. Both use the same printer therefore it takes about a minute to print a sheet of 82 \u00b5PAD designs. Now, the adhesive layer cut takes about 6-7 minutes. As mentioned in methods, due to the use of spray adhesive the assembly of the \u00b5PAD sheets has to be fairly quick, in less than a minute. Finally, it takes around 4-5 minutes to perform the last", "to the automated fabrication process, the time frame for each step can be broken down. Both use the same printer therefore it takes about a minute to print a sheet of 82 \u00b5PAD designs. Now, the adhesive layer cut takes about 6-7 minutes. As mentioned in methods, due to the use of spray adhesive the assembly of the \u00b5PAD sheets has to be fairly quick, in less than a minute. Finally, it takes around 4-5 minutes to perform the last cut in order to individualize the chips from the assembled sheets. The remainder of the time can be allocated to scrapping the sticky mat or mounting the paper onto the craft mat. With this being said, one can recall the time frame of the current fabrication process mentioned above, where just cutting the adhesives and assembling took about 1 hour and 40 minutes whereas my automation takes at most 8 minutes. To add on, what manually takes 15 minutes to cut each \u00b5PAD out of the sheet, takes at most 5 minutes with my automation. Because of this it is easy to say that my automation significantly affected the fabrication process by decreasing the duration. In order to prove this statement, statistical comparison between the two cutting speeds can be used to show that the automation process does make a difference in the overall duration of the process which therefore consequently also positively affects the throughput. A systematic comparison was done to the two cutting speeds (10cm/sec and 15cm/sec) through a two-tailed Wilcoxon signed-rank test. Results of the test yielded a 15 for the positive signed rank and 0 for the negative signed rank. Following the Wilcoxon signed-ranked tests protocol, the smallest of the two becomes the test statistic (W). Using my Wilcoxon statistic (W=0), the p-value was found using formula 2 in the appendix which results in a p-value of 0.0625. So as to reduce risk of Type II error and due to the small sample size, the significance 37 threshold used was \u03b1=0.10. Because my p-value is less than the significance threshold, I can reject the null hypothesis. This means that there is statistically significant evidence that highlights an effect in trial duration when my automated system was used even with different speeds. Therefore, if one were to increase the speed even more to about 20cm/sec, we can hypothesize that the trial duration will be even less. This", "error and due to the small sample size, the significance 37 threshold used was \u03b1=0.10. Because my p-value is less than the significance threshold, I can reject the null hypothesis. This means that there is statistically significant evidence that highlights an effect in trial duration when my automated system was used even with different speeds. Therefore, if one were to increase the speed even more to about 20cm/sec, we can hypothesize that the trial duration will be even less. This though does not take account cutting accuracy, it was neglected when data was gathered because the two studied speeds had extremely similar behavior as the same cutting files were used and therefore yielded similar accuracy. Again, as mentioned earlier, this study focused on automating the fabrication process therefore, the sheet baking step was not accounted for during this data collection as it is an essential step that cannot be automated. Overall, this specific aim was successfully completed as I rejected my null hypothesis and confirmed that my automation increases time efficiency. 5.3 Specific Aim #2: Dispensing System As aforementioned, CardioLab manually pipettes each individual \u00b5PAD which can take around 13 to 14 minutes for the dispensing of solution to 32 \u00b5PADs from a researcher with minimal experience. It could be inferred that with more practice, a researcher can decrease their pipetting duration but with that comes usage error and accuracy decreases. Knowing this and looking at the results table below, one can see that my automation does perform faster than 13 minutes for the dispensing of solutions to 32 \u00b5PADs. By directly comparing the manual and automated methods, the conclusion that there is about a 1.5 to 2 minute difference between manual and automated dispensing can be made. Now, as previously stated, a researcher can become faster the more they practice. But even then, looking at the results table, when the pumping speed increased and the dwell time decreased, the trial duration significantly decreased as well. This can be proven through the systematic comparison of the two automated dispensing 38 speeds, the result of the two-tailed, Wilcoxon signed-rank test is 15 for positive signed ranks and 0 for negative signed ranks therefore my W is 0. A two-tailed test with a sample size of n=5 and an alpha of 0.05, has a critical value of 0 which ultimately results in my test statistic W equaling the critical value. With this", "trial duration significantly decreased as well. This can be proven through the systematic comparison of the two automated dispensing 38 speeds, the result of the two-tailed, Wilcoxon signed-rank test is 15 for positive signed ranks and 0 for negative signed ranks therefore my W is 0. A two-tailed test with a sample size of n=5 and an alpha of 0.05, has a critical value of 0 which ultimately results in my test statistic W equaling the critical value. With this being said, we can easily reject the null hypothesis because there is significant difference in trial duration when the pumping speed and dwell time change. Again, we can infer that if we were to increase pumping speed and decrease dwell time again, then the trial duration would decrease. Ultimately, these results successfully display the effectiveness of my automated methodologies. Table 9. Results for the Dispensing System. Trial # Pumping Speed (mL/ min) Dwell Time (sec) Trial Duration (min) Inaccuracy (mm) 1 0.4 12 11:21 1.401272771 2 0.4 12 11:04 1.532019855 3 0.4 12 11:07 0.9089077451 4 0.4 12 11:03 1.034419024 5 0.4 12 11:04 0.730466027 6 0.6 9 9:26 0.6934342481 7 0.6 9 9:18 0.6842423333 8 0.6 9 9:20 0.4705576561 9 0.6 9 9:19 0.4465967042 10 0.6 9 9:19 0.439562836 39 5.4 Specific Aim #2: Accuracy After using ImageJ to find the centroid of both the dispensed droplet and the \u00b5PAD well, the distance and angle between both centroids was determined. In most cases, the distance between both points was not outstanding. Generally, the solution droplets landed at or near the centroid of the \u00b5PAD well. This is not to say that all droplets landed in the same position because most did not, they landed around the \u00b5PAD well centroid. The greatest distance between both centroids is seen in trial 2\u2019s average inaccuracy of about 1.5mm. Taking a deeper look at all of trial 2\u2019s distances (Google Sheets is accessible in appendix B), there is an outlier of a 3.5mm difference in the x direction which ultimately affected the average distance. There can be many reasons for having one outstanding outlier. From XY positioning system movement hiccups to variance in the \u00b5PAD chip size on that tray, errors can and will occur. By looking at the results table above, the errors were mitigated along the way as there is no other significant outlier. The results both statistically and visually (result", "in appendix B), there is an outlier of a 3.5mm difference in the x direction which ultimately affected the average distance. There can be many reasons for having one outstanding outlier. From XY positioning system movement hiccups to variance in the \u00b5PAD chip size on that tray, errors can and will occur. By looking at the results table above, the errors were mitigated along the way as there is no other significant outlier. The results both statistically and visually (result images can be seen in appendix C), showcase that the dispensing automation is not precise but is highly accurate. Now, as mentioned earlier, a two-tailed Wilcoxon signed-ranked test was conducted to this set of data points. The test statistics were 15 for positive ranks and 0 for negative ranks as per Wilcoxon test statistic, W is the smaller of the two. Therefore, my W is again 0 for these data points. The critical value of a two-tailed test with a sample size of n=5 and an alpha of 0.05 is 0 which shows that my test statistic W=0. This means that I can easily reject my null hypothesis, determining that change in dispensing speed ultimately has a significant effect on the accuracy of the XY positioning and dispensing system. An interesting observation to note is that inaccuracy actually decreased with the increase of speed. If one were to theoretically compare it with manual pipetting, one can argue that accuracy can be estimated to decrease as manual pipetting speed increases whereas this automation system increased in accuracy. 40 6. Limitations & Future Work Generally, research and experimental projects have limitations and restrictions. This study is no exception, in the length of this project different limitations were observed and noted as significant. Firstly, throughout this study only printer paper was used. In other words, all experiments conducted were done only with the use of printer paper. This is to be noted because, as mentioned earlier in this paper, the \u00b5PADs used for CardioLab\u2019s CKD projects are made with filter and plasma membrane paper. It is pertinent to note that, due to the limited use of only printer paper, result variability can and may occur during the presented automated fabrication process when filter and plasma membrane paper is used for the creation of the \u00b5PADs. This is because the craft cutter software accounts for material. In order to mitigate this, as mentioned earlier,", "as mentioned earlier in this paper, the \u00b5PADs used for CardioLab\u2019s CKD projects are made with filter and plasma membrane paper. It is pertinent to note that, due to the limited use of only printer paper, result variability can and may occur during the presented automated fabrication process when filter and plasma membrane paper is used for the creation of the \u00b5PADs. This is because the craft cutter software accounts for material. In order to mitigate this, as mentioned earlier, the material chosen in the software was thicker than the layered printer paper structure during the trials. Next, due to the spray adhesive during the fabrication process, there is a time constraint. As per application directions located on the bottle, assembly should be made within 15 seconds of spraying for longevity and firmness. Now during the dispensing process, two limitations were observed. The first being that the smallest volume that the syringe pumping mechanism effectively delivered was 0.058mL. The second being that the XY positioning system stage has relative weight on it that ultimately causes momentum and movement hiccups when moving horizontally. Lastly, although mentioned throughout this paper, it is to be noted here that another significant limitation for the adaptability of the presented automations is that all experiments were conducted on CardioLab\u2019s predetermined \u00b5PAD design dimensions. Although limitations can be considered weak points, they can also be regarded as bridges for future research work. In this case, researchers can focus on refining and improving the presented automations. For example, one point of study can be following the automated fabrication process with the use of the filter and plasma membrane paper instead. Another can 41 be, subjecting blood to the syringe pump mechanism for dispensing, as originally intended, following the successful delivery of water presented in this paper. Additionally, creating an affordable syringe pump instead of the Chemyx Fusion 720 pump machine. Furthermore, future work can also be focused on increasing dispensing accuracy and fabricating a lighter XY stage that is still affordable and works well with the remaining material. Lastly, creating a bigger \u00b5PAD tray can be studied in relation to throughput. Overall, all the aforementioned limitations pave the way for future advancements and work. 7. Conclusion While \u00b5PADs are faster to fabricate than other microfluidic devices, the fabrication process can be time consuming especially if affordable material and equipment is used to manufacture them. Much like in SJSU\u2019s", "increasing dispensing accuracy and fabricating a lighter XY stage that is still affordable and works well with the remaining material. Lastly, creating a bigger \u00b5PAD tray can be studied in relation to throughput. Overall, all the aforementioned limitations pave the way for future advancements and work. 7. Conclusion While \u00b5PADs are faster to fabricate than other microfluidic devices, the fabrication process can be time consuming especially if affordable material and equipment is used to manufacture them. Much like in SJSU\u2019s CardioLab, where CKD groups manually fabricate and process significant amounts of \u00b5PADs. Though due to affordability, days need to be allocated in order to simply fabricate the chips. This highlights the need for the automatization of the fabrication and dispensing process of paper-based microfluidic devices in order to have high throughput results. This paper presented a streamlined fabrication process and an automated sample dispensing mechanism that both result in a time efficient, high throughput. The first specific aim of this study was based on the fabrication process where the use of a craft cutter, spray adhesive, and crafting software led to a significant decrease in fabrication time from around two hours to around 13 minutes for the same 82 \u00b5PAD chips. The second specific aim presented a dispensing mechanism made up of two main components: syringe pump and XY positioning system. Together, the mechanism delivered sample solutions onto 32 \u00b5PADs under 12 minutes as compared to manual pipetting taking around 13 minutes. Ultimately, it was 42 determined that both of my automation methodologies were successful in improving time efficiency by decreasing overall fabrication and processing time. 43 References 1. Boobphahom, S., M. N. Ly, V. Soum, N. Pyun, O.-S. Kwon, N. Rodthongkum, and K. Shin. Recent Advances in Microfluidic Paper-Based Analytical Devices toward High-Throughput Screening. Molecules 25:2970, 2020. 2. Chen, J. L., D. I. Njoku, C. Tang, Y. Gao, J. Chen, Y. Peng, H. Sun, G. Mao, M. Pan, and N. F. Tam. Advances in Microfluidic Paper \u2010 Based Analytical Devices (\u039c PADs): design, fabrication, and applications. Small Methods , 2024.doi:10.1002/smtd.202400155 3. DeChiara, N. S., D. J. Wilson, and C. R. Mace. An open software platform for the automated design of Paper-Based microfluidic devices. Scientific Reports 7:, 2017. 4. Edward P. Randviir, Craig E. Banks, Analytical methods for quantifying creatinine within biological media, Sensors and Actuators B: Chemical,Volume 183, 2013,Pages 239-252, ISSN 0925-4005, https://doi.org/10.1016/j.snb.2013.03.103. 5. Gimondi, S., H. Ferreira, R.", "N. F. Tam. Advances in Microfluidic Paper \u2010 Based Analytical Devices (\u039c PADs): design, fabrication, and applications. Small Methods , 2024.doi:10.1002/smtd.202400155 3. DeChiara, N. S., D. J. Wilson, and C. R. Mace. An open software platform for the automated design of Paper-Based microfluidic devices. Scientific Reports 7:, 2017. 4. Edward P. Randviir, Craig E. Banks, Analytical methods for quantifying creatinine within biological media, Sensors and Actuators B: Chemical,Volume 183, 2013,Pages 239-252, ISSN 0925-4005, https://doi.org/10.1016/j.snb.2013.03.103. 5. Gimondi, S., H. Ferreira, R. L. Reis, and N. M. Neves. Microfluidic Devices: a tool for nanoparticle synthesis and performance evaluation. ACS Nano 17:14205\u201314228, 2023. 6.. Guan, H., S. Du, B. Han, Q. Zhang, and D. Wang. A rapid and sensitive smartphone colorimetric sensor for detection of ascorbic acid in food using the nanozyme paper-based microfluidic chip. LWT 184:115043, 2023. 7. Kovac, I. Characterization of the Sensitivity and Limitations of Serum Creatinine Detection of Paper-Based Microfluidic Devices. , 2024. 8. Kulkarni, M. B., N. H. Ayachit, T. M. Aminabhavi, and B. W. Pogue. Recent advances in microfluidics-based paper analytical devices (\u00b5PADs) for biochemical sensors: From fabrication to detection techniques. Biochemical Engineering Journal 198:109027, 2023. 9. Kumar, S., P. Bhushan, A. K. Agarwal, and S. Bhattacharya. A historical perspective on paper 44 microfluidic based Point-of-Care diagnostics. In: Advanced functional materials and sensors. 2019, pp. 1\u20135.doi:10.1007/978-981-15-0489-1_1 10. Li, W., X. Ma, Y.-C. Yong, G. Liu, and Z. Yang. Review of paper-based microfluidic analytical devices for in-field testing of pathogens. Analytica Chimica Acta 1278:341614, 2023. 11. Ng, J. S., and M. Hashimoto. Fabrication of paper microfluidic devices using a toner laser printer. RSC Advances 10:29797\u201329807, 2020. 12. Rodriguez, N. M., W. S. Wong, L. Liu, R. Dewar, and C. M. Klapperich. A fully integrated paperfluidic molecular diagnostic chip for the extraction, amplification, and detection of nucleic acids from clinical samples. Lab on a Chip 16:753\u2013763, 2016. 13. Tambaru, D., R. H. Rupilu, F. Nitti, I. Gauru, and N. Suwari. Development of paper-based sensor coupled with smartphone detector for simple creatinine determination. AIP Conference Proceedings , 2017.doi:10.1063/1.4978168 14. Wang, S., L. Ge, X. Song, J. Yu, S. Ge, J. Huang, and F. Zeng. Paper-based chemiluminescence ELISA: Lab-on-paper based on chitosan modified paper device and wax-screen-printing. Biosensors and Bioelectronics 31:212\u2013218, 2011. 15. Webster, A. C., E. V. Nagler, R. L. Morton, and P. Masson. Chronic kidney disease. The Lancet 389:1238\u20131252, 2016. 16. Yuan, M., C. Li, Y. Zheng, H. Cao,", "Suwari. Development of paper-based sensor coupled with smartphone detector for simple creatinine determination. AIP Conference Proceedings , 2017.doi:10.1063/1.4978168 14. Wang, S., L. Ge, X. Song, J. Yu, S. Ge, J. Huang, and F. Zeng. Paper-based chemiluminescence ELISA: Lab-on-paper based on chitosan modified paper device and wax-screen-printing. Biosensors and Bioelectronics 31:212\u2013218, 2011. 15. Webster, A. C., E. V. Nagler, R. L. Morton, and P. Masson. Chronic kidney disease. The Lancet 389:1238\u20131252, 2016. 16. Yuan, M., C. Li, Y. Zheng, H. Cao, T. Ye, X. Wu, L. Hao, F. Yin, J. Yu, and F. Xu. A portable multi-channel fluorescent paper-based microfluidic chip based on smartphone imaging for simultaneous detection of four heavy metals. Talanta 266:125112, 2024. 45 Appendix A 1. Consistency (%) = # \ud835\udc5c\ud835\udc53 \ud835\udc53\ud835\udc62\ud835\udc5b\ud835\udc50\ud835\udc61\ud835\udc56\ud835\udc5c\ud835\udc5b\ud835\udc4e\ud835\udc59 \ud835\udc5a\ud835\udc56\ud835\udc50\ud835\udc5f\ud835\udc5c\ud835\udc50\u210e\ud835\udc4e\ud835\udc5b\ud835\udc5b\ud835\udc52\ud835\udc59\ud835\udc60 # \ud835\udc5c\ud835\udc53 \ud835\udc5a\ud835\udc56\ud835\udc50\ud835\udc5f\ud835\udc5c\ud835\udc50\u210e\ud835\udc4e\ud835\udc5b\ud835\udc5b\ud835\udc52\ud835\udc59\ud835\udc60 \ud835\udc5d\ud835\udc5f\ud835\udc56\ud835\udc5b\ud835\udc61\ud835\udc52\ud835\udc51 2. P-value= \ud835\udc52\ud835\udc65\ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc5a\ud835\udc52 \ud835\udc5d\ud835\udc4e\ud835\udc61\ud835\udc61\ud835\udc52\ud835\udc5f\ud835\udc5b\ud835\udc60 2 \ud835\udc5b 3. d= ( \u2206 \ud835\udc65 ) 2 + ( \u2206 \ud835\udc66 ) 2 4. Average= \ud835\udc60\ud835\udc62\ud835\udc5a \ud835\udc5c\ud835\udc53 \ud835\udc63\ud835\udc4e\ud835\udc59\ud835\udc62\ud835\udc52\ud835\udc60 \ud835\udc5b\ud835\udc62\ud835\udc5a\ud835\udc4f\ud835\udc52\ud835\udc5f \ud835\udc5c\ud835\udc53 \ud835\udc63\ud835\udc4e\ud835\udc59\ud835\udc62\ud835\udc52\ud835\udc60 5. Standard Deviation= \u03a3 ( \ud835\udc65\ud835\udc56 \u2212 \u00b5 ) 2 \ud835\udc5b 46 Appendix B 6. Arduino IDE Code 7. Python Code 8. Data Collection & Analysis Google Sheet 47 Appendix C 9. Picture of printed \u00b5PAD design on Itari printer. 10. Picture of printed \u00b5PAD design on Phomemo printer. 48 11. Pictures of printed \u00b5PAD tray after solution dispensing at 0.4mL/min. 12. Pictures of printed \u00b5PAD tray after solution dispensing at 0.6mL/min 49", "CharacterizationoftheSensitivityandLimitationsofSerumCreatinineDetectionof Paper-basedMicrofluidicDevices IvanaKovac TableofContents 1. Introduction\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026..3 1.1. Background\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026..3 1.2. ManagementofChronicKidneyDisease(CKD)\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u20264 1.3. DetectionofCKD\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u20265 1.4. Riskfactors\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026..6 1.5. Significance\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026.6 1.6. Earlydetectionandfunctionalitytesting\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026.8 2. LiteratureReview\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026.9 2.1. TheoryBehindtheChemistryfordetectionofCKD\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026...9 2.2. CurrentCapabilitiesforserumcreatininemeasurement\u2026\u2026\u2026\u2026\u2026\u2026\u202610 2.2.1. DetectionsystemusingtheJaffechemicalreaction 2.3. Paper-basedmicrofluidicsforserumcreatininedetection\u2026\u2026\u2026\u2026.\u2026\u202611 2.4. Paper-basedmicrofluidicswithsmartphonedetectionsystems\u2026\u2026\u2026.\u202613 2.5. Fabricationofpaper-basedmicrofluidicdevices\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026...14 2.5.1. Waxprinting 2.5.2. Thermalprinting 2.6. Feasibilityofcreatininedetectionusingpaper-basedmicrofluidic devices\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u202616 2.6.1. Theoryofuseofmicrofluidicdeviceforcreatininedetection 3. ResearchObjective\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u202617 3.1. Objectives\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026 3.2. Justifications\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026 1 4. Materials&Methods\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026.19 4.1. Materials\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u202619 4.1.1. Materialsafety 4.2. Aim1Methods:Devicefabrication\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026...20 4.2.1. Outlineoffabricationmethodusingathermalprinter 4.3. Aim2Methods:SensitivityandLimitationdetection\u2026\u2026\u2026\u2026\u2026\u2026\u2026...23 4.3.1. Outlineofsensitivityandlimitationdetectionmethods 5. Results&Discussion\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026.26 6. Conclusion\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026.28 7. FutureWork\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026...28 8. Appendix\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u202630 9. References\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026..33 2 1. Introduction 1.1. Background Chronickidneydisease(CKD)isaconditionthataffectsover10%oftheglobal populationItischaracterizedbythedeteriorationofkidneyfunctionleadingtopoorblood. 27 filtration.Asthekidney\u2019slowerinfunctionality, fluidandwastecontinuestobuildupinthe bloodstreamleadingtoavarietyofotherhealthissues.EarlyCKDshowsnosignsorsymptoms withthemainmethodofdiagnosisbeingCKDspecificbloodandurinetests.Patientswith diabetesorhypertensionareatahigherriskofdevelopingCKDwiththoseprognosesbeingthe leadingcausesofCKD.Otherriskfactorsincludeobesity, heartdisease,acutekidneyinjury (AKI),andafamilyhistoryofCKD.Notably, theincidenceandprevalenceofCKDhavebeen steadilyincreasingwithCKDbecomingthefifthleadingcauseofdeathgloballyby2040asseen inFigure1. Figure1.Epidemiologyofchronickidneydisease. 3 CKDisquicklybecomingpartofthetop5leadingcauseofdeathgloballywithacurrent rateofoneintenpersons,earlydetectionbecomescriticalinallocatingpropertreatmentfor optimaloutcomes.AlthoughCKDdoesnothaveacure,thereareseveralmethodstoslowthe progressionofthediseaseinorderforittobemanageablewithlimitedheavymedical interventions. 1.2. ManagementofCKD ThemanagementofCKDaimstoslowdiseaseprogression,controlsymptoms,andmitigate complications.ManylatestageCKDpatientsonlyhaveheavilyinvasiveproceduresavailableto themsuchasdialysisorkidneytransplants.However, conservativemanagementisgrowingin recognition.Conservativemanagementistreatmentwithoutinvasiveproceduressuchasdialysis centeredaroundlifestylemodificationsandpharmaceuticals.Ideally, conservativemanagement therapiescanbeutilizedtoslowtheprogressionofthediseaseandachievethegreatest possibilityofsurvival.Treatmentincludeschangesindietandlifestylesuchasadoptinga plant-dominant,low-proteindiet,usingpharmacologicalagentstoslowCKDprogression, reducecardiovascularrisk,andmanageuraemiaandotherassociatedsymptoms.Earlydetection andappropriateevaluationofdecreasedkidneyfunctionarecrucialinslowingtheprogressionof kidneydisease.Comprehensivemanagementstrategiesareessentialtomitigatetheimpactof CKDonpatients'qualityoflifeandreducehealthcarecostsassociatedwithend-stagerenal disease. 4 1.3. DetectionofCKD TheprogressionofCKDcanbequiteaggressive,butcanbereducedorstoppedwiththe correcttreatmentandmonitoring.TheclassificationofCKDisdoneinfivestages,thefirststage beingtheleastsevereandthefifthbeingthemostsevereasseeninFigure2. Figure2.TableshowingthestagesofCKD,theresultingeGFRlevelsandtheassociated symptoms Thehealthofone\u2019skidneysisdeterminedbytheglomerularfiltrationrate(GFR).GFRis thebloodfiltrationrateofthekidneysandiscalculatedusingage,race,sexandserumcreatinine levels.OneofthemostcommonstrategiesforapproximatingthemeasurementoftheGFRis throughestimation(eGFR)bydeterminingthecreatininecontentlevelsintheblood.Currently, creatininelevelsaremeasuredthroughalaboratorybloodandurinetestwhereurineiscollected overaperiodof24hours.Patientsthengoinforabloodtestwithresultreportingtimeaveraging 1-2days.PreviousSJSUBMEresearchhasdevelopedaviable,cheaperalternativeforthe screeningofkidneyfunctionthroughthedevelopmentofapaperbasedmicrofluidicdevice.This devicemeasuresthelevelsofserumcreatinineinnon-clinicalsettingswithrapidresults.Past 5 researchhasindicatedthatthroughtheuseofwholeblood,thedesignofthepaperbased microfluidicdeviceiscapableofproducinganeGFR.Thisprojectseekstovalidatethedesign andcharacterizethesensitivityandlimitsofserumcreatininedetectionofthedevice. 1.4. RiskFactors CKDtypicallyevolvesthroughseveralstages,beginningwithkidneydamagewithnormal functiontocompletekidneyfailure,markedbyaprogressivedeclineinglomerularfiltrationrate (GFR).CompletefailureoccurswhentheeGFRdecreasestolessthan5mL/minper1\u00b773m\u00b2 3 CKDoftenremainsasymptomaticinitsearlystages,whichcanmakeearlydiagnosis challenging.Indicationoftheprogressionofthediseaseisdetectedwhenthediseasebecomes advancedwithaeGFRoflessthan30mL/minper1\u00b773m\u00b2 . Therateatwhichfunctionalityof 3 thekidneysislostalsoheavilyvarieswithfactorssuchasetiology, exposuresandinterventions contributingtotherates.Asthediseaseadvances,patientsmayexperiencesymptomssuchas progressiveuremia,anemia,volumeoverload,electrolyteabnormalities,mineralandbone disorders,andacidemia(Ref).TherapiesforthetreatmentofCKDarechronicdialysisorkidney transplants,bothofwhicharehighlyinvasiveanddebilitatingtreatments.Oneofthekey challengesinmanagingCKDistheprogressivedeclineinkidneyfunction,whichleadstoawide rangeoftreatment-relatedchallengesandpooroutcomes. 4 1.5. Significance ChronicKidneyDisease(CKD)isaglobalpublichealthconcernwithasubstantialimpacton patientmorbidityandmortality. Itrepresentsaprogressivedeteriorationinkidneyfunctionover timewithoutacure.Asthekidneylosesitsabilitytofilterbloodcausingexcessfluidandwaste 6 toremaininthebloodstream.Thiscanleadtovariousothercomplicationssuchasheartdisease orend-stagerenaldisease.CKDisaprevalentconditionworldwide.AccordingtotheGlobal BurdenofDiseaseStudy2017,CKDranked27thintermsofglobalcausesofdisability-adjusted lifeyearwithapproximately10%ofadultsworldwideaffectedbyCKD.Notably, theincidence andprevalenceofCKDhavebeensteadilyincreasingwithCKDbecomingthefifthleading causeofdeathgloballyby2040 Severalriskfactorscontributetothedevelopmentand. 1 progressionofCKDthatincludehypertensionanddiabetes.Althoughmanystudiessupportthe associationbetweenhypertensionandCKD,thecausationofandeffectofoneontheother remainscontroversial. 2 Asubstantialportionofpatientsadmittedtointensivecareunitsworldwidehavebeen admittedduetoacutekidneyinjury(AKI).Duringthetimeframeof1989to1999,people acrossGermanyandtheUnitedKingdom35.8%of41,972individualswereloggedintothe RiyadhIntensiveCareUnitProgramdatabasewithAKI Similarly,aresearchstudypriorhad. 24 cited67%of5383ICUpatientsinNorthAmericahadAKI.Thedouble-digitpercentagetrend repeatsforNewZealand,Australia,andItaly. 23 TheconditionofAKIisdiagnosedwhenthekidneysnolongerfunctionproperly accordingtoliterature.TheNHSofEnglandstates,\u201cIfthekidneysshutdowncompletely, this mayrequiretemporarysupportfromadialysismachine,orleadtodeath,\u201d Duetothehigh. 25 risknatureofthecondition,earlydetectionisacriticalneed.AKIstartsoff inpatientswitha diagnosisofchronickidneydisease(CKD).CKDismonitoredinpatientsbycomparingserum creatininelevels Inordertocomparethesesamples,acostlyandtime-consumingprocessis. 23 putinplacethroughcollectingandanalyzingthesamplesfrompatients.Theclosestdevice 7 currentlyonthemarkettothepaperbasedmicrofluidicdeviceistheNovaBiomedical StatSensorCreatinine.TheStatsensorisahandheld,singleusebiosensorforthetestingof creatininelevelsinwholebloodinahospitalsettingthatdeliversresultsinamatterofseconds. Thisdeviceshallbecompetitivewiththisdeviceasitwillprovidepatientstheconvenienceof beingabletomonitortheircreatininelevelsathomeswithresultsinamatterofminutes.The innovationofthispaperbasedmicrofluidicdeviceshallprovideamoreeffectiveandcost efficientalternativetotraditionalcreatininelevelmonitoringforthosewithCKD. 1.6. Earlydetectionandfunctionalitytesting Accurateassessmentsandfollow-upofrenalfunctionarecrucialfortheearlydetectionand managementofCKD.Onecommonlyusedmarkeriscreatinine,awasteproductthatisfiltered bythekidneys.Asrenalfunctionalitydecreases,theabilitytofilteroutcreatininealsodecreases leavingmoreofthiswasteproductintheblood.Monitoringtheselevelscanprovidevaluable informationaboutkidneyfunctionandhelpdetectearlystagesofCKD.Regularbloodtestscan measureserumcreatininelevelswhichthencanbeusedtocalculatetheeGFRwhichisakey indicatorinkidneyfunction.Increasedaccessibilitytobloodtestingcanincreasetheabilityto monitorthedetectionandprogressionofCKD.Withregulartesting,patientsandphysicianscan makechangestotreatmentfasterleadingtomoreeffectivetreatments.Rapiddiagnosticswiththe useofmicrofluidicpoint-of-caredevices,testingcanbemademorereadilyavailabletopatients enablingthemtoachieveahighervisibilitytotheprogressionoftheirdisease.Thisliterature reviewseekstogainadeeperanalysisofthecurrentstrategiesinPOCmicrofluidicdevices, morespecifically, paper-basedmicrofluidicplatformsusedinthedetectionofcreatininelevelsin patientswithCKD. 8 2. Literaturereview 2.1. TheoryBehindtheChemistryfordetectionofCKD Creatinineisoneofthechemicalcompoundwastesourbodiesproducefromtheenergy producingprocessesinourmuscles(2).Oneoftherolesofthekidneyistobreakdowncreatinine andexcreteitthroughurine.IndividualswhoexperienceCKDorAKIhavekidneysthatarenot abletobreakdowncertaincompounds,oneofthosebeingcreatinine.Individualswithhealthy kidneysmaintaincreatininelevelsinbloodserumthatrangefrom0.59mg/dLto1.35mg/dL. IndividualsthatexperienceCKDorkidneyfailureexhibitmuchhigherlevelsofcreatinine concentrationinthebloodduetothekidneyinabilitytofilterthecreatinine.Figure3outlinesthe purposeofcreatininetestsmostimportantlyusedtodiagnoseandmonitortheprogressionofthe disease. Figure3.PurposeandimportanceofcreatininetestsinCKDpatients Thisprojectseekstocharacterizethesensitivityandlimitsofserumcreatininedetection basedontheJaffereaction.Jaffe\u2019sreactionisachemicalreactionwhichgeneratesacolorchange basedontheconcentrationofcreatininethatreactswithpicricacidinanalkalinesolution.The 9 objectiveofthisprojectistovalidatethedesigndevelopedfrompreviousSJSUBMEresearch. Thisprojectwillusetheexistingdesign,butdevelopthedesignusingadifferentfabrication method.Theuseofathermalprinterwasimplementedratherthanthewaxprinterwhichwas usedinpreviousresearchandproventobeanunreproduciblemethod.Thereplicationofall existingstepsfrompreviousresearchwasconductedtodeterminethesensitivityandlimitation ofthedevice\u2019scapabilityofdetectingserumcreatinineinwholeblood.Inpreviousstudies,there werelimitationswiththedesign,asitdidnotconsistentlyallowwholebloodsamplestotravel fromtheinputtooutputchamber. Currentresearchconsistedoftestingthecurrentdesign,then proceedbasedondataretrievedtodetermineiftheinputchamber\u2019sdimensionsneedtobe increased.Oncethatisvalidated,furtherexperimentstodeterminethesensitivityandlimitsof thepaperbasedmicrofluidicdeviceshallbeconducted. 2.2. Currentcapabilitiesforserumcreatininemeasurement Currently, establishedorganizationsuselabtestingforthedetectionofserumcreatinine. Thismethodinvolvesusingequipmenttoseparatetheserumfromwholeblood.Acuvetteis filledwithasampleandamachineprocessperformsthereading.Thismethodiscapableof processingmultiplesamplesinarun.AnothercapabilityistheNovaStatSensorCreatinine,a point-of-careanalyzerthatiscapableoftakingcreatininemeasurementsinthe\u201c27to1,056 \u03bc mol/L\u201drangeTooperatethisdevice,aproprietarybiosensorstripisloadedintoanelectronic. 5 unit,thenapatientfingertipislanced,adropofbloodisappliedontheinsertedteststrip,andthe meterdisplayscreatininereadingsinmg/dLandglomerularfiltrationrateinmL/minHowever,. 6 thisdevicecanbecostlytooperatewithteststripsbeingsoldonlineinlotsoffiftyinthe$800 pricerange,averaging$16peruseinadditiontothecostoftheelectronicmeter. Inorderto 10 measureserumcreatininewithlimitedresources,severalpreviousteamsofSJSUresearchers havedevelopedapaper-basedmicrofluidicdevice. 7 2.2.1. DetectionsystemusingtheJaffechemicalreaction Serumcreatinineproducesanorangecolorwhenreactedwithpicricacidinanalkaline mediumknownastheJaffereaction.Thedevelopmentofthecolorismeasuredaftersometime andthesaturationisusedtodeterminetheconcentrationofcreatinineOncethebloodsample. 16 wasintroducedtothereactionarea,thedevicewasheatedinordertoinduceaJaffereaction resultinginyellow-orangecompleximages.ImagesofthesereactionsweretakenbyaCMOS cameraandanalyzedforthesaturationofcolortodeterminethecreatinineconcentration.Image analyzingsoftwareusesRGBintensityvaluestodetermineaconcentrationofcreatinine.Asthe concentrationofcreatinineincreases,theRGBvaluesdecrease.Reactiontimeandreaction temperatureaffecttheintensityofthecolorwiththeintensityincreasingastimeandtemperature increaseuntilasaturationpoint.Thatpointwasfoundtobe37degreeCelsiusand5minutesof reactiontemperatureandtimerespectively. 11 2.3. Paper-basedmicrofluidicsforserumcreatininedetection Inrecentyears,therehasbeenanincreasingfocusonthedevelopmentofrapiddiagnostics withthedevelopmentofpoint-of-caremicrofluidicdevices.Thesedevicesarecapableof performinglaboratorytestinginshortertimeframeswithouttheuseoflaboratoryequipment. Point-of-caretestingreferstothetestingthatcanbeconductedoutsideofthelaboratory includingphysicianoffices,clinicsandpatienthomes.PropertiesrequiredbytheWorldHealth OrganizationofPOCtestsincludeaccuracy, sensitivity, specificity, userfriendliness,rapidand 11 robust,equipmentfreeanddeliveredatsettingsawayfromtraditionallabinfrastructureThe. 8 advantagesofPOCtestingovertraditionaltestingincluderapidresultsenablingafasterand moreeffectivetreatmentresponse,reducedhealthcarecosts,increasedaccessibility, andlesser reagentandsamplesizes Rapiddiagnostictestsrefertoimmunoassays,urinalysis,. 9 environmentalmonitoring,bioterrorism,foodsafety, andveterinarymedicine.Immunoassays leadto detectionofinfectious,respiratory, cardiovasculardiseases,oncologyandwomen\u2019s health.Urinalysisaidsinthedetectionofpregnancy, glucose,protein,ketoneleukocytes,nitrite, bilirubin,anddrugabuse.Rapidtestingextendsbeyonddiseasedetectiontoenvironmental contaminationtesting,bioterrorssuchasanthrax,foodsafetytesting Theuseofpoint-of-care. 10 testinghasenabledtheearlydetectionandcontinuousmonitoringofdiseasesleadingtoearly interventionandprevention. Thereareseveralmicrofluidicdevicesthatenablepoint-of-carediagnostictesting.These miniaturizedsystemsallowforthemanipulationofsmallvolumesoffluidswithinmicrochannels toprovidevesselsforamultitudeofreactions.Thesereactionsarethenreadbytheuseror detectionplatformresultinginadiagnosis.Microfluidicsystemsincludecentrifugal microfluidics,digitalmicrofluidics,dropletmicrofluidics,continuousflow/pneumatic microfluidicsandpaper-basedmicrofluidics.Allthesetechniquesimploredifferenttechniquesto performoperationssuchaspumping,valving,andmixingonachip. 5 Paper-basedmicrofluidics,alsoknownas\"lab-on-paper\"or\"paper-basedanalytical devices,\"havegainedprominenceaslow-cost,portable,andversatiletoolsforperforming variousanalyticalanddiagnostictasks.Thesedevicesutilizethecapillaryflowoffluidsthrough porouspapersubstrates,enablingthemanipulationanddetectionofanalytes.Paper-based microfluidicsreliesonthepassivetransportoffluidsthroughpaperchannels,eliminatingthe 12 needforexternalpumpsorpowersources.Keyprinciplesincludethewickingeffect,capillary flow, andtheuseofhydrophobicbarrierstocontrolfluidmovementPaper-basedmicrofluidic. 11 deviceshavedemonstratedremarkableutilityindiagnosingdiseases,especiallyin resource-limitedsettings.Theyareusedfordetectingvariousanalytes,includingbiomarkers, pathogens,andnucleicacids.Suchdeviceshavebeenappliedtothediagnosisofmalaria,HIV, anddenguefever, amongothers. 12 Establishedformatsofpaper-basedmicrofluidicsforPOCtestingaredipsticksandlateral flowassays.Thesearecommonlyfoundinuseforpregnancytests,glucosemotioning,infection testing,etc.Theyareequippedwithacolorimetricchartordetectionequipmenttoanalyzethe results. 2.4. Paper-basedmicrofluidicswithsmartphonedetectionsystems Researchconductedfoundthatapaper-basedchiphasthepossibilitytobeusedtogether withasmartphoneapplicationtodeterminethehumanserumcreatininelevelbasedonJaffe reactiontheory. InthisresearchthereactionzonewastreatedwithpicricacidandNaOHreagent andthendriedfor20minat35C.ToallowtheJaffereactiontooccur, thechipwasheatedfor5 minutesat37CandthecreatinineconcentrationwasderivedbyanalyzingtheRGBintensity. TheresultsfromthisstudyvalidatedthatbasedonJaffereactioncreatininemeasurementsfrom 0.2-8mg/dLcanbedetected.RGBcoloranalysismustbeanalyzedafternoshorterofa5minute durationtoallowforreactionofRandBtoappear. 13 2.5. Fabricationofpaper-basedmicrofluidicdevices Thefabricationofpaper-basedmicrofluidicdevicescanbedoneintwoorthree-dimensionto allowforthetransportationoftheliquidinboththehorizontalorverticaldirections.The 13 fabricationtechniquescanbederivedintotwocategories:physicalblockingofporesandtwo dimensionshapingandcutting. 2.5.1. Waxprinting Waxprintingisapopularandaccessibletechniqueforfabricatingpaper-basedmicrofluidic devices.Itinvolvesthedepositionofhydrophobicwaxontopapersubstratestocreatepatterns, channels,andreservoirsforfluidmanipulationandanalyticalapplications.Solidwaxisrubbed ontoapapersheetthroughascreenintheshapeofthedesiredpatternandthenthepaperis placedintotheoven.Thewaxmeltsintothepaperrevealinghydrophobicbarriersdiffusedinto thepaper. Washburn\u2019sequationcanbeusedtopredicttheflowofthemoltenwaxandthefinal widthofthehydrophobicbarriers. whereListhedistancecoveredbythewaxfront,gistheviscosity, cistheeffectivesurface tension,Distheaverageporediameterandt isthetime.Theinnerwidthofthehydrophobic channelcanbedefinedas whereWCistheinnerwidthofthehydrophobicchannel,WPistheinnerwidthofthe printedchannel,andListheadditionaldistancethatthewaxspreadsperpendiculartothelength ofthechannel.Thismodelisbasedontheassumptionsthatthedistancethewaxspreadsonto thepaterisconstant,theviscosityofthewaxdoesnotchange,theheatsourcehasaconstantand uniformtemperature,andtheamountofwaxisnotlimiting,andnaturalconvectionisnegligible. Waxprintingisalow-costsolutionhowever, itislimitinginflexibilityofthescreenpatternand lowreproducibilitybetweenbatches. 7 Anotherexampleofwaxprintingfabricationisseeninthepaperconsistsoflayeringtwo filterpapersmountedonacardboardbacking.TheupperfilterpaperwasAdvantecqualitative 14 filterpaperwithaporesizeof6micrometersandathicknessof0.2millimeters.Thelowerlayer", "Theuseofpoint-of-care. 10 testinghasenabledtheearlydetectionandcontinuousmonitoringofdiseasesleadingtoearly interventionandprevention. Thereareseveralmicrofluidicdevicesthatenablepoint-of-carediagnostictesting.These miniaturizedsystemsallowforthemanipulationofsmallvolumesoffluidswithinmicrochannels toprovidevesselsforamultitudeofreactions.Thesereactionsarethenreadbytheuseror detectionplatformresultinginadiagnosis.Microfluidicsystemsincludecentrifugal microfluidics,digitalmicrofluidics,dropletmicrofluidics,continuousflow/pneumatic microfluidicsandpaper-basedmicrofluidics.Allthesetechniquesimploredifferenttechniquesto performoperationssuchaspumping,valving,andmixingonachip. 5 Paper-basedmicrofluidics,alsoknownas\"lab-on-paper\"or\"paper-basedanalytical devices,\"havegainedprominenceaslow-cost,portable,andversatiletoolsforperforming variousanalyticalanddiagnostictasks.Thesedevicesutilizethecapillaryflowoffluidsthrough porouspapersubstrates,enablingthemanipulationanddetectionofanalytes.Paper-based microfluidicsreliesonthepassivetransportoffluidsthroughpaperchannels,eliminatingthe 12 needforexternalpumpsorpowersources.Keyprinciplesincludethewickingeffect,capillary flow, andtheuseofhydrophobicbarrierstocontrolfluidmovementPaper-basedmicrofluidic. 11 deviceshavedemonstratedremarkableutilityindiagnosingdiseases,especiallyin resource-limitedsettings.Theyareusedfordetectingvariousanalytes,includingbiomarkers, pathogens,andnucleicacids.Suchdeviceshavebeenappliedtothediagnosisofmalaria,HIV, anddenguefever, amongothers. 12 Establishedformatsofpaper-basedmicrofluidicsforPOCtestingaredipsticksandlateral flowassays.Thesearecommonlyfoundinuseforpregnancytests,glucosemotioning,infection testing,etc.Theyareequippedwithacolorimetricchartordetectionequipmenttoanalyzethe results. 2.4. Paper-basedmicrofluidicswithsmartphonedetectionsystems Researchconductedfoundthatapaper-basedchiphasthepossibilitytobeusedtogether withasmartphoneapplicationtodeterminethehumanserumcreatininelevelbasedonJaffe reactiontheory. InthisresearchthereactionzonewastreatedwithpicricacidandNaOHreagent andthendriedfor20minat35C.ToallowtheJaffereactiontooccur, thechipwasheatedfor5 minutesat37CandthecreatinineconcentrationwasderivedbyanalyzingtheRGBintensity. TheresultsfromthisstudyvalidatedthatbasedonJaffereactioncreatininemeasurementsfrom 0.2-8mg/dLcanbedetected.RGBcoloranalysismustbeanalyzedafternoshorterofa5minute durationtoallowforreactionofRandBtoappear. 13 2.5. Fabricationofpaper-basedmicrofluidicdevices Thefabricationofpaper-basedmicrofluidicdevicescanbedoneintwoorthree-dimensionto allowforthetransportationoftheliquidinboththehorizontalorverticaldirections.The 13 fabricationtechniquescanbederivedintotwocategories:physicalblockingofporesandtwo dimensionshapingandcutting. 2.5.1. Waxprinting Waxprintingisapopularandaccessibletechniqueforfabricatingpaper-basedmicrofluidic devices.Itinvolvesthedepositionofhydrophobicwaxontopapersubstratestocreatepatterns, channels,andreservoirsforfluidmanipulationandanalyticalapplications.Solidwaxisrubbed ontoapapersheetthroughascreenintheshapeofthedesiredpatternandthenthepaperis placedintotheoven.Thewaxmeltsintothepaperrevealinghydrophobicbarriersdiffusedinto thepaper. Washburn\u2019sequationcanbeusedtopredicttheflowofthemoltenwaxandthefinal widthofthehydrophobicbarriers. whereListhedistancecoveredbythewaxfront,gistheviscosity, cistheeffectivesurface tension,Distheaverageporediameterandt isthetime.Theinnerwidthofthehydrophobic channelcanbedefinedas whereWCistheinnerwidthofthehydrophobicchannel,WPistheinnerwidthofthe printedchannel,andListheadditionaldistancethatthewaxspreadsperpendiculartothelength ofthechannel.Thismodelisbasedontheassumptionsthatthedistancethewaxspreadsonto thepaterisconstant,theviscosityofthewaxdoesnotchange,theheatsourcehasaconstantand uniformtemperature,andtheamountofwaxisnotlimiting,andnaturalconvectionisnegligible. Waxprintingisalow-costsolutionhowever, itislimitinginflexibilityofthescreenpatternand lowreproducibilitybetweenbatches. 7 Anotherexampleofwaxprintingfabricationisseeninthepaperconsistsoflayeringtwo filterpapersmountedonacardboardbacking.TheupperfilterpaperwasAdvantecqualitative 14 filterpaperwithaporesizeof6micrometersandathicknessof0.2millimeters.Thelowerlayer wasabloodcellplasmaseparationmembranewithathicknessof300micrometers.Thislayer allowsfortheseparationofcellsandfibrinogenfromwholeblood.Thechanneldesignwas printedusingawaxprinter. Theupperlayercontainedthereactionanddetectionzonewhereas thelowerlayercontainedtheinletzone.Thelayerswerefabricatedusingasinglefacewax printingtechniqueandbakedat120degreesCelsiusfor120seconds.Thesecondsideofthe upperlayerwasbakedat100degreesCelsiusfor80secondstoallowforthe3-Dstructureofthe reaction/detectionzone.Thetwolayerswerebondedtogetherusing3mdouble-sidedtapeand attachedtothecardboardbacking Thismethodwaspreviouslyadoptedtogeneratesample. 14 devices,howeverinconsistenciesinprintsandthediscontinuationofthewaxprinterhas preventedtheuseofthismethodinthefuture. 2.5.2. Thermalprinting Asthemanufacturingofwaxprintershasbeendiscontinued,therehavebeenmany methodsinconsiderationforanalternativeprintingmethodforpaperbasedmicrofluidicdevices. Oneofthemethodsselectedasmostsuitableforthisobjectiveistheuseofthermaltransfer printing.Resultsfromthermaltransferprintingweresuggestedtobecomparabletothosefroma waxprinter Oneofthekeyadvantagesofusingthermaltransferprintersisthattheytendtobe. 21 easilyportableprintersthatuselowheattemperatures(<100C)totransfersolidwax-basedink fromathermalribbonontopaper Oneofthedisadvantagestotheseprintersisthatcurrently. 22 theyareonlycapab;eofsmallfabricationsonspecificsubstrateformats However, anew. 22 thermalprinterthathasbeenrecentlydesignedpresentscapabilitiesthatmakeitpossibletoprint 15 onlettersizedpaperwhichisthetypicaldimensionsforprintingmicrofluidicpaperbased devices. 26 2.6. Feasibilityofcreatininedetectionusingpaper-basedmicrofluidicdevices Therearedifferentanalyticalmethodsforquantifyingcreatininewithinbiologicalmedia incurrentusetoday. Creatinineisamoleculecreatedwithintheskeletalmusclesasaproductof workdonewhichisreflectedbasedonthestateofone\u2019srenalandmusclefunction.Thereare twoprominentchemicalmethodsusedinmedicinetothedetectionofCKD,whicharethe enzymaticreactionandJaffereaction.Bothmethodshavegreatviabilityforpotentiallybeing usedwithapaper-basedmicrofluidicdeviceduetotheirlowcreatinineconcentrationlevel detectioncapabilities.OneofthekeylimitationsofusingtheJaffereactionisthatitishighly non-specific,whichcouldresultinlargelyinaccuratedataresults.However, therearemethods thatcanbeimplementedtoallowthismethodtobemorespecific.Thechangeincolorbasedon thereactioncanbedeterminedatafixedwavelengthof510nmusingaspectrophotometerand visuallydeterminedwithanorange-redcolor TheuseofJaffereactionwiththesimultaneous. 17 useofaquantitativedeviceallowsforthedetectionofcreatininetobeaviablemethod. 2.6.1. Theoryofuseofmicrofluidicdeviceforcreatininedetection Thetheoryforthisdeviceisthatasampleofwholebloodcontainingserumcreatininewillbe placedonthesamplepad.Thissamplewilltravelthroughachannelwithpredeterminedpore size,length,andstructureTheparametersareforthelateralflowassay(LFA),aframeworkfor. 18 paper-basedmeasurementof\u201canalytesincomplexmixtures,\u201dsimilartoCOVID-19antibodytest 16 kits.Theporesallowtheanalytestopasstothedetectionzonewhilenotallowingtheredblood cellsandothernon-analytestopass. 19 3. Researchobjective 3.1. Objective Thegoalofthisprojectistodevelopapaper-basedmicrofluidicdeviceforthedetection ofcreatininewhichallowsforthepossibilitytoeliminatetheneedforcostlyandtimeconsuming testingmethodsforindividualsaffectedbyCKD.Thisresearchwillconsistofmeetingtwoaims toachievetheresearchobjective.Thefirstaimtomeetthisobjectiveistodetermineaviable fabricationmethodforthefabricationofapaperbasedmicrofluidicdeviceforthedetectionof creatinine.PreviousresearchconductedbySJSUstudentsusedawaxprinterforthefabrication ofthepaper-basedmicrofluidicdevice.However, waxprintershavebeendiscontinuedleadingto conductingresearchonwhatisthebestupdatedfabricationmethodforthedevice.Thefirst methodwewillbetestingtoobtaininsightifitisaviableoptionfortheprintingofthedeviceis athermalprinter. Thermalprintershavebeenresearchedinrecentyearsandhavebeendeemed bystudiestobeaviablealternativefortheprintingprocessofpaper-basedmicrofluidicdevices. Inthefundamentalstagesofprintingwiththeseprinters,theresultsindicatedsomedrawbacks withtheprinter\u2019sprintingcapabilities.Thegoalistoattempttoreplicatetheresultsfrom literaturetoprovideadditionalevidencethatthermalprintersareaviablealternativeprinting method.However, iftheprintingqualityfromthermalprintersusedinthisprojectdeemthe methodtobeanunusabledevicefortheproject,alternativeprintingmethodswillbeconsidered. Someoftheotheroptionstoconsiderasacontingencyplanaretheuseofinkjetprintersorwax dipping. 17 Thesecondaimtomeetthisobjectiveistocharacterizethecreatininedetection sensitivityandlimitsofthedevicewhichwillbeincorporatedintoamachinelearningmodel. ThegoalofthisaimistodeterminethesensitivityandlimitsofthedevicebasedontheJaffe reactionthroughthetestingofvariousamountsofcreatinine.Throughthecollectionofdata,the determinationofthelowerandupperlimitsofthedevice\u2019scapabilitiesforthedetectionof creatinine,whichwilllaterbeusedinamachinelearningmodelforthedevelopmentofa smartphoneapplication. Thetwoaimsforthisprojectshallbeaccomplishedtoachievetheobjectiveofthe research,whichistodevelopalowcostpaperbasedmicrofluidicdeviceforthedetectionof creatinineforindividualswithCKD. 3.2. Justification Currenttestingmethodsforkidneyfunctionareanecessityforindividualsaffectedby CKDandarequitecostlyandcanbealengthyprocesstoreceiveresults.Theresearchaimsto providealowcostmethodthatwillprovideindividualswiththeirresultsinamatterofminutes. Forindividualswiththiscondition,itisvitaltoreceiveresultsrapidlytodeterminethenextsteps toavoidkidneyfailure.Dependingontheseverityoftheindividual\u2019sconditionandwherethey resideintheworld,resourcestomonitortheirCKDmaynotbenearandcanbecostlytofind andreceive.Thepaper-basedmicrofluidicdevicefordetectionofcreatininewilleliminatethe needtotravelfarandwaitforresults,allowingforamoreefficientmethodthatwillbenefitthe livesoftheseindividuals. 4. MaterialsandMethods 4.1. Materials 18 Table1.ListofmaterialsandtheirusesItem Justification/purposeEquipmentThermalPrinter UsedinthefabricationofthepaperdevicesKerr666FurnaceUsedtobakethedevicesafterprintingCylindricalFlasksUsedtomixsolutionsWestcottCarboTitaniumCutter Usedintheassemblyofthedevicetoallowforreproducibility Micropipette UsedtotransferthesolutionsontothedeviceinameasuredandcontrolledmannerIphone14 Usedtocaptureimagesofthedevicereactionsite,likelytobeoneofthemostcommondevicesusedbythegeneralpopulationiPhonecamerastand UsedtosetthecameraatafixeddistanceandallowforrepeatabilityinimagecapturingLightbox UsedtocontrolenvironmentalfactorswhentakingimagesofthedeviceConsumablesPicricacid UsedinthecreationoftheJaffereactionHydrochloricacid UsedinthecreationoftheJaffereactionCreatinineanhydrous UsedinthecreationoftheJaffereaction Sodiumhydroxide UsedinthecreationoftheJaffereaction Whatman4filterpaper Usedtoprintthedevicedesignon PallCompanyGxmembrane Usedtoprovideplasmacellseparation Micropipettes UsedtopipettesolutionsPlasticcards UsedtoaddrigiditytothedeviceMicropipettetips Usedinconjunctionwiththepipette 4.1.1. Materialsafety PicricacidandNaOHarebothhazardoussolutionsthatneedtobehandledwithcare duringthisexperiment.Picricacidcancauseskinirritationsandrecommendedtobecautioned bywearingtheproperpersonalprotectiveequipment.TherecommendedPPEareprotective glovesandeyewearalongwiththoroughlywashinghandsafterhandling.Anyareasthathave beencontaminatedmustbewashedwithsoapandwaterincludinganyclothingandmaterials. 19 Thissolutioncanbehandledundernormalroomventilationandadditionalhandlingprecautions canbefoundintheAppendixA. NaOHmaybecorrosivetometalsandcausessevereskinburnsandeyedamage.Italso shouldbehandledwhilewearingPPEofgloves,gogglesandalabcoat.Afterhanding,hands shouldbethoroughlywashedalongwithanythingelsethathascomeintocontactwiththe solution.NaOHisthemoredangerouschemicalofthetwoandshouldbehandledwithmore precautionasoutlinedinAppendixA.Allworkwiththischemicalshouldbeconductedundera ventilationfumehood. 4.2. Aim1Methods:DeviceFabrication 4.2.1. Outlineoffabricationmethodusingathermalprinter Thegoalofthisprojectistoverifythefabricationmethodforthepaperbased microfluidicdeviceforthedetectionofcreatinineinwholebloodandthedevice\u2019ssensitivityand limits.Thefollowingprocedureallowedforthedeterminationofthebestfabricationmethodand thelimitsofthedevice.OncetheprintingmethodofusingtheItariportableprinterwas determined,theprintswereusedonthefilterpaperandplasmaseparationmembranetodevelop aprototypeofthemicrofluidicdevice.Themicrofluidicdeviceconsistsofa3layerdesign;atop filterpaperlayer, middleplasmaseparationmembranelayer, andbottomplasticcardbacking layer. ThelayerswiththeprinteddevicedesignwillbecutusingaWestcottcutterforprecise dimensionsofthedeviceandlayeredtogetherusingtape. Thedeterminationof aviablefabricationmethodforthepaperbasedmicrofluidicdevice beganbytestingtheHPRTMT800thermalprinter. Testingofthisprinter\u2019scapabilitieswas 20 conductedonWhatmanGrade4filterpaper. AftervigoroustestingwiththeHPRTMT800 printer, theresultsdidnotproveviableforourstudy. ThedevicedesignisshowninFigure4.ThisdesignwasadaptedfromapreviousSJSU undergraduateCardiolabteam.Thisdesignofthispaper-basedmicrofluidicdeviceconsistsof twolayersmountedonaplasticcardbacking.Thedimensionsofthedeviceare10mmby40mm overallwiththetopfilterlayercontaininga 2mmby2mmsamplepadand3mmby3mm reactionpad,and8mmby2mmcapillarychannelmembraneasthemiddlelayer. Oncethe designofthedevicewasuploadedandprintedbytheprinter, thesheetsofpaperunderwenta bakingprocesstoallowforthesolidinkwaxtomeltintothepaperformingbarriersaroundthe hydrophilicareasofthedesign.Thisbarrierallowsforthetransferoffluidthroughcapillary action.Thebakingperiodforthefilterpaperlayerwasconductedfor30minutesat100Cand10 minutesat100Cfortheplasmamembranewhichwasdeterminedtobethebestbaketimeand temperatureduringpreliminarytesting.Inordertodeterminetheviabilityofthehydrophilic areasofthedesign,ahydrophobicitytestwasconductedbyadd8 ofDIwatertotheinput\u00b5\ud835\udc3f zoneofthefilterpaperlayerwhereitwasnecessaryforliquidtoflowthrough.Thistest indicatedifthefilterpaperallowedliquidtoflowthroughaswellasiftheliquidstayedinthe designatedarea,thereforeindicatingwhetheritisaviableprintornot.Ifthewatertravels outsideofthedevicedesignordoesnotflowthroughthefilterpaperindesignatedareas,it wouldbedeterminedthatthehydrophilicareasarenotfunctional.Theexpectationwastoseea differenceintheviabilityofthebarriersandthehydrophilicareasbasedupontheprintingand bakingmethod.Futureresearchshouldconsideramorethoroughtestusingthecontactangletest totestdropletsofwateronthedevicedesigntodeterminewhetherthedesigncanbeconsidered hydrophobic.Anothertesttoconsiderforfutureresearchisamatlabcodetodeterminewhether 21 animageofthebakedfilterpaperdesignandmembranelayerdesignare\u201cviableprints\u201dforthe continuationoftesting.Inthecasethatthethermalprinterisunabletoprovideaviabledevice, alternateprintingmethodswereconsideredsuchasinkjetprinterandwaxdipping.Onceenough datahadbeencollectedtoestablishabaselineprocedureforviableprints,preparationofthe devicebeganforaddingchemicalsolution.ThiswasduetotheItraportableproducingconsistent highqualityprintsonboththefilterpaperandplasmamembrane. Figure4.DesignofdeviceadaptedfrompreviousSJSUCardioLabteam 22 Figure5.FilterpaperprintwithHPRTMT800 Figure6.LeftprintwithItriaprinter, rightwithHPRTMT800 4.2.2. Alternatefabricationmethods Thetwoalternativemethodssuggestedtomoveforwardareinkjetprintingorwax dipping.Fortheprocessofinkjetprinting,theuseofacommercialinkjetprinterisapossible alternativetoprintthedesignonafullyhydrophobizedfilterpaper. Toallowthefilterpaperto becomehydrophobicforthisprocess,itwillbetreatedwithpolystyrenesolution.Afterprinting ofthedesigniscomplete,toluenesolutionwillbeusedtoremovethepolystyrenesolutionwhere thedesignmustbehydrophobic.Thesecondalternativemethodistheuseofwaxdipping,which willbeperformedbycreatingmoldsforthehydrophilicareasoftheensignusingironanda magnet.Thefilterpaperwiththeironandmagnetattachedwillthenbedippedintomeltedwax 23 andallowedtodryforapproximately1minutetoproducethedevicedesign.Ahydrophobicity testwillbeusedtodeterminetheviabilityofthesedeviceprintsasusedintheoriginalprinting method. 4.3. Aim2Methods:Sensitivityandlimitationdetection 4.3.1. Outlineofsensitivityandlimitationdetectionmethods Figure7:High-levelprocedureoftestingthesensitivityandlimitationsofthedevice Thesecondportionofthisprojectistodeterminethesensitivityandlimitsofthedevice basedonthecolorimetricJaffereaction.Tostartthetesting,apicricacidandsodiumhydroxide (NaOH)solutionwascreated.Tocreatethissolution,anSOPwascompletedandsubmittedto theuniversity. Thissolutionwasappliedtothereactionzoneofthedeviceandallowedto completelydryovernightbeforeaddingothersolutions.Followingthisstep,9solutionsof 24 variousamountsofcreatininewith0.1MHClwerecreatedtobetested.Arangebetween0.0135 g/mLto0.00007421g/mLofcreatinineforthetestingsolutionsbasedonknowledgeofthe effectsoflowesttomostseverereactionsofcreatininelevelswereused.Solutionswerecreated withthefollowingcreatininelevels:threewithbelownormalcreatininelevels(0.00742 mg/dL,0.275936mg/dL,0.54452mg/dL),threewithnormalcreatininelevels(0.812968 mg/dL,1.081484mg/dL,1.35mg/dL),and3withabovenormalcreatininelevels(3.9mg/dL,9.9 mg/dL,15.9mg/dL). Thereagentsneededforthedeviceshallbecreatedwith2MNaOHand0.04Mpicric acidfromstocksolution.Thedilutionofpicricacidwasperformedbytransferring70.5mLof 1.3%stockpicricacidtoa100mLErlenmeyerflaskwith29.5mLdistilledwaterintheflaskto obtain0.04MPicricacid.TheshallbestoredinthePIVlabacidcabinet.Tocreatethesolution of2MNaOH,approximately7.9gofsodiumhydroxidepelletswasmeasuredonasmallglass traywhichwasmixedina100mLErlenmeyerflasktocreatethe2Msodiumhydroxidesolution. Heatisreleasedwhenthereactionoccurs,soitisrecommendedtoplacethe100mLErlenmeyer flaskwithapproximately50mLofdistilledwaterinitinanicebathorbeakerofcoldwateras weslowlyaddthesodiumhydroxidepellets.Havingonly50mLofdistilledwaterintheflaskas thepelletsareaddedallowsforaneasiermixingprocess,oncethepelletsweredissolved, distilledwaterwasaddedtothe100mLline.Oncethetworeagentswerecreated,a50:50 solutionof2MNaOHand0.4Mpicricacidsolutionwascreatedusing2.5 ofeachreagent.10\u00b5\ud835\udc3f ofthe50:50solutionof2MNaOHand0.4Mpicricacidontothereactionzoneofthefilter\u00b5\ud835\udc3f paperlayerofthedeviceandallowittodryovernight.Thewasdonebyallowingthefilterpaper piecestodryoverameshwiretray. Then8 ofthepredeterminedcreatinineconcentration\u00b5\ud835\udc3f solutionswasdispensedusingamicropipetteonthereactionpadtodeterminethevisualcolor 25 changeofthereaction.Photographsofthereactionweretakenforeachconcentrationat0,5,10, and15minutestodeterminecolorstabilizationandthelimitsofthecolorchangebasedonthe presentcreatininelevels.Oncethepreliminarytestswerecompletedbasedonlyonthecolor changelimitsofthedevice,testingwasthenperformedbyplacingthecreatininesolutiondroplet ontheinputzoneofourdeviceanddetermineifthedeviceallowsforthecreatininetotravelto thereactionzoneacrosstheplasmamembranetoreactonthereactionzone.Thephotographson thereactionweretakenwithaniPhone14camerainalightingtent.Oncetheimagesweretaken, theywereuploadedandanalyzedinImageJsoftware.TheimagesproducedanRGBoutputin ImageJandhadaconsistentRGBoutputcorrespondingtothereactionduetocreatininelevels. TheRGBresultsproducedbyimageJcorrespondtoJaffe\u2019sreactionwherecreatininebindswith picricacidtoproduceanorangetoredcolorcomplex.FollowingthecollectionofRGBvalues foreachimage,theresultswerethenanalyzedfurtherinminitabbyperformingarepeated measures2wayanovatestoneachcolorchannelasthedependentvariable,andtimeand concentrationastheindependentvariables.Thesetestsallowedfortheresultstodemonstrate whichfactorshadasignificanteffectonthecolorchangeduetotheJaffereaction.Oncethe resultswereunderstoodtobesignificant,aposthoctest,Tukey, wasperformedtodetermine whenthelargesteffectduetotimeandconcentrationtookplaceforeachcolor. Tocompletethe analysisofthesensitivityandlimitsofthedevice,aregressionplotwascreatedtovisually identifythetypeofcurvedemonstratedbythedataandwhethertheoutlierswerevery significant. 5. Results&Discussion 5.1. FabricationMethod 26 TheresultsfromtestingoftheHPRTMT800indicatedlimitationsasroughly70%ofthe printsweredeemedunusable.TestingwiththeHPRTMT800printerconcludedaftertesting differentinkresolutionandtypesofpaper, whichallprovidedthesameresults.Thisledtofurther researchintoslightlyhigherqualitythermalprintersonthemarket.ThisledtofindingtheItria portableprinter. OnceselectingtheItriaprinter, printingresultsofboththefilterpaperlayerand membranelayerdeemedsuccessful.Althoughthisprinterproducedmoreproficientresults,there werestillabout10-15%unusableprints.Theseresultswerefoundbyperformingabasicwater droplettestwith8 ofDIwateronafullybakedfilterpaperlayersinputzone.After5minutes\u00b5\ud835\udc3f imagesweretakentoidentifyifthedropletwentthroughthefilterpaperasexpectedandifit movedoutsidethebordersofthedevicedesign.Duetotimeconstraints,onlyabasic hydrophobicitytestwasconductedandverificationwasconsideredcompletewiththedevice properlyfunctioningduringtesting.Futureresearchwiththisprojectshouldperformacontact angletesttoprovethatthedevicedesignitselfishydrophobicandisfunctionalforitsintended use. AfterthoroughtestingwiththeItriaprinterandcomparingtheresultsfrompreviousstudies performed,thisprintingmethodcanbedeemedthemostsuitablefor 5.2. CharacterizationoftheCreatinineDetectionSensitivity&Limits Thedeterminationofthesensitivityandlimitsofthedevice\u2019sabilitytodetectcreatinine", "Thissolutionwasappliedtothereactionzoneofthedeviceandallowedto completelydryovernightbeforeaddingothersolutions.Followingthisstep,9solutionsof 24 variousamountsofcreatininewith0.1MHClwerecreatedtobetested.Arangebetween0.0135 g/mLto0.00007421g/mLofcreatinineforthetestingsolutionsbasedonknowledgeofthe effectsoflowesttomostseverereactionsofcreatininelevelswereused.Solutionswerecreated withthefollowingcreatininelevels:threewithbelownormalcreatininelevels(0.00742 mg/dL,0.275936mg/dL,0.54452mg/dL),threewithnormalcreatininelevels(0.812968 mg/dL,1.081484mg/dL,1.35mg/dL),and3withabovenormalcreatininelevels(3.9mg/dL,9.9 mg/dL,15.9mg/dL). Thereagentsneededforthedeviceshallbecreatedwith2MNaOHand0.04Mpicric acidfromstocksolution.Thedilutionofpicricacidwasperformedbytransferring70.5mLof 1.3%stockpicricacidtoa100mLErlenmeyerflaskwith29.5mLdistilledwaterintheflaskto obtain0.04MPicricacid.TheshallbestoredinthePIVlabacidcabinet.Tocreatethesolution of2MNaOH,approximately7.9gofsodiumhydroxidepelletswasmeasuredonasmallglass traywhichwasmixedina100mLErlenmeyerflasktocreatethe2Msodiumhydroxidesolution. Heatisreleasedwhenthereactionoccurs,soitisrecommendedtoplacethe100mLErlenmeyer flaskwithapproximately50mLofdistilledwaterinitinanicebathorbeakerofcoldwateras weslowlyaddthesodiumhydroxidepellets.Havingonly50mLofdistilledwaterintheflaskas thepelletsareaddedallowsforaneasiermixingprocess,oncethepelletsweredissolved, distilledwaterwasaddedtothe100mLline.Oncethetworeagentswerecreated,a50:50 solutionof2MNaOHand0.4Mpicricacidsolutionwascreatedusing2.5 ofeachreagent.10\u00b5\ud835\udc3f ofthe50:50solutionof2MNaOHand0.4Mpicricacidontothereactionzoneofthefilter\u00b5\ud835\udc3f paperlayerofthedeviceandallowittodryovernight.Thewasdonebyallowingthefilterpaper piecestodryoverameshwiretray. Then8 ofthepredeterminedcreatinineconcentration\u00b5\ud835\udc3f solutionswasdispensedusingamicropipetteonthereactionpadtodeterminethevisualcolor 25 changeofthereaction.Photographsofthereactionweretakenforeachconcentrationat0,5,10, and15minutestodeterminecolorstabilizationandthelimitsofthecolorchangebasedonthe presentcreatininelevels.Oncethepreliminarytestswerecompletedbasedonlyonthecolor changelimitsofthedevice,testingwasthenperformedbyplacingthecreatininesolutiondroplet ontheinputzoneofourdeviceanddetermineifthedeviceallowsforthecreatininetotravelto thereactionzoneacrosstheplasmamembranetoreactonthereactionzone.Thephotographson thereactionweretakenwithaniPhone14camerainalightingtent.Oncetheimagesweretaken, theywereuploadedandanalyzedinImageJsoftware.TheimagesproducedanRGBoutputin ImageJandhadaconsistentRGBoutputcorrespondingtothereactionduetocreatininelevels. TheRGBresultsproducedbyimageJcorrespondtoJaffe\u2019sreactionwherecreatininebindswith picricacidtoproduceanorangetoredcolorcomplex.FollowingthecollectionofRGBvalues foreachimage,theresultswerethenanalyzedfurtherinminitabbyperformingarepeated measures2wayanovatestoneachcolorchannelasthedependentvariable,andtimeand concentrationastheindependentvariables.Thesetestsallowedfortheresultstodemonstrate whichfactorshadasignificanteffectonthecolorchangeduetotheJaffereaction.Oncethe resultswereunderstoodtobesignificant,aposthoctest,Tukey, wasperformedtodetermine whenthelargesteffectduetotimeandconcentrationtookplaceforeachcolor. Tocompletethe analysisofthesensitivityandlimitsofthedevice,aregressionplotwascreatedtovisually identifythetypeofcurvedemonstratedbythedataandwhethertheoutlierswerevery significant. 5. Results&Discussion 5.1. FabricationMethod 26 TheresultsfromtestingoftheHPRTMT800indicatedlimitationsasroughly70%ofthe printsweredeemedunusable.TestingwiththeHPRTMT800printerconcludedaftertesting differentinkresolutionandtypesofpaper, whichallprovidedthesameresults.Thisledtofurther researchintoslightlyhigherqualitythermalprintersonthemarket.ThisledtofindingtheItria portableprinter. OnceselectingtheItriaprinter, printingresultsofboththefilterpaperlayerand membranelayerdeemedsuccessful.Althoughthisprinterproducedmoreproficientresults,there werestillabout10-15%unusableprints.Theseresultswerefoundbyperformingabasicwater droplettestwith8 ofDIwateronafullybakedfilterpaperlayersinputzone.After5minutes\u00b5\ud835\udc3f imagesweretakentoidentifyifthedropletwentthroughthefilterpaperasexpectedandifit movedoutsidethebordersofthedevicedesign.Duetotimeconstraints,onlyabasic hydrophobicitytestwasconductedandverificationwasconsideredcompletewiththedevice properlyfunctioningduringtesting.Futureresearchwiththisprojectshouldperformacontact angletesttoprovethatthedevicedesignitselfishydrophobicandisfunctionalforitsintended use. AfterthoroughtestingwiththeItriaprinterandcomparingtheresultsfrompreviousstudies performed,thisprintingmethodcanbedeemedthemostsuitablefor 5.2. CharacterizationoftheCreatinineDetectionSensitivity&Limits Thedeterminationofthesensitivityandlimitsofthedevice\u2019sabilitytodetectcreatinine wasperformedbygenerating748samplesfromverylowtohighconcentrationsofcreatininein 0.1Mhydrochloricacid.14samplesweretestedat9distincttestingpointsbetween 0.00007421mg/mlto0.219mg/ml.Outofthe9testingpoints,therewerethreewithbelow normalcreatininelevels(0.00742mg/dL,0.275936mg/dL,0.54452mg/dL),threewithnormal creatininelevels(0.812968mg/dL,1.081484mg/dL,1.35mg/dL),and3withabovenormal creatininelevels(3.9mg/dL,9.9mg/dL,15.9mg/dL).Resultswerecollectedbydispensing8\u00b5\ud835\udc3f 27 ofthedesignatedcreatinineconcentrationandplacedontheinputpad.Resultswererecordedby takingimageswithaniphone14cameraat0,5,10,and15minutestoprovidedatatobe analyzedwithImageJsoftware.TheImageJsoftwarewasusedtocollecttheRGBdatafromthe reactionzoneofeachdeviceimagetodeterminewhetherRed,Green,orBluevalueisaffected mostbytheconcentrationofcreatinineinasampleandtime.Thesevalueswerecollectedforall imagescollectedandorganizedtoperformarepeatedmeasuretwowayanovatestforeachcolor channelwithtimeandconcentrationastheindependentvariables.The2wayanovatestwas conductedwiththethreecolorchannelsasthedependentvariableandthe4timepointsand9 concentrationsastheindependentvariables,whichcanbeidentifiedinthetablesinAppendixB. Fromthesestatisticaltests,itwasconcludedthatbothtimeandconcentrationhadasignificant impactonthegreencolorchannel,whileconcentrationonlyhadasignificantimpactonthered andbluechannel.Oncetheresultswereconsideredstatisticallysignificant,aposthocTukeytest wasperformedoneachofthecolorchannelsfortheirsignificantvariablesandresultsfromthe testcanbefoundinAppendixC.TheTukeytestforthegreencolorchanneldeterminedthat colorchangestartedtakingsignificanteffectnoearlierthan5minutesafterimagewastakenand continuedtochangeuntilthe15minutemark.Thisindicatesthatthelowerlimitofdetectionhas beenidentified,butfurthertestingneedstoextendtheupperlimittoallowverificationofthe upperlimitwithclosertimepointsbetweentheupperandlowerlimittoidentifythemost accuratelimitsofdetection.Forthegreencolorchannel,whenconsideringtheeffectofthe amountofcreatininehadontheresults,theearliestsignificantresultswerebetween0.0074 mg/dland1.08mg/dl.Thepointsthatjumpedfromundernormaltonormalandnormaltoabove normalcreatininelevelsalsodemonstratedsignificantresultswhichwasexpectedfromthe resultsoftheTukeytest.Ascatterplotforthegreencolorvaluechannelvaluevs.concentration 28 foundinAppendixDsupportstheresultsoftheTukeytest.Theseresultsdemonstratethe smalleststatisticallysignificantchangeinconcentrationleadingtoabetterunderstandingofthe sensitivityofthisdevice.Boththeredandbluecolorchannelsonlydemonstratedsignificant changewiththeconcentrationvariablefromtherepeatedmeasuretwowayanovatest.Thisled toperformingtheTukeytestonlyonconcentrationforthesetwocolorchannels.Theresultsof theTukeytestfoundinAppendixCforboththeredchannelandbluechannelindicatedthat therewassignificantcolorchangeassmallandearlyasbetween0.0074mg/dland0.2759mg/dl andtherecontinuedtobesignificantcolorchangeuntilthelast2concentrationstested.This identifiesaneedtoexpandtherangeofconcentrationstestedtobeabletoverifythesensitivity ofthedevice.Futureanalysisshouldconsiderusingrepeatedmeasurethreewayanovatestrather thantwowayanovatoreducethechanceoferroraseachpairwisehasa5%chanceoferror. Furtheranalysiswasperformedbycreatingahistogramoftheaveragevaluesofthecolor channelsvstheconcentrationwitherrorbarscalculatedbythestandarddeviation.Theseresults indicatedaconsistentstandarddeviationamongallresultsthatwererelativelylowwhichallows thedatatobedeterminedashavingminimalerror. Thereweresomeoutlierswhichcouldhave occurredduetoerrorsduringtesting.Thisallowedforabetterdeterminationofthesensitivityof thedeviceandfurtherwithalinearregressionplot.Thelinearregressionplotdemonstrateda smallslopeasexpected,howeveralargerslopewasexpectedandcanbeexpectedduringfurther testingofconcentrationsthatarecloserinproximitytooneanother. Preliminarytestingbeganbytestingcompletedeviceprototypeswithwholeporcine blood.Thisoccurredduetoearlyonduringtesting,itwasindicatedthateitherthepicricacidor hydrochloricacidbeingusedhadexpiredorbeenstoredincorrectlyasitdrieddownclearrather thanapaleyellowcolor. Resultsfromwholeporcinebloodtestingalignedcloselytoliteratureas 29 therewerenoredbloodcellsvisibleonthereactionpadafter15and30minutes.Duetolimited timeframeandlackofmaterials,visualanalysiswasonlytakenintoconsiderationforthe porcineblood.FuturetestingshouldconsideranalyzingporcinebloodtestswithimageJ. Theoretically, accordingtoliterature,thedetectionlimitsofthedeviceusingwholebloodshould beveryclosetothevaluesfound. 6. Conclusion Theresultscollectedfromthisresearchmettheaimsoftheexperimentprovingthata paper-basedmicrofluidicdeviceiscapableofdetectingthelevelofkidneyfunctionbasedon one\u2019screatininelevels.TheresultsofthisstudydeterminedthattheuseoftheItriaportable printerisaviableprintingmethodforthedeterminationofapaperbasedmicrofluidicdevicefor thedetectionofcreatinine.Theexperimentsalsocharacterizedthelowestpointthedeviceisable todetectcreatinineandsubsequentcreatininelevelswereidentifiedbasedoncolorimetry. Althoughtheseresultsidentifiedthesensitivityandlimitsofthedevice,thereismuchmore imageanalysistestingthatmustoccurtodevelopastrongcharacterizationofthesensitivityand limits. 7. FutureWork Thelongtermgoalofthisprojectistodevelopasmartphonediagnostictoolforat-home testingofcreatininelevelsusingapatient\u2019sbloodsample.Thedevicecoupledwiththe smartphoneappwillbeabletodeterminethelevelofcreatinineinthebloodbycapturingan imageofthereactionsiteandcomparingittoknowndatatodeterminealevel.Thisprojectaims toprovidethedataneededinordertodeveloptheAIimageanalysistechnology. Byproviding colorimetricdatabasedonalargevarietyofcreatinineconcentrations,developerscanfurtherthe 30 applicationofthisprojecttoprovideafullpoint-of-caresystemtopatientswithchronickidney disease. APPENDIXA: Materialsafetydatasheet Picricacid: \u25cf Avoidbreathingdust,fumesormist. \u25cf Contaminatedworkclothingmustnotbeallowedoutoftheworkplace. \u25cf Wearprotectiveglovesandeyeprotection. \u25cf IFONSKIN:Washwithplentyofsoapandwater. \u25cf Specifictreatment(Washareasofcontactwithwater). \u25cf Ifskinirritationoccurs:Getmedicalattention. \u25cf Washcontaminatedclothingbeforereuse. \u25cf Disposeofcontentsinaccordancewithlocal,state,federalandinternational regulations Precautionsinthehandlingandstorageofthesolutionareasfollows: \u25cf usinggloves \u25cf washinghandsthoroughlyafterhandling, \u25cf Avoidcontactwitheyesandskin. \u25cf Protectfromfreezingandphysicaldamage. \u25cf Donotallowthismaterialtodryout. \u25cf Donotletdrypicricacid(crystals)formincontaineroronthecapthreadsofcontainer. \u25cf Keepawayfromheat. \u25cf Keepawayfromsourcesofignition. 31 \u25cf Keepawayfromdirectsunlightorstrongincandescentlight. \u25cf Groundallequipmentcontainingmaterial. \u25cf Emptycontainersmaycontainhazardousresidueandposeafirerisk.Donotingest.Do notbreathedust. \u25cf Takeprecautionarymeasuresagainstelectrostaticdischarges. \u25cf Avoidshockandfriction. NaOH: \u25cf Keeponlyintheoriginalcontainer. \u25cf Donotbreathedust. \u25cf Washskinthoroughlyafterhandling. \u25cf Avoidreleasetotheenvironment. \u25cf Wearprotectivegloves/protectiveclothing/eyeprotection/faceprotection. \u25cf IFSWALLOWED:Rinsemouth.DoNOTinducevomiting. \u25cf IFONSKIN(orhair):Takeoff immediatelyallcontaminatedclothing.Rinseskinwith water/shower. \u25cf IFINHALED:Removepersontofreshairandkeepcomfortableforbreathing. ImmediatelycallaPOISONCENTER/doctor. \u25cf IFINEYES:Rinsecautiouslywithwaterforseveralminutes.Removecontactlenses,if presentandeasytodo.Continuerinsing.ImmediatelycallaPOISONCENTER/doctor. P363Washcontaminatedclothingbeforereuse. \u25cf Absorbspillagetopreventmaterialdamage. \u25cf Storelockedup. 32 \u25cf Storeinacorrosiveresistantcontainerwitharesistantinnerliner. \u25cf Disposeofcontents/containertoanapprovedwastedisposalplant. ThehandlingandstorageofNaOHIsasfollows: \u25cf Absorbspillagetopreventmaterialdamageduetocorrosivenesstometal. \u25cf Avoidcontactwitheyes,skin,andclothing. \u25cf Washhandsafterhandling. \u25cf Donotmixwithacids. \u25cf Followgoodhygieneprocedureswhenhandlingchemicalmaterials. \u25cf Useonlyinwellventilatedareas. APPENDIXB: 2wayAnovatestforredcolor 2wayAnovatestforgreencolor: 33 2wayAnovatestforbluecolor 34 APPENDIXC: TukeyPairwiseComparison(red):concentration TukeyPairwiseComparison(green):concentration 35 TukeyPairwiseComparison(green):time TukeyPairwiseComparison(blue):concentration 36 AppendixD: 37 References 1. ForemanKJ,MarquezN,DolgertA,etal.Forecastinglifeexpectancy, yearsoflifelost,andall-causeandcause-specificmortalityfor250causesofdeath:referenceandalternativescenariosfor2016\u201340for195countriesandterritories.Lancet2018;392:2052\u201390.2. KuE,LeeBJ,WeiJ,WeirMR.HypertensioninCKD:corecurriculum2019.AmJKidneyDis2019;74:120\u201331.3. Pfister, M.,Nolin,TD.,&Arya,V. (2012,January1).OptimizingDrugDevelopmentandUseinPatientsWithKidneyDisease:Opportunities,Innovations,andChallenges. https://scite.ai/reports/10.1177/0091270011415414Pfister,M.,Nolin,TD.,&Arya,V.(2012,January1).OptimizingDrugDevelopmentandUseinPatientsWithKidneyDisease:Opportunities,Innovations,andChallenges.https://scite.ai/reports/10.1177/00912700114154144. Kosack,C.S.,W. deKieviet,K.Bayrak,A.Milovic,andA.L.Page.EvaluationoftheNovaStatSensor\u00aeXpress(TM)Creatininepoint-of-carehandheldanalyzer. PLoSOne10:e0122433,2015.5. StatSensor\u00aeandStatSensorXpress\u00aeCreatinineandeGFRMetersat<https://novabiomedical.com/statstrip-creatinine/index.php>6. Ostermann,M.,andR.W. S.Chang.AcutekidneyinjuryintheintensivecareunitaccordingtoRIFLE*.Crit.CareMed.35:1837,2007.7. Azhar, M.,&Dendukuri,D.(2017).Microfluidicplatformsforpointofcare(POC)medicaldiagnostics.MedicalBiosensorsforPointofCare(POC)Applications, 255\u2013273.https://doi.org/10.1016/b978-0-08-100072-4.00011-38. Poon,T. C.W. (2013).Opportunitiesandlimitationsofcurrentpoint-of-caretestinginliverdisease.Gastroenterology, Research,6(2),39-469. Shi,Z.,Lu,Y.,&Yu,L.(2017).Microfluidicpaper-basedanalyticaldevicesforpoint-of-carediagnosis.NextGenerationPoint-of-CareBiomedicalSensorsTechnologiesforCancerDiagnosis, 365\u2013396.https://doi.org/10.1007/978-981-10-4726-8_1610.Martinez,A.W.,Phillips,S.T.,&Whitesides,G.M.(2008).Three-dimensionalmicrofluidicdevicesfabricatedinlayeredpaperandtape.ProceedingsoftheNationalAcademyofSciences,105(50),19606-19611.11.Yetisen,A.K.,Akram,M.S.,Lowe,C.R.(2013).Paper-basedmicrofluidicpoint-of-carediagnosticdevices.LabonaChip,13(12),2210-2251. 38 12.Fu,L.-M.,Tseng,C.-C.,Ju,W.-J.,&Yang,R.-J.(2018).Rapidpaper-basedsystemforhumanserumcreatininedetection.Inventions(Basel),3(2),34\u2013.https://doi.org/10.3390/inventions302003413.Tseng,C.-C.,Yang,R.-J.,Ju,W.-J.,&Fu,L.-M.(2018).Microfluidicpaper-basedplatformforwholebloodcreatininedetection.ChemicalEngineeringJournal, 348,117\u2013124.https://doi.org/10.1016/j.cej.2018.04.19114.Tseng,C.-C.,Lu,S.-Y.,Chen,S.-J.,Wang,J.-M.,Fu,L.-M.,&Wu,Y.-H.(2022).MicrofluidicaptasensorPOCdevicefordeterminationofwholebloodpotassium.AnalyticaChimicaActa,1203,339722\u2013339722.15.EdwardP. Randviir, CraigE.Banks,Analyticalmethodsforquantifyingcreatininewithinbiologicalmedia,SensorsandActuatorsB:Chemical,Volume183,2013,Pages239-252,ISSN0925-4005,https://doi.org/10.1016/j.snb.2013.03.103.16.Tambaru,D.,R.H.Rupilu,F. Nitti,I.Gauru,andSuwari.Developmentofpaper-basedsensorcoupledwithsmartphonedetectorforsimplecreatininedetermination.,2017.doi:10.1063/1.497816817.KatarzynaM.Koczula,A.G.Lateralflowassays.EssaysBiochem.60:111,201618.Ghosh,R.;Gopalakrishnan,S.;Savitha,R.;Renganathan,T.;Pushpavanam,S.FabricationofLaserPrintedMicrofluidicPaper-BasedAnalyticalDevices(LP- \u039c PADs)forPoint-of-CareApplications.Sci.Rep.2019,9(1),1\u201311. 19.Ruiz,R.A.,Gonzalez,J.L.,Vazquez-Alvarado,M.,Martinez,N.W.,&Martinez,A.W.(2022).Beyondwaxprinting:Fabricationofpaper-basedmicrofluidicdevicesusingathermaltransferprinter. AnalyticalChemistry, 94(25),8833\u20138837.https://doi.org/10.1021/acs.analchem.2c01534 20.Walters,C.D.Inf.Technol.Libr. 2004,23(1),30\u221236 21.Mendoza,J.D.AcuteKidneyInjury:Causes,Diagnosis,andTreatments.NovaBiomedical,2011,188pp.22.Ostermann,M.,andR.W. S.Chang.AcutekidneyinjuryintheintensivecareunitaccordingtoRIFLE*.Crit.CareMed.35:1837,2007.23.Acutekidneyinjuryat<https://www.nhs.uk/conditions/acute-kidney-injury/>24.HPRTMT800.https://www.hprt.com/Product/ConsumerElectronics/MT800.html(accessed2022-03-27).25.CsabaP. Kovesdy, Epidemiologyofchronickidneydisease:anupdate2022,KidneyInternationalSupplements,Volume12,Issue1,2022,Pages7-11,ISSN2157-1716,https://doi.org/10.1016/j.kisu.2021.11.003. 39", "Design and Validation of an In-vitro Pulsatile Flow Environment for Blood Clot Growth Characterization near Bileaflet MHVs Through Fluid Mechanics Yhanira Medina-Amaro and Santosh Dasari Technical Advisor: Dr. Alessandro Bellofiore BME 198B Senior Design Project II Department of Biomedical Engineering Charles W. Davidson College of Engineering San Jose State University May 15th, 2025 Key Words In-vitro model, thrombogenesis, Mechanical Heart Valve (MHV), fluid mechanics, shear stress, hemodynamics, platelet aggregation, pulsatile flow, onset time, hemocompatibility. Abstract This study presents the design, development, and validation of an innovative in-vitro pulsatile flow system, the ThromboGenicity Tester, to evaluate blood clot formation near bileaflet mechanical heart valves (MHVs). The device replicates physiological cardiac flow conditions with enhanced accuracy, utilizing hemocompatible materials and precise motor control mechanisms to prevent unintended thrombosis within the experimental setup. Validation was conducted through rigorous flow dynamics testing and comparative analysis against established physiological flow benchmarks. Results indicate that the ThromboGenicity Tester provides a reliable and physiologically representative platform for studying thrombus formation. The ultimate objective of this research is to contribute to reducing patient dependency on lifelong anticoagulant therapy and the associated risks, thereby improving the quality of life for patients with MHVs. Table of Contents Key Words ...................................................................................................................................... 2 Abstract ........................................................................................................................................... 2 Table of Contents ........................................................................................................................... 3 Figure List ....................................................................................................................................... 4 Table List ........................................................................................................................................ 4 Executive Summary ....................................................................................................................... 5 Literature Review .......................................................................................................................... 5 Biomedical Motivation ................................................................................................................... 8 Statement of Need .......................................................................................................................... 8 Materials and Methods .................................................................................................................. 8 I. Hardware System Configuration and Mechanical Design ............................................... 9 \u25cb Electrical Drive Unit System Configuration .................................................................... 9 \u25cb Mechanical Design and Assembly ................................................................................... 9 \u25cb Motor Control Logic and Programming ......................................................................... 11 II. Verification of Flow System ............................................................................................. 12 III. Whole Blood Experimentation and Thrombosis Validation ....................................... 13 \u25cb Porcine Blood Setup and Storage ......................................................................................... 13 Results ........................................................................................................................................... 14 I. Thrombogenicity Tester ..................................................................................................... 14 II. Flow Measurement Data .................................................................................................. 15 III. Cell Platelet Count Data ................................................................................................. 16 IV. Clot Formation Visualization .......................................................................................... 17 V. Sound Wave Data .............................................................................................................. 17 Discussion ..................................................................................................................................... 18 Conclusion .................................................................................................................................... 18 Future Work ................................................................................................................................. 19 Safety ............................................................................................................................................. 20 I. Safety Issues with Blood .................................................................................................... 20 II. Electrical Safety Issues ..................................................................................................... 20 Acknowledgements ...................................................................................................................... 22 References ..................................................................................................................................... 23 Cost Analysis ................................................................................................................................ 24 Appendix ....................................................................................................................................... 26 Figure List Figure 1 Electrical components of the TGT device including Arduino Uno, stepper motor, andmotor drive Figure 2 Exploded assembly diagram of the TGT mechanical component. Figure 3 Logic map of program", ".......................................................................................... 17 V. Sound Wave Data .............................................................................................................. 17 Discussion ..................................................................................................................................... 18 Conclusion .................................................................................................................................... 18 Future Work ................................................................................................................................. 19 Safety ............................................................................................................................................. 20 I. Safety Issues with Blood .................................................................................................... 20 II. Electrical Safety Issues ..................................................................................................... 20 Acknowledgements ...................................................................................................................... 22 References ..................................................................................................................................... 23 Cost Analysis ................................................................................................................................ 24 Appendix ....................................................................................................................................... 26 Figure List Figure 1 Electrical components of the TGT device including Arduino Uno, stepper motor, andmotor drive Figure 2 Exploded assembly diagram of the TGT mechanical component. Figure 3 Logic map of program functionality. Figure 4 Assembled Thrombogenicity Tester including: filling compartment, valve housing, blood circulation loop, and Transonic flow meter. Figure 5 Flow rate vs. time plot of TGT containing blood analog for verification. Figure 6 Flow rate vs. time for one simulated cardiac cycle by the TGT containing porcine blood. Figure 7 Paired comparison of Baseline vs. Post-clot platelets count across 181 samples Figure 8 Left image displays a clean MHV pre experiment, the right image displays the same MHV post experiment displaying clot formation Figure 9 Thrombogenesis detection: combined flow rate and acoustic analysis Table List Table 1 Reynolds and Womersley number values divided between cardiac phases. Executive Summary Heart valve disease is a debilitating condition that is normally caused by damage to the heart over time. This disease affects millions of individuals worldwide, with more than 5 million Americans suffering from this affliction and a total of 25,000 deaths per year (CDC 2024). From this number, the most affected populations are especially racial and ethnic minorities due to having the disease being underdiagnosed and undertreated 1 . The creation of mechanical heart valves (MHVs) has revolutionized the treatment of heart valve disease, enabling interventions that can save the life of patients while having a lower cost than prosthetic heart valves. By replacing a deteriorated heart valve with a MHV, patients can experience improved quality of life and extended lifespan. However, MHVs significantly elevate the risk of thromboembolism in patients who receive these implants. Research into the thrombogenic properties of MHVs has been hindered by the inadequacy of existing in-vitro simulation environments. Effective in-vitro blood simulations and MHV testing depend on appropriate material selection and hemocompatibility. Commonly utilized materials include medical-grade silicone, polycarbonate, and stainless steel, chosen for their inert properties and resistance to thrombosis and hemolysis. These materials must exhibit smooth surfaces with minimal roughness to reduce shear-induced platelet activation. Further reduction in protein adhesion and blood component activation", "who receive these implants. Research into the thrombogenic properties of MHVs has been hindered by the inadequacy of existing in-vitro simulation environments. Effective in-vitro blood simulations and MHV testing depend on appropriate material selection and hemocompatibility. Commonly utilized materials include medical-grade silicone, polycarbonate, and stainless steel, chosen for their inert properties and resistance to thrombosis and hemolysis. These materials must exhibit smooth surfaces with minimal roughness to reduce shear-induced platelet activation. Further reduction in protein adhesion and blood component activation can be achieved by employing antithrombogenic coatings, such as hydrophilic or superhydrophobic layers 2,9 . Additionally, all materials used in simulation setups must undergo hemocompatibility testing compliant with ISO 10993-4 standards to ensure they do not adversely affect the blood dynamics under investigation. Therefore, if an in vitro simulation environment accurately reproduces the physiological conditions of cardiac circulation through pulsatile flow and maintains a proper heart valve orientation reflective of in vivo positioning, then it will not induce clots nor generate blood cell damage, which will provide a suitable platform for reliable thrombogenicity analysis of mechanical heart valves. Literature Review Before diving deeper into the project details, it is important to understand the prior scientific knowledge of in vitro thrombogenicity testing simulations. Selecting proper materials based on the project's needs is important for any good design. Material selection and hemocompatibility are critical considerations for the simulation setup and the mechanical heart valve (MHV) testing. Materials such as medical-grade silicone, polycarbonate, and stainless steel are commonly used for the simulation setup due to their inertness and resistance to thrombosis and hemolysis. These materials must have smooth, low-roughness surfaces to minimize shear-induced platelet activation and may benefit from anti-thrombogenic coatings, such as hydrophilic or superhydrophobic layers, to further reduce protein adhesion and blood component activation 2,9 . Additionally, materials for simulation setups must be put through ISO 10993-4 compliant hemocompatibility testing to ensure they do not interfere with the studied blood dynamics. For MHVs, durable materials like pyrolytic carbon, titanium alloys, and cobalt-chromium alloys are often selected for their structural integrity and wear resistance, although these may also be enhanced with coatings to improve hemocompatibility 2 . The design of MHVs inherently creates regions of high shear stress, which can activate platelets and promote thrombus formation, necessitating careful evaluation of their hemodynamic performance 1 . Simulation setups must accurately replicate physiological conditions, including pulsatile flow, pressure, and shear stresses, to effectively evaluate", "For MHVs, durable materials like pyrolytic carbon, titanium alloys, and cobalt-chromium alloys are often selected for their structural integrity and wear resistance, although these may also be enhanced with coatings to improve hemocompatibility 2 . The design of MHVs inherently creates regions of high shear stress, which can activate platelets and promote thrombus formation, necessitating careful evaluation of their hemodynamic performance 1 . Simulation setups must accurately replicate physiological conditions, including pulsatile flow, pressure, and shear stresses, to effectively evaluate the thrombogenic potential of the MHVs and not the simulation materials themselves 5 . These considerations ensure the simulation environment and the tested MHVs meet rigorous hemocompatibility and material stability standards. Although significant progress has been made, challenges remain in standardizing these evaluations and optimizing materials for simulation fidelity and long-term MHV performance. The next most important aspect of in vitro simulations is the pulsatile fluid flow. The representation of pulsatile flow, which mimics the cyclic nature of cardiac output, is critical in evaluating the performance of MHVs and blood-contacting materials. Most studies utilize in vitro setups to replicate the human cardiovascular system's physiological pressure and flow profiles. For example, the MarioHeart system employs a toroidal flow loop to simulate the pulsatile flow patterns in the aortic root, achieving realistic amplitude and waveform shapes using a programmable control system 6 . Similarly, pulse duplicators are widely used for in vitro simulations, such as in Susin et al. (2017), where an elastic and anatomically accurate aortic model, along with precise pulsatile waveforms. This allowed for a detailed analysis of valve leaflet dynamics and downstream shear stresses. Flow representation in these systems relies heavily on time-resolved parameters, including velocity fields, pressure gradients, and shear stress distributions. Techniques such as particle image velocimetry (PIV) are frequently employed to visualize and quantify these parameters. For example, Arjunon et al. (2015) used PIV within a pulsatile flow system to assess shear-induced platelet activation around bileaflet mechanical heart valves (MHVs), highlighting the complex relationship between valve design and local flow dynamics. Numerical modeling also plays a significant role, with computational fluid dynamics (CFD) used to predict flow patterns and platelet activation indices under various conditions 7 . These models provide insights into areas prone to flow separation, stagnation, or excessive shear, which are critical for understanding thrombogenic risks. Several studies emphasize the importance of matching the compliance and geometry of the test setup to human anatomy.", "mechanical heart valves (MHVs), highlighting the complex relationship between valve design and local flow dynamics. Numerical modeling also plays a significant role, with computational fluid dynamics (CFD) used to predict flow patterns and platelet activation indices under various conditions 7 . These models provide insights into areas prone to flow separation, stagnation, or excessive shear, which are critical for understanding thrombogenic risks. Several studies emphasize the importance of matching the compliance and geometry of the test setup to human anatomy. For example, compliant silicone aortic models, with realistic distensibility and anatomical features such as the sinuses of Valsalva, are critical to capturing the physiological pressure-volume relationship during the cardiac cycle 3 . Furthermore, standardized setups like the Thrombosis Tester Helmholtz Institute Aachen (THIA3) ensure controlled boundary conditions to isolate the effects of valve geometry on flow field characteristics 3,6 . The MarioHeart model deserves special attention due to its unique design and ability to address some limitations in previous in vitro setups. It utilizes a closed-loop toroidal system with low surface-to-volume ratios, which minimizes undesired material interactions and ensures physiologically accurate flow dynamics 4 . The system's design enables the generation of non-turbulent, pulsatile flow patterns that closely resemble those in the ascending aorta. By incorporating porcine blood in some experiments, MarioHeart has demonstrated thrombus formation patterns on the suture ring of mechanical heart valves that mirror in vivo observations. Furthermore, its modular design allows for precise control of flow rates and patterns, making it highly suitable for testing both the thrombogenic potential of MHVs and the efficacy of new anticoagulant therapies. This adaptability, combined with its realistic simulation capabilities, highlights MarioHeart as a valuable tool in advancing the study of cardiovascular device hemocompatibility and performance 4 . Once the simulation design details of in vitro simulations are understood, it is logical to understand how it affects MHV thrombosis testing. Thrombogenicity is critical to evaluating mechanical heart valves (MHVs) and their compatibility with blood. Common testing methods include assays for thrombin-antithrombin complexes, platelet activation, hemolysis indices, and imaging techniques like scanning electron microscopy for visualizing thrombus formation. In vitro models offer controlled environments for these evaluations, isolating thrombogenic factors better than in vivo studies, which often introduce variability due to tissue and systemic responses 1,8 . Valve geometry significantly impacts hemodynamic performance and thrombogenic potential. Bileaflet MHVs, for example, create distinct regions of high shear stress around the hinge mechanisms,", "their compatibility with blood. Common testing methods include assays for thrombin-antithrombin complexes, platelet activation, hemolysis indices, and imaging techniques like scanning electron microscopy for visualizing thrombus formation. In vitro models offer controlled environments for these evaluations, isolating thrombogenic factors better than in vivo studies, which often introduce variability due to tissue and systemic responses 1,8 . Valve geometry significantly impacts hemodynamic performance and thrombogenic potential. Bileaflet MHVs, for example, create distinct regions of high shear stress around the hinge mechanisms, which are prone to platelet activation. In contrast, bioprosthetic valves mimic physiological flow patterns more closely but suffer from structural degradation over time. Design features, such as leaflet number and orifice shape, are optimized to balance durability and hemocompatibility 1 . Many in vitro studies rely on blood analogs, such as glycerin-water solutions, to simulate flow conditions while minimizing variability. While these provide greater experimental control, they cannot replicate the complex rheological properties of whole blood, including its cellular and protein components. Real blood provides physiologically relevant insights but introduces challenges like variability in clotting time and platelet response 7 . Particle image velocimetry (PIV) and computational fluid dynamics (CFD) are widely used to analyze flow fields, predict regions of flow separation, and quantify shear stresses. These methods provide critical insights into the fluid dynamics around MHVs and are increasingly used in conjunction with experimental setups to validate results. Emerging technologies, such as machine learning, have begun to show promise in optimizing valve design based on in vitro and computational data 1 . Despite advances, existing in vitro models face limitations in replicating physiological conditions, particularly compliance of native tissues and the complex interactions between blood cells and flow. Additionally, platelet activation and hemolysis thresholds still need to be better defined, requiring further investigation 1 . Compliance with international standards such as ISO 5840 for valve performance and ISO 10993-4 for blood-contacting biomaterials is essential for reproducibility and clinical translation. Standardized testing ensures comparability across studies and reduces variability in evaluating thrombogenic potential 8 . Biomedical Motivation Mechanical heart valves (MHVs) are vital devices that replace diseased valves but carry a high risk of thromboembolism, necessitating lifelong anticoagulant therapy. This not only burdens patients but also raises healthcare costs and bleeding risks. Understanding thrombus formation near MHVs is still limited due to inadequacies in current in-vitro testing systems. The proposed research aims to fill this gap by developing the ThromboGenicity", "Standardized testing ensures comparability across studies and reduces variability in evaluating thrombogenic potential 8 . Biomedical Motivation Mechanical heart valves (MHVs) are vital devices that replace diseased valves but carry a high risk of thromboembolism, necessitating lifelong anticoagulant therapy. This not only burdens patients but also raises healthcare costs and bleeding risks. Understanding thrombus formation near MHVs is still limited due to inadequacies in current in-vitro testing systems. The proposed research aims to fill this gap by developing the ThromboGenicity Tester (TGT) 2.0, a sophisticated in-vitro system that simulates the heart's pulsatile flow and gravity effects. This platform will facilitate studies on clot dynamics, ultimately improving MHV design and patient outcomes. Given the reliance on these devices and the risks of anticoagulation therapy, addressing the limitations of current in-vitro models for MHV research is critical. This project aims to enhance patient care and accelerate innovations in cardiovascular device design. Statement of Need The lack of a standardized in vitro whole blood-based testing of mechanical heart valves indicates a clear need for further research and aiding in developing future MHVs. Although the Mario Heart system has significantly contributed to the ability to simulate realistic conditions with whole blood, it has yet to prove that the thrombosis shown was caused solely by the MHV rather than the system itself. Furthermore, there has not been any system that can aid in capturing the exact time of thrombosis formation upon the MHVs during experimentation. Therein lies the real need for further research. A system that can not only replicate physiological flow but can also prove the system's thromboresistance while providing the ability to capture and study early on onset clotting on MHVs would provide tremendous insights into understanding the underlying mechanisms of MHV thrombosis. Such a system would not only bridge a critical gap in current cardiovascular research but also be a powerful tool for improving valve design, reducing the need for anticoagulant therapy, and ultimately reducing patients' risk of life-threatening complications. Advancing this study area is essential for enhancing patient safety, informing regulatory standards, and accelerating innovation in developing next-generation mechanical heart valves. Materials and Methods The ThromboGenicity Tester (TGT) 2.0 is an electromechanical system consisting of several subsystems: an electrical drive unit for flow generation, mechanical design and assembly that couples the drive unit to the actual fluid test section, and lastly, motor control logic via programming that generates a motor motion", "and ultimately reducing patients' risk of life-threatening complications. Advancing this study area is essential for enhancing patient safety, informing regulatory standards, and accelerating innovation in developing next-generation mechanical heart valves. Materials and Methods The ThromboGenicity Tester (TGT) 2.0 is an electromechanical system consisting of several subsystems: an electrical drive unit for flow generation, mechanical design and assembly that couples the drive unit to the actual fluid test section, and lastly, motor control logic via programming that generates a motor motion profile to the test section in order to replicate physiological flow. The following subsections outline how the subsystems were designed, manufactured, integrated, verified, and validated. I. Hardware System Configuration and Mechanical Design \u25cb Electrical Drive Unit System Configuration The electrical drive unit generates and regulates the fluid test section's controlled systolic and diastolic rotational actuation. It consists of four main components. A MeanWell \u24c7 LRS-350-48 Power Supply outputs 48 volts and 7.3 Amps, for a maximum power of 350 Watts required to supply adequate power to all electrical components. A STEPPERONLINE \u24c7 Nema 34 34HS38-5204S CNC stepper motor was selected for its precise, high-torque output and reliability, enabling consistent actuation required for generating a prolonged pulsatile motion profile. A STEPPERONLINE \u24c7 DM860A motor driver was used to provide high power out to the motor, processing microcontroller signals, and providing microstepping capabilities. Lastly, an Arduino Uno microcontroller delivered the programmed control profile to drive the electrical system. All of the mentioned components were wired following a schematic designed to ensure safe power distribution, signal integrity, and ease of troubleshooting. Figure 1. Electrical components of the TGT device including Arduino Uno, stepper motor, andmotor driver. \u25cb Mechanical Design and Assembly Computer-aided design software, SolidWorks, was utilized to design each critical component of the TGT system. Refer to the exploded assembly diagram in Figure 2 for identification of part numbers referenced below. The electrical components (parts 1, 3, 5, and 7) were strategically positioned within the assembly to provide sufficient space for wiring and operational clearance during test section movement. Custom mechanical designs were developed for all additional components required for system functionality. The base plate (part 1) measures 40 cm in length, 35 cm in width, and 1.30 cm in thickness. These dimensions ensure the device maintains a footprint appropriate for manual transportation and compatibility with standard laboratory incubators. Mounted on the base are two T-slot aluminum extrusions (part 4). These extrusions", "positioned within the assembly to provide sufficient space for wiring and operational clearance during test section movement. Custom mechanical designs were developed for all additional components required for system functionality. The base plate (part 1) measures 40 cm in length, 35 cm in width, and 1.30 cm in thickness. These dimensions ensure the device maintains a footprint appropriate for manual transportation and compatibility with standard laboratory incubators. Mounted on the base are two T-slot aluminum extrusions (part 4). These extrusions are secured to the base plate via countersunk screws that thread through the bottom of the plate. Their purpose is to elevate the motor assembly from the base surface. The motor sub-assembly, comprising the motor bracket (part 6) and the motor itself (part 7), attaches directly to these aluminum extrusions. This configuration provides rigid support for both the fluid test section and its containing bracket. Connected to the motor shaft is part 8, a stainless steel flanged shaft coupler designed to secure the shaft to the additional components requiring rotation. This coupler firmly tightens onto both interfacing components to minimize slippage and efficiently transfer rotational motion from the stepper motor to the attached component. On the flanged side of the shaft coupler, the test section retaining bracket is secured. This bracket consists primarily of parts 9 and 10, which are connected by four hollow steel tubes. This sub-assembly was specifically designed to hold the fluid test section in a vertical position, ensuring physiological orientation of the mechanical heart valve during testing. Lastly, the circular closed-loop test section itself was designed, consisting of two semicircular tubes (part 11), a fluid filling section (part 12), and gated stopcocks (part 13). On the side opposite to the filling section is the valve housing. As depicted in detail in section A, part 14 is the actual valve housing itself, comprising two sections between which the mechanical heart valve (part 15) is placed and secured. It is important to mention that the mechanical heart valve is positioned such that its leaflets are perpendicular to the ground when in the open position during clockwise rotation or systole. While gravity aids the closure of the MHV during counterclockwise motion (diastole), it does not interfere with valve function during systole. Overall, the designed system ensures physiological orientation of the mechanical heart valve while optimizing operational efficiency. The configuration facilitates straightforward setup procedures, streamlined experimentation protocols, simplified cleaning processes,", "It is important to mention that the mechanical heart valve is positioned such that its leaflets are perpendicular to the ground when in the open position during clockwise rotation or systole. While gravity aids the closure of the MHV during counterclockwise motion (diastole), it does not interfere with valve function during systole. Overall, the designed system ensures physiological orientation of the mechanical heart valve while optimizing operational efficiency. The configuration facilitates straightforward setup procedures, streamlined experimentation protocols, simplified cleaning processes, and accurate data collection. This integrated design approach allows researchers to conduct reliable and reproducible mechanical heart valve performance testing under controlled conditions that closely simulate in vivo hemodynamics. Additionally, all materials which come into contact with blood are manufactured from hemocompatible materials in accordance with ISO 10993-4 standard. Figure 2. Exploded assembly diagram of the TGT mechanical component. \u25cb Motor Control Logic and Programming In order for the designed system to function and simulate the pulsatile flow of the heart. A Stepper motor control logic was developed using the C++ programming language. This Arduino based control system simulates the pulsatile rhythm of the heart through precise stepper motor movements. The system alternates between systolic and diastolic phases of the cardiac cycle using a state machine approach, which divides the cycle into four distinct sections: acceleration and deceleration during both systole and diastole. This creates a circular alternating motion, which spins the fluid filled test section with the mechanical heart valve in a pulsatile and physiological manner. It is important to mention that the stepper motor travels the same number of steps in each phase, systole, and diastole. This detail is vital to ensure that gravity's effects only aid in the closure of the mechanical heart valve and provide the proper valve orientation throughout each cycle. This approach also has the added advantage of allowing wired sensing equipment to remain unchanged throughout the experiment's run time. The motor's acceleration profile is mathematically calculated to produce a natural, smooth transition between speeds rather than abrupt changes. The program begins with a physics-derived delay value and progressively adjusts it throughout the motion cycle, ensuring the motor acceleration and decelerations are performed in a controlled manner that mimics biological tissue movement. Speed limits were also implemented to enforce and maintain precise operational control. Lastly, position tracking is active throughout the operation. The system continuously monitors the motor's location relative to the starting position.", "is mathematically calculated to produce a natural, smooth transition between speeds rather than abrupt changes. The program begins with a physics-derived delay value and progressively adjusts it throughout the motion cycle, ensuring the motor acceleration and decelerations are performed in a controlled manner that mimics biological tissue movement. Speed limits were also implemented to enforce and maintain precise operational control. Lastly, position tracking is active throughout the operation. The system continuously monitors the motor's location relative to the starting position. After running for a predetermined duration, the system completes its current cycle before automatically returning to the original position and entering a stationary holding state. This design approach provides valuable performance for hardware testing applications while incorporating failsafe mechanisms and allowing potential future enhancements. Figure 3 provides a logic map of the program's function. Figure 3. Logic map of program functionality. II. Verification of Flow System Initial verification of the TGT system was conducted using a blood analog fluid. The fluid motion was continuously monitored using an ultrasonic flow meter to verify accurate replication of systolic and diastolic phases by the device. Flow profiles were assessed to ensure repeatable pulsatile motion and verify the integrity of the system under dynamic conditions. III. Whole Blood Experimentation and Thrombosis Validation \u25cb Porcine Blood Setup and Storage Porcine whole blood with heparin was acquired from Lampire Biological Laboratories. Initial hemoanalysis measurements were recorded using a calibrated hematology analyzer Heska Element HT5. Blood from four bottles ~2L was combined, homogenized, and adjusted to a target hematocrit range of 36-42% by diluting it with NaCl if necessary. The blood was then redistributed into the original blood bottles ~450mL/bottle and stored at room temperature for testing. \u25cb Protamine Preparation A protamine cocktail was prepared by dissolving 200mg of protamine in 2 mL of buffer solution, for each trial. Vials containing the cocktail were stored at room temperature for immediate use right before starting the TGT simulation. \u25cb Blood Trial Setup and Execution 1. MHV Preparation Four St. Jude MHVs were used for validation trials. Each valve was weighed using a precision balance and kept in a labeled weight boat. After inserting it in the TGT valve housing, flow direction was clearly marked. 2. Blood Administration After checking that the junction between Tygon tubing and ports of the valve housing and filling compartment were correctly sealed using PVC clamps, the tubing was filled with porcine blood through", "\u25cb Blood Trial Setup and Execution 1. MHV Preparation Four St. Jude MHVs were used for validation trials. Each valve was weighed using a precision balance and kept in a labeled weight boat. After inserting it in the TGT valve housing, flow direction was clearly marked. 2. Blood Administration After checking that the junction between Tygon tubing and ports of the valve housing and filling compartment were correctly sealed using PVC clamps, the tubing was filled with porcine blood through the fill valve using a funnel, ensuring air bubbles were eliminated at the most of the operator\u2019s ability. Then, the previously prepared vial of protamine was injected into the system, ensuring any fluid came out of the system and that no bubbles were introduced. The filling holes were closed using two caps and further sealed with teflon tape around the caps, and electrical tape around the complete filling valve compartment. 3. Incubation and Monitoring Within 5 minutes after protamine addition, the filled loop system was placed atop the TGT platform, which was already transferred into a 37\u00b0C incubator. The flow meter was clamped at the upper section of the Tygon loop while a microphone was placed close to the MHV to capture acoustic \u201cclicks\u201d during closure. The system was then run, with flow data and audio recorded continuously for 45 minutes. \u25cb Post-Trial Analysis At the end of the incubation period, the TGT motor would stop automatically. The TGT loop was then removed from the incubator. Post-trial flow profiles were recorded for each trial. Each MHV was extracted and reweighed to assess clot formation. Blood samples were analyzed pre- and post-trial using the Heska Element HT5 hematology analyzer to evaluate cellular integrity and coagulation activity. Results I. Thrombogenicity Tester Figure 4. Assembled Thrombogenicity Tester including: filling compartment, valve housing, blood circulation loop, and Transonic flow meter. II. Flow Measurement Data Figure 5. Flow rate vs. time plot of TGT containing blood analog for verification. Figure 6 . Flow rate vs. time for one simulated cardiac cycle by the TGT containing porcine blood. The figures 5 and 6 display data from an ultrasonic flow meter attached to the test section to monitor flow rate throughout a total experiment. The first figure shows 9 cardiac cycles extracted from a much larger data set to display the pulsatile nature of flow. A sharp peak is seen during systole followed by a recovery. Then", "blood analog for verification. Figure 6 . Flow rate vs. time for one simulated cardiac cycle by the TGT containing porcine blood. The figures 5 and 6 display data from an ultrasonic flow meter attached to the test section to monitor flow rate throughout a total experiment. The first figure shows 9 cardiac cycles extracted from a much larger data set to display the pulsatile nature of flow. A sharp peak is seen during systole followed by a recovery. Then a small portion of reverse flow is captured indicating diastole. The second graph shows an isolated cardiac cycle in order to calculate the stroke volume. Simpson\u2019s approximation was used to calculate the integral of the flow rate vs time yielding in a stroke volume of 100.78ml. Table 1. Reynolds and Womersley number values divided between cardiac phases. The Reynolds and Womersley numbers for blood flow are calculated through a sophisticated process accounting for blood's complex properties. The Reynolds number (Re = \u03c1 vd/ \u03bc) uses pig blood density (1053 kg/m\u00b3), vessel diameter (0.0254m), and flow velocity derived from volumetric flow rate (Q/( \u03c0\u00d7( d/2)\u00b2), converting L/min to m\u00b3/s). Blood's non-Newtonian viscosity is calculated via the Carreau-Yasuda model: \u03bc = \u03bc\u221e + (\u03bc \u2080 - \u03bc\u221e)\u00d7(1 + (\u03bb\u00d7\u03b3\u0307) \u1d43)^((n-1)/a), where \u03bc\u221e = 0.0035 Pa\u00b7s, \u03bc \u2080 = 0.056 Pa\u00b7s, \u03bb = 1.902s, n = 0.22, and a = 0.64. The model incorporates shear rate (\u03b3\u0307 = 8v/d), with corrections for temperature (exp(-0.02\u00d7(T-37))) and hematocrit ((hematocrit/0.40)^2.5). The Womersley number (Wo = (d/2)\u00d7\u221a(2 \u03c0 f/ \u03bd)) characterizes pulsatile flow effects by relating vessel diameter to frequency (heart rate/60 Hz) and kinematic viscosity (\u03bc/\u03c1), revealing how transient inertial forces affect blood flow profiles across vessel diameters. III. Cell Platelet Count Data Figure 7. Paired comparison of Baseline vs. Post-clot platelets count across 181 samples. As mentioned in the methods section, a hematology analyzer was used to measure baseline and post-experimental blood markers. Figure X shows the aggregate platelet count drop between before and after the experimentation. A drastic drop in platelet count is indicative of thrombocytopenia. IV. Clot Formation Visualization Figure 8. Left image displays a clean MHV pre experiment, the right image displays the same MHV post experiment displaying clot formation. V. Sound Wave Data Figure 9. Thrombogenesis detection: combined flow rate and acoustic analysis. Lastly Figure 9 shows a combined monitoring of flow rate and acoustic signals, divided into three distinct time", "shows the aggregate platelet count drop between before and after the experimentation. A drastic drop in platelet count is indicative of thrombocytopenia. IV. Clot Formation Visualization Figure 8. Left image displays a clean MHV pre experiment, the right image displays the same MHV post experiment displaying clot formation. V. Sound Wave Data Figure 9. Thrombogenesis detection: combined flow rate and acoustic analysis. Lastly Figure 9 shows a combined monitoring of flow rate and acoustic signals, divided into three distinct time periods. In the Pre-Clot Phase (blue region, before 18.3 minutes), the flow exhibits strong pulsatile patterns, with peaks reaching approximately 20 L/min, while the sound energy demonstrates moderate and consistent patterns. During the Clot Formation Region (yellow region, from 18.3 to 22 minutes), the flow gradually decreases in amplitude, indicating progressive flow restriction. Here, the sound energy displays a significant spike between 20.5 and 21.7 minutes, reaching maximum values of 0.0035 a.u. This acoustic signature coincides with the steepest decline in flow rate, likely representing turbulent flow and vibrations from partial occlusion. Finally, in the Post-Clot Phase (green region, after 22 minutes), the flow stabilizes at a lower amplitude with peaks around 8 L/min, and the sound energy settles into a lower, more stable pattern. Discussion The development and validation of the proposed Thrombogenicity Tester (TGT) demonstrated successful emulation of physiological pulsatile flow conditions, enabling in-vitro thrombogenesis testing near bileaflet mechanical heart valves. The rotational motion of the system, designed to mimic systolic and diastolic flow, replicated cardiac blood circulation with similar cyclic motion and acoustic consistency, as confirmed by \u201cclicking\u201d sound during valve closure. Similarly, flow rate recording and hematological analysis validated the system\u2019s capacity to sustain pulsatile flow conditions over a 45-minute simulation period with real-time data collection. Blood trials with porcine blood and protamine infusion showed that thrombus occured predictably on MHV surfaces, validating the capability of the proposed system to provide the environment where to detect early clot onset on MHV without inducing thrombogenesis. Moreover, the system enables real-time monitoring with a non-invasive method- ultrasound flow meter- to measure flow, avoiding the use of invasive probes and reducing interference with native flow fields. However, Reynolds number calculations indicated the presence of flow regimes that exceeded physiological parameters, potentially as a cause of turbulent effects due to the device's inner surfaces. Additionally, the calculated stroke volume slightly surpassed the normal values, suggesting that the stepper motor\u2019s acceleration", "early clot onset on MHV without inducing thrombogenesis. Moreover, the system enables real-time monitoring with a non-invasive method- ultrasound flow meter- to measure flow, avoiding the use of invasive probes and reducing interference with native flow fields. However, Reynolds number calculations indicated the presence of flow regimes that exceeded physiological parameters, potentially as a cause of turbulent effects due to the device's inner surfaces. Additionally, the calculated stroke volume slightly surpassed the normal values, suggesting that the stepper motor\u2019s acceleration profile may require refinement to more precisely match human hemodynamics. Despite these challenges, the platelet count and acoustic data showed consistent blood-cell integrity and valve performance, highlighting the TGT\u2019s hemocompatibility and functional reliability for thrombogenicity research. Conclusion Thrombogenicity analysis of mechanical heart valves is an important research area for bettering the understanding of clot formation and developing less thrombogenic artificial heart valves, benefiting patients who rely on lifelong anticoagulant therapies after having these devices implanted on them. This project addresses that need by developing a physiologically representative and hemocompatible pulsatile flow environment capable of supporting non-invasive in-vitro thrombogenesis studies through flow measurement and sound analysis. The real-time collection of flow data and sound provided insights into valve function and thrombus development with minimal fluid disturbance. The TGT was designed following design requirements to be compact, modular, and easy to assemble for standardized procedures, while also simulating flow dynamics of blood circulation pumped by the heart, this by using a controlled stepper motor. The system showed a good performance in replicating systolic and diastolic phases, maintained flow conditions for the duration of each simulation test, and did not show additional thrombogenesis induction by the system itself when in contact with blood. Furthermore, this developed system has a great potential for broader cardiovascular research and the study of thrombogenicity of cardiovascular devices, improving, this way, the safety and performance of these devices. By enabling controlled and reproducible simulation of blood flow, the platform can contribute to the design and testing of a wide range of implantable devices where risk of thrombosis is a critical concern. Future Work Future mechanical heart valve (MHV) designs must enhance thromboresistance for patients across all age groups. There is a need for improved testing methodologies to validate the thrombogenic performance of these valves, and ongoing research will focus on innovative in vitro, in silico, and preclinical platforms to better study flow-induced clotting and valve design effects 11", "platform can contribute to the design and testing of a wide range of implantable devices where risk of thrombosis is a critical concern. Future Work Future mechanical heart valve (MHV) designs must enhance thromboresistance for patients across all age groups. There is a need for improved testing methodologies to validate the thrombogenic performance of these valves, and ongoing research will focus on innovative in vitro, in silico, and preclinical platforms to better study flow-induced clotting and valve design effects 11 . Laser-based techniques, such as Particle Image Velocimetry (PIV) and Laser Doppler Velocimetry (LDV), yield insights into flow dynamics but face challenges with opaque blood, limiting their use in whole-blood testing. Innovations aim to merge flow characterization with thrombogenicity testing. Pulsatile flow systems, like MarioHeart, can simulate conditions with clear blood analogues or whole blood, yet isolating MHV-specific thrombogenicity remains difficult 11 . While ideally a dual-fluid platform testing system would operate with both clear and whole blood. A twin system, consisting of one for flow and another for thrombosis, could be a practical alternative. This would require matched flow profiles to accurately correlate flow and clotting data 11 . The future of our TGT at the CardioLab of San Jose State University intends to play a crucial role in the aforementioned twin system. As explored throughout this paper, the TGT successfully implemented physiological MHV placement, pulsatile flow, and thrombosis capture. Although these results are promising, improvements are still being made to this system. The first is to improve the motor control algorithm; the current system uses an array of step delays to create a linear acceleration curve of the stepper motor during the systolic and diastolic phases. Many widely used mock circulation loop (MCL) systems utilize linear actuators following a sinusoidal acceleration profile 12 . This sinusoidal approach can also be applied to the TGT\u2019s loop style pulsatile system, which could improve flow rate behavior and further bring it close to MCL systems. Thus bringing us closer to the proposed twin testing system 11 . One way to improve the TGT is through better design, material selection, and experimental protocols. In future developments, the test section should utilize a rigid PMMA tube, which will completely eliminate kinks in the test section. This change would also reduce the total number of materials that the blood comes into contact with, potentially enhancing the tester\u2019s biocompatibility. Currently, the process of filling the", "close to MCL systems. Thus bringing us closer to the proposed twin testing system 11 . One way to improve the TGT is through better design, material selection, and experimental protocols. In future developments, the test section should utilize a rigid PMMA tube, which will completely eliminate kinks in the test section. This change would also reduce the total number of materials that the blood comes into contact with, potentially enhancing the tester\u2019s biocompatibility. Currently, the process of filling the test section with blood leaves much to be desired, with the main concern being the blood's exposure to air during filling and throughout the day. In the future, we intend to mandate the use of blood bags to store and deliver blood to the test section in order to eliminate air contact. Safety I. Safety Issues with Blood Handling large amounts of porcine blood poses a potential risk of contact with bloodborne pathogens. To ensure safety during experiments involving blood, we wear the necessary personal protective equipment (PPE), including lab coats, nitrile gloves, safety goggles or face shields, and polypropylene sleeve covers. These measures help prevent direct contact with the blood and its potential pathogens. We also ensure that the porcine blood is procured from university-approved suppliers to minimize the risk of infectious blood in the first place. Additionally, lab materials and the surrounding environment are at risk of blood contamination. Although we follow strict protocols to avoid contamination and spills, eliminating 100% of the contamination has proven to be impossible.For this reason, we adhere to the strict bleach cleaning procedures provided by the university's health and safety department. This ensures that the lab itself and all equipment are free of blood before we can conclude that day's research and vacate the laboratory. II. Electrical Safety Issues During our experiments with whole blood and blood analogs, we position the fluid-filled test section loop close to electrical components, such as the Arduino, power supply, motor driver, and stepper motor. Since the power supply and motor/driver combination are high-voltage DC systems, there is an inherent risk of equipment failure, fire, and, most importantly, electrocution of laboratory staff. We follow specific safety steps to ensure the safe use of our electrical equipment. First, we conduct the fluid filling of the test section in a separate location away from the TGT itself to avoid accidental spills on the electronics. Additionally, we use Teflon plumber's", "Arduino, power supply, motor driver, and stepper motor. Since the power supply and motor/driver combination are high-voltage DC systems, there is an inherent risk of equipment failure, fire, and, most importantly, electrocution of laboratory staff. We follow specific safety steps to ensure the safe use of our electrical equipment. First, we conduct the fluid filling of the test section in a separate location away from the TGT itself to avoid accidental spills on the electronics. Additionally, we use Teflon plumber's and heavy-duty electrical tape to secure all areas at risk of leaks, ensuring they are watertight. Before attaching the test section to the TGT body, we wipe down and dry all wet surfaces with paper towels to mitigate electrical risks further. Future designs will integrate improved electrical safety features to minimize risks further. Lastly, we also ensure that all electrical equipment is never plugged into the same breaker lines and avoid daisy-chaining power strips to prevent overloading circuits, which can lead to potential hazards. Future designs will integrate improved electrical safety features to minimize risks further. Acknowledgements Our team would like to thank our technical advisor, Dr. Alessandro Bellofiore, for his support throughout the duration of this project. Additionally, we would like to thank the National Institute of Health (NIH) for funding this project, which was essential in making this work possible. We are also grateful to Joey Arey and Daniela Vivanco for their collaboration during the blood trial protocols and learning alongside us. References [1] Arjunon, S., Ardana, P. H., Saikrishnan, N., Madhani, S., Foster, B., Glezer, A., & Yoganathan, A. P. (2015). Design of a pulsatile flow facility to evaluate thrombogenic potential of implantable cardiac devices. Journal of biomechanical engineering, 137(4), 045001. https://doi.org/10.1115/1.4029579 [2] Bark DL Jr, Vahabi H, Bui H, Movafaghi S, Moore B, Kota AK, Popat K, Dasi LP. Hemodynamic Performance and Thrombogenic Properties of a Superhydrophobic Bileaflet Mechanical Heart Valve. Ann Biomed Eng. 2017 Feb;45(2):452-463. Doi: 10.1007/s10439-016-1618-2. Epub 2016 Apr 20. PMID: 27098219; PMCID: PMC5073049. [3] Bhattacharya, S., Berdanier, R. A., Vlachos, P. P., & Key, N. L. (2016). A New Particle Image Velocimetry Technique for Turbomachinery Applications. Journal of Turbomachinery. https://doi.org/10.1115/1.4033672 [4] Brubert, Jacob, et al. \u201cHemocompatibility of styrenic block copolymers for use in prosthetic heart valves.\u201d Journal of materials science. Materials in Medicinee vol. 27,2 (2016): 32. doi:10.1007/s10856-015-5628-7 [5] Chakraborty, S., Simon, M. G., & Bellofiore, A. (2025). Rethinking mechanical heart valves in", "Doi: 10.1007/s10439-016-1618-2. Epub 2016 Apr 20. PMID: 27098219; PMCID: PMC5073049. [3] Bhattacharya, S., Berdanier, R. A., Vlachos, P. P., & Key, N. L. (2016). A New Particle Image Velocimetry Technique for Turbomachinery Applications. Journal of Turbomachinery. https://doi.org/10.1115/1.4033672 [4] Brubert, Jacob, et al. \u201cHemocompatibility of styrenic block copolymers for use in prosthetic heart valves.\u201d Journal of materials science. Materials in Medicinee vol. 27,2 (2016): 32. doi:10.1007/s10856-015-5628-7 [5] Chakraborty, S., Simon, M. G., & Bellofiore, A. (2025). Rethinking mechanical heart valves in the aortic position: new paradigms in design and testing. Frontiers in Cardiovascular Medicine , 11 . https://doi.org/10.3389/fcvm.2024.1458809 [6] Devos M, Liesdek OCD, Suyker WJL, van Tuijl S, Schutgens REG, van de Vosse FN, de Heer LM, Rutten MCM. MarioHeart: Novel In-Vitro Flow Model for Testing Heart Valve Prostheses and Anticoagulant Therapies. ASAIO J. 2023 May 1;69(5):e192-e198. Doi: 10.1097/MAT.0000000000001915. Epub 2023 Mar 12. PMID: 36913553. [7] Garc\u00eda Garc\u00eda, J., Garc\u00eda Carrascal, P., Castro Ruiz, F., Manuel Mart\u00edn, F., & Fern\u00e1ndez, J. A. (2017). Effects of bifurcation-specific and conventional stents on coronary bifurcation flow. An experimental and numerical study. Journal of biomechanics, 54, 64\u201372. https://doi.org/10.1016/j.jbiomech.2017.01.043 [8] Min Yun B, Aidun CK, Yoganathan AP. Blood damage through a bileaflet mechanical heart valve: a quantitative computational study using a multiscale suspension flow solver. J Biomech Eng. 2014 Oct;136(10):101009. doi: 10.1115/1.4028105. PMID: 25070372; PMCID: PMC4151159. [9] Nobili M, Sheriff J, Morbiducci U, Redaelli A, Bluestein D. Platelet activation due to hemodynamic shear stresses: damage accumulation model and comparison to in vitro measurements. ASAIO J. 2008 Jan-Feb;54(1):64-72. doi: 10.1097/MAT.0b013e31815d6898. PMID: 18204318; PMCID: PMC2756061. [10] Vahidkhah, Koohyar, et al. \u201cFlow-Induced Damage to Blood Cells in Aortic Valve Stenosis.\u201d Annals of Biomedical Engineering vol. 44,9 (2016): 2724-36. doi:10.1007/s10439-016-1577-7 [11] Weber, Marbod, et al. \u201cBlood-Contacting Biomaterials: In Vitro Evaluation of the Hemocompatibility.\u201d Frontiers in bioengineering and biotechnology vol. 6 99. 16 Jul. 2018, doi:10.3389/fbioe.2018.00099 [12] Yin, W., Alemu, Y., Affeld, K. et al. Flow-Induced Platelet Activation in Bileaflet and Monoleaflet Mechanical Heart Valves. Annals of Biomedical Engineering 32, 1058\u20131066 (2004). https://doi.org/10.1114/B:ABME.0000036642.21895.3f [13] Zakaria, M.S., Ismail, F., Tamagawa, M. et al. Review of numerical methods for simulation of mechanical heart valves and the potential for blood clotting. Med Biol Eng Comput 55, 1519\u20131548 (2017). https://doi.org/10.1007/s11517-017-1688-9 Cost Analysis Name Quantity Price Total Cost Materials Chemicals Protamine Sulfate 4.4 kg $100.00 $100.00 PBS buffer solution 44 mL $0.49 $21.91 0.9% NaCl solution 1 L $29.00 $4.83 TrueClot clotting sol 1 $45.00 $45.00", "Monoleaflet Mechanical Heart Valves. Annals of Biomedical Engineering 32, 1058\u20131066 (2004). https://doi.org/10.1114/B:ABME.0000036642.21895.3f [13] Zakaria, M.S., Ismail, F., Tamagawa, M. et al. Review of numerical methods for simulation of mechanical heart valves and the potential for blood clotting. Med Biol Eng Comput 55, 1519\u20131548 (2017). https://doi.org/10.1007/s11517-017-1688-9 Cost Analysis Name Quantity Price Total Cost Materials Chemicals Protamine Sulfate 4.4 kg $100.00 $100.00 PBS buffer solution 44 mL $0.49 $21.91 0.9% NaCl solution 1 L $29.00 $4.83 TrueClot clotting sol 1 $45.00 $45.00 TrueClot solution 1 $75.99 $75.99 Laboratory Equipment Transonic flow meter 1 $1,500.00 $1,500.00 Bambu Lab 3D printer 1 $699.00 $699.00 Heska Element HT5 1 $7,000.00 $7,000.00 Scale 1 $600.00 $600.00 Incubator 1 $1,200.00 $1,200.00 Laboratory Supplies Stopcocks 8 $1.75 $14.00 Valve Housings 2 - - Tygon Clear Tube 1 $120.77 $120.77 3D printer filament 1 $158.00 $158.00 Arduino Uno 1 $32.00 $32.00 Motor Driver 1 $6.39 $6.39 St. Jude MHV 4 $250.00 $1,000.00 Screws, bolts 12 $15.00 $15.00 Samples Porcine Blood 8 L $2000.00 $4000.00 Software Arduino Uno - - - LabChart - - - Equipment Repair/Maintenance 1 $1,000.00 $1,000.00 Facilities E233J Total $17,592.89 Appendix I. BME 198A Proposal NSF Proposal Final TGT.pdf II. Technical Memorandums Technical Memos III. Motor Control Code Motor code on Github IV. Flow data LabChart", "Design and Evaluation of a Novel Trileaflet Mechanical Heart Valve to Achieve Natural Flow B.S. Biomedical Engineering Biomedical Engineering Department Charles W. Davidson College of Engineering May 16th, 2025 Seth Gonzalez Christopher Gudiel-Gutierrez Advised by: Alessandro Bellofiore P.H.D. Table of Contents Executive Summary 3 Introduction 3 Materials and Methods 4 Results 10 Discussion 15 Conclusion 19 Future works 19 References 21 2 Executive Summary The Centers for Disease Control states that more than 5 million people are diagnosed with heart valve disease each year, and it is the cause of more than 25,000 deaths yearly within the United States. Our research focuses on the design and evaluation of a novel trileaflet mechanical heart valve (MHV) that addresses limitations in current designs. Literature has shown that trileaflet valves demonstrate more physiological blood flow characteristics 6 . One reason that trileaflet valves have been shown to have optimal flow dynamics is due to the alignment of the valve with the native aortic root\u2019s Valsalva sinuses 10 . However, existing mechanical trileaflet valves often fail to replicate the natural blood flow dynamics of native valves due to creating a high-velocity region between the leaflets 13 . By incorporating the curved wall design of the bileaflet iValve created by researchers at the University of British Columbia 5 and applying it to a mechanical trileaflet heart valve, our research designed and created a trileaflet valve capable of promoting a more natural physiological flow. Introduction Cardiovascular diseases remain a leading cause of mortality worldwide, with heart valve diseases affecting millions of patients. Mechanical heart valves (MHVs) are critical in valve replacement procedures, as they have been used to replace diseased valves in over 3 million patients 1 . Currently, there are no trileaflet MHVs approved clinically, as bileaflet MHVs dominate the market. As a prosthetic heart valve, the proposed device would fall under the FDA\u2019s Class III designation. Researchers at the University of British Columbia have created a new revolutionary design for a bileaflet MHV known as the iValve, which incorporates a curved leaflet design. With this new design, they demonstrated that curved leaflet structures can significantly improve 3 hemodynamics by reducing flow disturbances and shear stress 5 . Current trileaflet MHVs often fail to replicate the natural blood flow dynamics of native valves due to creating a high-velocity region between the leaflets 13 . These high-velocity regions create non physiological flow patterns and can lead", "created a new revolutionary design for a bileaflet MHV known as the iValve, which incorporates a curved leaflet design. With this new design, they demonstrated that curved leaflet structures can significantly improve 3 hemodynamics by reducing flow disturbances and shear stress 5 . Current trileaflet MHVs often fail to replicate the natural blood flow dynamics of native valves due to creating a high-velocity region between the leaflets 13 . These high-velocity regions create non physiological flow patterns and can lead to increased shear stresses , which have the potential to contribute to complications such as thrombosis and the need for lifelong anticoagulation therapy 13 . To improve upon current Trileaflet MHV technology, our research has taken this curved leaflet design used on the iValve and applied it to a trileaflet configuration. Our hypothesis is that this approach will optimize central flow and reduce high-shear regions, more closely mimicking the natural hemodynamics of native heart valves. Using SolidWorks 3D software, we were able to successfully apply the curved design to the leaflets of a trileaflet MHV. Using computational fluid dynamics simulations in COMSOL to validate its performance. Materials and Methods The trileaflet mechanical heart valve was designed using SolidWorks, a computer-aided design (CAD). All design work was conducted using SolidWorks 2023\u20132024 edition, installed on a personal laptop running Windows 11. Solid modeling tools within SolidWorks were used to iteratively develop and refine the geometry of the valve housing and the leaflet structures. The design process began by referencing a prior bileaflet design created by Allesandro Bellofiore from the Department of Biomedical Engineering at San Jose State University for their trileaflet valve geometry. Specifically, the outer ring of this previous design was imported and used as a dimensional and anatomical foundation for the initial valve structure. 4 Development of the Curved Leaflet Design The leaflet geometry was developed independently and drew inspiration from the curved leaflet design demonstrated in the University of British Columbia\u2019s bileaflet iValve, as seen in Figure 1. The new design incorporates curvature into each leaflet to promote centralized, Figure 1: Image of iValve with curved leaflet design 5 circular outflow during systole. Unlike prior models, the trileaflet leaflet shape was not derived from preexisting CAD files or literature. Instead, the geometry was developed through an original trial-and-error process, manually adjusting hinge angle, hinge, and leaflet thickness in successive SolidWorks configurations. The design objective was to achieve a configuration", "British Columbia\u2019s bileaflet iValve, as seen in Figure 1. The new design incorporates curvature into each leaflet to promote centralized, Figure 1: Image of iValve with curved leaflet design 5 circular outflow during systole. Unlike prior models, the trileaflet leaflet shape was not derived from preexisting CAD files or literature. Instead, the geometry was developed through an original trial-and-error process, manually adjusting hinge angle, hinge, and leaflet thickness in successive SolidWorks configurations. The design objective was to achieve a configuration in which all three leaflets open 90\u00b0 during peak systole and form a uniform circular outflow jet, minimizing shear concentrations and high-velocity regions commonly seen in existing mechanical heart valves. Each iteration was manually evaluated for leaflet symmetry, manufacturability, and clearance from the valve housing. The butterfly hinge design seen in Figure 5 was selected due to its mechanical simplicity and ease of integration within the compact valve housing. The butterfly hinge allowed for a consistent and reliable pivoting motion of each leaflet while maintaining a small footprint and minimal assembly complexity. The closing angle of each leaflet was determined to be 47\u00b0. The finalized assembly included the newly developed curved leaflets seated within the outer ring. Assembly constraints and mates in 5 SolidWorks was used to simulate leaflet articulation and identify any potential interference. The clearance between the leaflet edge and the housing wall was set to 0.01 mm to prevent mechanical locking or friction. The finalized view of the entire MHV can be seen in Figure 4. COMSOL Setup The validation process was conducted using COMSOL Multiphysics 6.1, a simulation software widely used for physics-based and engineering modeling. Students at San Jos\u00e9 State University have free access to the software, which was utilized to run simulations on the mechanical heart valve (MHV). After completing the final CAD assembly of the MHV, it was imported into COMSOL, along with the aortic root geometry, which was passed down from previous project groups. It is important to note that the MHV was done on SOLIDWORKS; therefore, when importing into COMSOL, there were some issues. The issues involved the opening and closing mechanism of the MHV in COMSOL. It was not possible to either open or close the MHV due to the file being saved as a single part instead of an assembly. The solution to the issue was to go back into SOLIDWORKS and manually open or close the MHV, depending", "project groups. It is important to note that the MHV was done on SOLIDWORKS; therefore, when importing into COMSOL, there were some issues. The issues involved the opening and closing mechanism of the MHV in COMSOL. It was not possible to either open or close the MHV due to the file being saved as a single part instead of an assembly. The solution to the issue was to go back into SOLIDWORKS and manually open or close the MHV, depending on the validation process, and then save the MHV as a file. For example, when running a simulation during a systole environment, the MHV has to be manually opened on SOLIDWORKS, then saved as a file, and then imported into COMSOL to keep its shape to model the systole environment. Once both 3D CAD models were imported, they were carefully aligned by using the move and rotation geometry under the transform dropdown menu. The positioning of the MHV within the aortic root was rotated to a degree in which the valve leaflets were oriented parallel with the sinus bulbs. Following the alignment of the valve, the function union was used to form the two different objects into a single object. Since the tests done are using stationary analysis, the union command is needed to keep both the MHV and the aortic 6 root in place. Following the union function, a form composite faces command was used on the aortic root in order to simplify the geometry by merging multiple faces into a single large face. This command is useful on the aortic root due to the curved features and different faces the aortic root has. The final step when building the geometry was the remove detail command, which is an automatic setting useful to remove: vertices with continuous tangent, short edges, small faces, sharp faces, and thin domains. The feature is an automatic setting operated by COMSOL in order to simplify the complexity in the import of the MHV and aortic root. Once the final details are achieved, the final build is achieved as shown in Figure 2, which illustrates the final version of the aortic root and MHV in COMSOL. Figure 2: Frontal View (Left Image) and Top View (Right Image) of MHV and Aortic Root in COMSOL COMSOL Systole Laminar Flow Simulation As mentioned earlier, due to issues with the import of the MHV, the use of a", "simplify the complexity in the import of the MHV and aortic root. Once the final details are achieved, the final build is achieved as shown in Figure 2, which illustrates the final version of the aortic root and MHV in COMSOL. Figure 2: Frontal View (Left Image) and Top View (Right Image) of MHV and Aortic Root in COMSOL COMSOL Systole Laminar Flow Simulation As mentioned earlier, due to issues with the import of the MHV, the use of a different method to incorporate the MHV into the COMSOL software was used. The first simulation that ran consisted of a systole environment in the cardiac cycle. In a systole environment, the blood travels from the left ventricle into the aorta. In this cycle, the heart reaches its peak pressure 4 . 7 During the systole environment, the MHV is fully open; therefore, when importing the MHV into COMSOL, a necessary step was to save the MHV file in SOLIDWORKS as a fully opened MHV. Following the steps mentioned earlier, the simulation parameters were defined, such as geometry settings, material properties, boundary conditions, mesh generation, and computation. The simulation parameters used for the validation process were a standard laminar flow; turbulence would have been used, but it was not available due to our COMSOL licensing. The material properties used for laminar flow were water, due to COMSOL not having blood as a material and to serve as a starting point. The boundary conditions used for laminar flow were applied to the body of the aortic root and not the MHV itself. The wall boundary was applied to the aortic root and the MHV due to laminar flow being held at a stationary simulation. The last boundary set was the inlet and outlet condition, where the inlet was set on the opening of the left ventricle side, and the outlet was set on the opening of the aorta side. This is due to the cardiac cycle environment in which the systole process allows blood to pump from the left ventricle to the aorta. The inlet value was determined to be 120 mmHg with an outlet of 0 Pa according to literature 4. It is important to note that after 5 different simulations, the inlet and outlet conditions were switched to a velocity inlet instead of a pressure gradient. The velocity inlet value was 1.2 m/s, which was utilized in the Literature 5", "to the cardiac cycle environment in which the systole process allows blood to pump from the left ventricle to the aorta. The inlet value was determined to be 120 mmHg with an outlet of 0 Pa according to literature 4. It is important to note that after 5 different simulations, the inlet and outlet conditions were switched to a velocity inlet instead of a pressure gradient. The velocity inlet value was 1.2 m/s, which was utilized in the Literature 5 . Therefore, our results will be presented by using velocity inlets instead of pressure inlets. This step was preferred due to the results that better physiologically matched. The next step was to generate a mesh on the valve and the aortic root. The mesh is needed for solving simulations by dividing the geometry into smaller elements that the software can analyze. The mesh used was a coarser setting due to by having anything finer than coarser, the simulation would not compute. After the mesh, the simulation would begin, which is where our results are retrieved from. The data was extracted when we took a cut line in 8 3D across the midsection of the aortic root to derive the data and plot it. This data will show the velocity profile of the simulation at any given line point, but I decided to plot data from the given point to assess the jet peak velocity. The aim is to visualize the jet flow velocity to validate the data as a central jet flow. COMSOL Diastole Laminar Flow Simulation Similar to the systole simulation, the diastole environment had 3D imports from SOLIDWORKS into COMSOL. The diastole import had a different approach due to the left ventricle relaxing and filling up with blood during diastole. The aortic pressure is much greater than the left ventricular pressure, allowing a reverse pressure gradient and pushing the valve leaflets shut. Therefore, when importing a diastole environment MHV, the leaflets had to be shut on SOLIDWORKS and then saved and imported into COMSOL. Besides the different orientation of the leaflets, the same procedure is applied to the model, similar to a systolic environment. However, one change necessary was the change of the inlet and outlet boundaries. As mentioned earlier, in a diastolic environment, the greater pressure from the aorta side pushes down the valve leaflets, which can be described as the pressure being driven down into the", "MHV, the leaflets had to be shut on SOLIDWORKS and then saved and imported into COMSOL. Besides the different orientation of the leaflets, the same procedure is applied to the model, similar to a systolic environment. However, one change necessary was the change of the inlet and outlet boundaries. As mentioned earlier, in a diastolic environment, the greater pressure from the aorta side pushes down the valve leaflets, which can be described as the pressure being driven down into the left ventricle side. One way to simulate the occurrence is by applying the inlet boundary on the aorta side and the outlet on the left ventricle side, the opposite of the systole simulation. The inlet boundary value was 80 mmHg, and the outlet boundary value was 0 mmHg, similar to the systole testing, which was retrieved from literature 4 . Once that was settled, the simulation was computed. The computation served a different purpose from the systole testing because in diastole, there are no specific velocities one is looking for. Instead, the purpose of the simulation at the diastole phase was to find any openings within the MHV when fully closed. The simulation will therefore help the team make any improvements to the MHV before moving on 9 to any future work. Based on the following results, we aim to move into a Fluid-Structure Interaction (FSI) interface and then further work on the MHV. The main goal is to achieve a centralized jet flow and find any weak points to further modify the MHV for future work. Results Heart Valve design results. Figure 3: Isometric view, open valve, and closed valve view of the completed MHV designs. Figure 4: Female and Male sections of the butterfly valve. 10 Systole Results Figure 5 : First Run using Pressure Inlet at 120 mmHg Figure 6 : Slice Plot at Mid-Section in Systole (YZ Plane) using Velocity Inlet 11 Figure 7 : Cut Line 3D across Cross-Section at Systole (XZ Plane) Figure 8 : Line Graph across Cross-Section at Systole (XZ Plane) 12 Figure 9 : Cut Line 3D across Center-Line at Systole (YZ Plane) Figure 10 : Line Graph across Center-Line at Systole (YZ Plane) 13 Diastole Results Figure 11 : Max/Min at Diastole Across Center Plane (YZ Plane) Figure 12 : Velocity Magnitude at 85 mm at Diastole (YZ Plane) 14 Figure 13 : Velocity Magnitude at 88 mm at Diastole", "Line 3D across Cross-Section at Systole (XZ Plane) Figure 8 : Line Graph across Cross-Section at Systole (XZ Plane) 12 Figure 9 : Cut Line 3D across Center-Line at Systole (YZ Plane) Figure 10 : Line Graph across Center-Line at Systole (YZ Plane) 13 Diastole Results Figure 11 : Max/Min at Diastole Across Center Plane (YZ Plane) Figure 12 : Velocity Magnitude at 85 mm at Diastole (YZ Plane) 14 Figure 13 : Velocity Magnitude at 88 mm at Diastole (YZ Plane) Discussion The final design of the novel trileaflet mechanical heart valve (MHV) featured three symmetrically curved leaflets inspired by the iValve\u2019s bileaflet geometry developed at the University of British Columbia. Each leaflet was designed to open to 90\u00b0 during systole and close completely during diastole. The curvature of the leaflets was engineered to encourage a central, circular jet outflow. The butterfly hinge mechanism Figure 4 allowed consistent pivoting while minimizing internal footprint and mechanical complexity. The closing angle of each leaflet was set to 47\u00b0, and a 0.01 mm clearance was introduced between leaflet edges and the housing to reduce friction and avoid mechanical locking. The full valve assembly can be seen in Figure 3 from an isometric view, as well as fully open and closed. 15 Systole Results Figure 14: Past Group Work The systolic simulation initially employed a pressure inlet configuration, which resulted in non-physiological velocity magnitudes, as seen in Figure 4 . The maximum velocity observed exceeded 5 \u00d7 10 \u2076 m/s, which significantly deviates from expected physiological values near 1.0 m/s for the aortic valve during systole 4 . Furthermore, the flow visualization exhibited an asymmetric velocity distribution, with higher magnitudes concentrated along the superior region of the aortic bulb and minimal velocity near the inferior wall. These inconsistencies highlighted the need for an alternative simulation input. To address this, a velocity inlet boundary condition of 1.2 m/s, derived from literature benchmarks 5 , was applied. As shown in Figure 5 , this adjustment resulted in a more physiologically relevant velocity field, with a peak velocity of 2.5 m/s at the YZ plane. The flow was directed primarily through the center of the valve, and lower-velocity zones were observed along the vessel walls, consistent with laminar boundary layer development. While the velocity 16 distribution appeared slightly asymmetric, this was attributed to the orientation of the cross-sectional slice rather than geometric flow abnormalities.Subsequent analysis using", "was applied. As shown in Figure 5 , this adjustment resulted in a more physiologically relevant velocity field, with a peak velocity of 2.5 m/s at the YZ plane. The flow was directed primarily through the center of the valve, and lower-velocity zones were observed along the vessel walls, consistent with laminar boundary layer development. While the velocity 16 distribution appeared slightly asymmetric, this was attributed to the orientation of the cross-sectional slice rather than geometric flow abnormalities.Subsequent analysis using a 3D cut-line across the XZ plane, Figures 6 and 7 , provided a clearer depiction of the central flow jet. The velocity profile exhibited a parabolic distribution with a maximum velocity of 3.06 m/s at approximately 25 mm downstream of the valve, indicating the presence of a skewed but centralized jet. The parabolic shape and wall-adjacent near-zero velocities further support physiological accuracy. While 25 mm is offset from the anatomical midline of the aortic root, it still provides valuable insight into the flow development. A complementary analysis was performed along a longitudinal centerline through the valve and aortic root, Figures 9 and 10 . The velocity gradually increased from the inlet condition of 1.2 m/s, peaking twice\u2014once at ~17 mm and again at ~65 mm. The first peak corresponded to the geometric entrance of the aortic sinus, while the second peak is likely a consequence of flow convergence downstream. The trough between these peaks may be indicative of flow separation or recirculation within the sinus region. While the velocity profiles observed across various cross-sectional analyses suggest the emergence of a centralized jet flow, the current evidence is insufficient to conclusively validate its effectiveness. Although a peak velocity of 3.06 m/s was recorded and parabolic flow characteristics were present, the jet lacked full symmetry and centrality throughout the entire aortic root, particularly when evaluated across multiple planes. This inconsistency raises questions about the repeatability and stability of the flow generated by our novel trileaflet design. Furthermore, when comparing our results to those from a previous senior project, Figure 14 , their simulation outputs clearly demonstrate a more well-defined and centered jet flow. Their velocity field showed greater uniformity and alignment with physiological expectations, 17 indicating superior flow optimization. In contrast, the observed flow asymmetries and variability in peak velocity positions suggest that further refinement and validation are required. Diastole Results In contrast to systole, the diastolic phase was simulated using", "by our novel trileaflet design. Furthermore, when comparing our results to those from a previous senior project, Figure 14 , their simulation outputs clearly demonstrate a more well-defined and centered jet flow. Their velocity field showed greater uniformity and alignment with physiological expectations, 17 indicating superior flow optimization. In contrast, the observed flow asymmetries and variability in peak velocity positions suggest that further refinement and validation are required. Diastole Results In contrast to systole, the diastolic phase was simulated using a reversed pressure gradient to replicate physiological conditions in which aortic pressure exceeds left ventricular pressure. The valve was modeled in a fully closed state, and the inlet and outlet boundaries were reversed accordingly. The objective was to assess valve competency and identify any leakage pathways. The velocity magnitude slice plot in Figure 11 revealed a maximum velocity of 1.97 mm/s along the center plane of the aortic root. While low in magnitude, the presence of non-zero velocity through the valve suggests incomplete sealing. Notably, flow was observed through the center and peripheral regions of the valve, particularly between the housing and the leaflet edges. Further characterizing leakage, additional slice plots were taken just upstream, Figure 12 , and downstream, Figure 13 , of the valve. At 85 mm from the inlet, localized velocity magnitudes reached ~0.3 m/s, indicating minor gaps at the leaflet tips. Downstream at 88 mm, velocity magnitudes as high as 1.4 m/s were observed near the hinge regions, with fluid entering the centerline from these lateral pathways. While some degree of leakage is a known limitation in mechanical valve design, the current level of backflow highlights the need for refinement in the leaflet sealing and hinge configuration. Despite this, the diastolic simulation provides valuable insight into potential areas for design optimization. 18 Conclusion In conclusion, our initial objective of creating a trileaflet mechanical heart valve with curved leaflets inspired by the iValve was successfully completed. This valve design incorporated a central jet flow, which used three symmetrically curved leaflets and a butterfly hinge mechanism that allowed for full 90\u00b0 opening and complete closure with minimal friction. Fluid simulations conducted under systolic and diastolic conditions demonstrated preliminary formation of a centralized jet during systole and minor regurgitant flow during diastole. While these results suggest that the curved leaflet design promotes favorable flow patterns, the evidence remains inconclusive, and further validation is required. Continued refinement and additional simulations", "valve design incorporated a central jet flow, which used three symmetrically curved leaflets and a butterfly hinge mechanism that allowed for full 90\u00b0 opening and complete closure with minimal friction. Fluid simulations conducted under systolic and diastolic conditions demonstrated preliminary formation of a centralized jet during systole and minor regurgitant flow during diastole. While these results suggest that the curved leaflet design promotes favorable flow patterns, the evidence remains inconclusive, and further validation is required. Continued refinement and additional simulations will be essential to fully assess the valve\u2019s hemodynamic performance. Overall, the results affirm that the integration of curved leaflets into a trileaflet mechanical design holds promise for improving hemodynamic performance and reducing risks associated with conventional MHVs. This design may offer a path forward toward developing a clinically viable trileaflet MHV that more closely replicates the natural function of the human aortic valve. Future Works The next phase of this project will focus on refining the hinge mechanism to enhance the manufacturability. While the current butterfly hinge design proved effective in simulation, its small-scale precision requirements present challenges for fabrication and assembly. The current butterfly hinge, though functional, presents challenges in terms of precision and assembly at small scales. By optimizing the hinge geometry and reducing intricate clearances, we aim to create a more manufacturable and scalable design. Following the hinge redesign, additional 19 simulations will be conducted in COMSOL using a Fluid-Structure Interaction (FSI) framework. This simulation approach offers a more physiologically accurate representation of valve dynamics by accounting for both fluid forces and leaflet motion. Concurrently, physical prototypes of the valve will be fabricated using high-resolution 3D printing or resin-based techniques. These prototypes will be evaluated for dimensional accuracy, leaflet mobility, and fit within the housing. Once fabricated, the valve aims to be put into a mock circulation loop to gather experimental data. This setup will enable comparison against our simulation results and provide a more comprehensive assessment of valve performance under dynamic physiological conditions. 20 References 1. Abbott. (n.d.). Mechanical heart valves . Abbott Structural Heart. Retrieved May 16, 2025, from https://www.structuralheart.abbott/int/campaign/mechanical-heart-valvesstructuralheart.abbott 2. Carrel, T., et al. Evolving technology: The TRIFLO tri-leaflet mechanical valve without oral anticoagulation: A potential major innovation in valve surgery. Frontiers in Cardiovascular Medicine , 10, 1220633, 2023. https://doi.org/10.3389/fcvm.2023.1220633 3. Centers for Disease Control and Prevention. (2024, September 30). About heart valve disease . https://www.cdc.gov/heart-disease/about/heart-valve-disease.html 4. Dave, R., et al. Shear stress", "assessment of valve performance under dynamic physiological conditions. 20 References 1. Abbott. (n.d.). Mechanical heart valves . Abbott Structural Heart. Retrieved May 16, 2025, from https://www.structuralheart.abbott/int/campaign/mechanical-heart-valvesstructuralheart.abbott 2. Carrel, T., et al. Evolving technology: The TRIFLO tri-leaflet mechanical valve without oral anticoagulation: A potential major innovation in valve surgery. Frontiers in Cardiovascular Medicine , 10, 1220633, 2023. https://doi.org/10.3389/fcvm.2023.1220633 3. Centers for Disease Control and Prevention. (2024, September 30). About heart valve disease . https://www.cdc.gov/heart-disease/about/heart-valve-disease.html 4. Dave, R., et al. Shear stress quantification in tissue engineering bioreactor heart valves: A computational approach. Journal of Functional Biomaterials , 15(3), 76, 2024. https://doi.org/10.3390/jfb15030076 5. Chaothawee L. (2012). Diagnostic approach to assessment of valvular heart disease using magnetic resonance imaging, part II: a practical approach for native and prosthetic heart valve stenosis. Heart Asia , 4 (1), 171\u2013175. https://doi.org/10.1136/heartasia-2012-010124 6. Goode, D., Scotten, L., Siegel, R., and Mohammadi, H. Can mechanical heart valves perform similarly to tissue valves? An in vitro study. Journal of Biomechanics , 174, 112270, 2024. 7. Jun, B. H., et al. Effect of hinge gap width of a St. Jude medical bileaflet mechanical heart valve on blood damage potential: An in vitro microparticle image velocimetry study. Journal of Biomechanical Engineering , 136(9), 091008, 2014. https://doi.org/10.1115/1.4027935 8. Laha, S., Fourtakas, G., and Das, P. K., et al. Smoothed particle hydrodynamics-based FSI simulation of the native and mechanical heart valves in a patient-specific aortic model. Scientific Reports , 14, 6762, 2024. https://doi.org/10.1038/s41598-024-57177-w 9. Pawlikowski, M., and Nieroda, A. Comparative analyses of blood flow through mechanical trileaflet and bileaflet aortic valves. Acta of Bioengineering and Biomechanics , 24, 2022. 10. Piatti, F., Sturla, F., Marom, G., Sheriff, J., Claiborne, T. E., Slepian, M. J., Redaelli, A., and Bluestein, D. Hemodynamic and thrombogenic analysis of a trileaflet polymeric valve using a fluid-structure interaction approach. Journal of Biomechanics , 48(13), 3641\u20133649, 2015. https://doi.org/10.1016/j.jbiomech.2015.08.009 11. Salmonsmith, J., Ducci, A., and Burriesci, G. Does transcatheter aortic valve alignment matter? Open Heart , 6, 2019. 21 12. Scotten, L. N., Goode, D., Siegel, R., Blundon, D. J., Dutton, J. W., and Mohammadi, H. Curved leaflet design for mechanical heart valves: A novel approach. medRxiv , 2024. https://doi.org/10.1101/2024.09.12.24313331 13. StatPearls. (2024). Physiology, pulse pressure . National Center for Biotechnology Information. Retrieved May 16, 2025, from https://www.ncbi.nlm.nih.gov/books/NBK482408/ 14. Sukta, C., and Uangpairoj, P. Simulation study on influence of leaflet shape and open angle of tri-leaflet mechanical heart valve on blood flow.", "matter? Open Heart , 6, 2019. 21 12. Scotten, L. N., Goode, D., Siegel, R., Blundon, D. J., Dutton, J. W., and Mohammadi, H. Curved leaflet design for mechanical heart valves: A novel approach. medRxiv , 2024. https://doi.org/10.1101/2024.09.12.24313331 13. StatPearls. (2024). Physiology, pulse pressure . National Center for Biotechnology Information. Retrieved May 16, 2025, from https://www.ncbi.nlm.nih.gov/books/NBK482408/ 14. Sukta, C., and Uangpairoj, P. Simulation study on influence of leaflet shape and open angle of tri-leaflet mechanical heart valve on blood flow. Suranaree Journal of Science and Technology , 28(4), 010061(1\u201310), 2021. 22", "1 FLUID-STRUCTURE INTERACTION COMSOL MODEL FOR UNSTEADY FLOW THROUGH A 3D BI-LEAFLET MECHANICAL HEART VALVE A Thesis Presented to The Faculty of the Department of Biomedical Engineering San Jos\u00e9 State University In Partial Fulfillment of the Requirements for the Degree Master of Science by Aya Zaky May 2025 \u00a9 2025 Aya Zaky ALL RIGHTS RESERVED The Designated Thesis Committee Approves the Thesis Titled FLUID-STRUCTURE INTERACTION COMSOL MODEL FOR UNSTEADY FLOW THROUGH A 3D BI-LEAFLET MECHANICAL HEART VALVE by Aya Zaky APPROVED FOR THE BIOMEDICAL ENGINEERING DEPARTMENT SAN JOS\u00c9 STATE UNIVERSITY May 2025 Alessandro Bellofiore, Ph.D. Department of Bimedical Engineering Lin Jiang, Ph.D. Department of Mechanical Engineering Yue Luo, Ph.D. Department of Industrial and Systems Engineering ABSTRACT FLUID STRUCTURE INTERACTION COMSOL MODEL FOR UNSTEADY FLOW THROUGH A 3D BI-LEAFLET MECHANICAL HEART VALVE by Aya Zaky In this study, we developed a 3D Fluid -Structure Interaction (FSI) model using COMSOL\u00a9 to analyze the unsteady flow through a bi -leaflet mechanical heart valve. The model incorporates sufficient complexity to test various factors, such as leaflet angle rotation, material selection for rigid components, fluid properties, and different simulation modes. Simultaneously, the model maintains a simplified structure and is developed using a single software platform, facilitating ease of modification and expansion by other researchers or developers. The mechanical heart valve (MHV) design was created using St. Jude\u2019s geometric parameters as a reference. The input and output pressure waveforms are derived from interpolated ventricular and aortic pressures and were used as simulation inputs. The resulting velocity, vorticity, shear, and pressure profiles were analyzed and compared to previously published laboratory results and available clinical data from All of US Research databases for validation. The output velocity and shear stress profiles aligned strongly with published laboratory results, highlighting the mo del\u2019s reliability and potential for practical use. The findings contribute to the ongoing development of computational cardiovascular simulations, bridging the gap between experimental and numerical analyses in MHV designs. Keywords: FSI, 2D CFD, 3D CFD, bi-leaflet MHV, numerical model. v DEDICATION To my beloved parents, Gamal and Fatima: Words will never be enough to express the depth of my gratitude. Thank you for the values you have lived by and passed on to me\u2014the strength to never give up and the courage to keep challenging myself, no matter how hard the road may get. This thesis is mo re than a piece of academic work\u2014it is my comeback,", "designs. Keywords: FSI, 2D CFD, 3D CFD, bi-leaflet MHV, numerical model. v DEDICATION To my beloved parents, Gamal and Fatima: Words will never be enough to express the depth of my gratitude. Thank you for the values you have lived by and passed on to me\u2014the strength to never give up and the courage to keep challenging myself, no matter how hard the road may get. This thesis is mo re than a piece of academic work\u2014it is my comeback, a testament to resilience after a long pause filled with uncertainty and self-doubt. In those quiet moments when I almost gave up, it was your voice, your example, and your unwavering belief in me that pu lled me forward. The values you instilled in me never left\u2014they stayed, they whispered, and they pushed me to rise again. This is for you. To my dearest Children: Zain and Sarah, You are my heart, my light, and my greatest source of joy. Just as I was inspired by my parents, I hope this work becomes something that inspires you in your own journeys. I hope it teaches you to dream without fear, to pursue your passions fiercely, and t o always believe in yourselves\u2014even when the path ahead feels unclear. You are my proudest accomplishment, and I thank you for giving my life more meaning than I ever imagined possible. With all my love, always. vi ACKNOWLEDGEMENTS To my advisor: Prof. Alessandro Bellofiore, Thank you from the bottom of my heart for your steady guidance, your patience, and your unwavering support throughout this journey. Your belief in my abilities, even when I doubted myself, meant more than I can express. You gave me the space to grow, the encouragement to push forward, and the wisdom to navigate the challenges along the way. Your mentorship went beyond academics, and I am truly grateful for the role you played in bringing this work to life. To my reading committee: Dr. Lin Jiang and Dr. Yue Luo I would like to express my sincere gratitude to my reading committee for their thoughtful review, valuable feedback, and support throughout the process of making this thesis come to light. Your guidance has been instrumental, and I am truly thankful for your time and expertise. To my wonderful husband, Mostafa Thank you for being my rock through every moment of this journey. Your love, encouragement, and quiet strength", "life. To my reading committee: Dr. Lin Jiang and Dr. Yue Luo I would like to express my sincere gratitude to my reading committee for their thoughtful review, valuable feedback, and support throughout the process of making this thesis come to light. Your guidance has been instrumental, and I am truly thankful for your time and expertise. To my wonderful husband, Mostafa Thank you for being my rock through every moment of this journey. Your love, encouragement, and quiet strength have been the foundation beneath me. During the hardest days, when I questioned if I could really do this, you were always there, reminding me of who I am and why I started. You held space for my dreams and stood by me through the long hours, the doubts, and the triumphs. This accomplishment is as much yours as it is mine. I couldn\u2019t have done this without you, and I wouldn\u2019t have wanted to. vii TABLE OF CONTENTS List of Tables ............................................................................................................. viii List of Figures ............................................................................................................ ix List of Abbreviations ................................................................................................. xi 1. Introduction .......................................................................................................... 1 1.1 Motivation ...................................................................................................... 1 1.2 Overview of Prosthetic Heart Valves ............................................................ 1 1.3 Significance of Numerical Simulation ........................................................... 3 1.4 Proposed Approach ........................................................................................ 5 2. Literature Review................................................................................................. 7 2.1 Heart Valve Functionality and Design Evolution: ......................................... 7 2.1.1 Native Heart Valve Function ................................................................ 7 2.1.2 Bioprosthetic Heart Valves ................................................................... 8 2.1.3 Bioprosthetic Heart Valves Market Size............................................... 9 2.2. Mechanical Heart Valve (MHV) Designs..................................................... 9 2.2.1 Ball and Cage MHV Design .............................................................. 11 2.2.2 Tilting-Disk MHV Design ................................................................. 12 2.2.3 Bi-leaflet MHV Design ...................................................................... 12 2.2.4 Tri-leaflet MHV Design ..................................................................... 13 2.3. Simulation Techniques of Flow Through MHVs ........................................... 14 2.3.1 Two Fixed MHV States (Open/Shut) Based Simulation Model ........... 14 2.3.2 Fluid-Structure Interaction (FSI)-Based Model Using Numerical Solver ............................................................................................................. 14 2.3.3 Existing Meshing Techniques for FSI of MHV .................................... 15 2.3.4 Existing Fluid-Structure Coupling Computational Methods ................ 16 2.3.5 Overview of State-of-the-Art FSI Solvers ............................................ 18 3. Research Plan ....................................................................................................... 20 3.1 Simulation Framework................................................................................... 20 3.2 Fluid Flow Modeling ..................................................................................... 20 3.3 MHV Design and Structural Modeling .......................................................... 21 3.4 Fluid-Structure Interaction Configuration ..................................................... 22 3.5 Summary of Research Goal and Specific Aims ............................................. 23 4. 2D FSI Bi-Leaflet MHV Simulation Model ........................................................ 25 4.1 Materials and Methods ................................................................................... 25 4.1.1 Geometrical Parameters ........................................................................ 25 4.1.2 Fluid and Solid Domain Materials Parameters ..................................... 27 4.1.3 FSI Solver Setup ................................................................................... 27 viii 4.2", "FSI Solvers ............................................ 18 3. Research Plan ....................................................................................................... 20 3.1 Simulation Framework................................................................................... 20 3.2 Fluid Flow Modeling ..................................................................................... 20 3.3 MHV Design and Structural Modeling .......................................................... 21 3.4 Fluid-Structure Interaction Configuration ..................................................... 22 3.5 Summary of Research Goal and Specific Aims ............................................. 23 4. 2D FSI Bi-Leaflet MHV Simulation Model ........................................................ 25 4.1 Materials and Methods ................................................................................... 25 4.1.1 Geometrical Parameters ........................................................................ 25 4.1.2 Fluid and Solid Domain Materials Parameters ..................................... 27 4.1.3 FSI Solver Setup ................................................................................... 27 viii 4.2 Results ............................................................................................................ 30 4.2.1 Mesh Sensitivity Analysis..................................................................... 30 4.2.2 2D Simulation Results .......................................................................... 31 5. 3D FSI Bi-leaflet MHV Simulation Model ......................................................... 34 5.1 Materials and Methods ................................................................................... 34 5.1.1 3D COMSOL Simulation Setup ........................................................... 34 5.1.2 3D FSI Solver Setup ............................................................................. 35 5.2 Results ............................................................................................................ 36 5.2.1 3D COMSOL Simulation Results ......................................................... 36 5.2.2 Validations of 3D Model CFD Simulation ........................................... 38 5.2.2.1 Comparison of 3D Simulation Results with Experimental Data ...................................................................................................... 39 5.2.2.2 Statistical Analysis of Supporting Clinical Data ...................... 41 5.2.2.3 Comparison of 3D Simulation Results and Clinical Data ........ 43 6. Limitations ........................................................................................................... 46 7. Conclusions .......................................................................................................... 49 References .................................................................................................................. 51 ix LIST OF TABLES Table 1. Comparison Summary of Fixed and Moving Grid FSI Methods. ........ 16 Table 2. Comparison Summary of Partitioned and Monolithic FSI Coupling Approaches ........................................................................................... 17 Table 3. Summary of Research Goal and Overview of Specific Research Aims and Their Corresponding Deliverables ................................................. 23 Table 4. Simulated Aortic Root Geometrical Parameters ............................... 26 Table 5. Summary of Quadratic Bezier Polygon Parameters Used in Sinus Curve .................................................................................................... 27 Table 6. Summary of MHV Parameters ............................................................ 27 Table 7. Cohort Summary for Clinical Datasets based on All of US Research Database ............................................................................................... 42 Table 8. Comparison of CFD Simulation Results versus Clinical Data........... 44 x LIST OF FIGURES Figure 1. Summary of COMSOL\u2019s Implementation of FSI Solver Integrating Fluid Mechanics and Solid Dynamics. ................................................. 5 Figure 2. Available MHV Designs: (a) Ball and Cage, (b) Tilting Disk, (c) Bi- leaflet, and (d) Tri-leaflet ...................................................................... 11 Figure 3. Mock Circulation Loop in the Cardio Lab at SJSU .............................. 25 Figure 4. 2D Geometry for Aortic Root with Bi-leaflet MHV Model ................. 26 Figure 5. Pressure Waveforms at (a) Inlet/Ventricular and (b) Outlet/Aortic...... 29 Figure 6. Summary of Mesh Sensitivity Analysis for the 2D COMSOL BMHV Model .................................................................................................... 30 Figure 7. Leaflet Deformation Angle with Different Mesh Settings ................... 31 Figure 8. Velocity Profiles at", "Cage, (b) Tilting Disk, (c) Bi- leaflet, and (d) Tri-leaflet ...................................................................... 11 Figure 3. Mock Circulation Loop in the Cardio Lab at SJSU .............................. 25 Figure 4. 2D Geometry for Aortic Root with Bi-leaflet MHV Model ................. 26 Figure 5. Pressure Waveforms at (a) Inlet/Ventricular and (b) Outlet/Aortic...... 29 Figure 6. Summary of Mesh Sensitivity Analysis for the 2D COMSOL BMHV Model .................................................................................................... 30 Figure 7. Leaflet Deformation Angle with Different Mesh Settings ................... 31 Figure 8. Velocity Profiles at Mid-opening, Max-opening, and Mid-closing Positions. ............................................................................................... 32 Figure 9. Vorticity Profile at Two Different Time Instants During the Valve Closing Phase. ....................................................................................... 33 Figure 10. (a) Cut Line at the Middle of Aortic Sinus, (b) Vorticity Magnitude at Different Time Instants. ........................................................................ 33 Figure 11. Simulated 3D Aortic Root with MHV Model in COMSOL. ................ 34 Figure 12. Simulated Time-Domain Aortic/Ventricular Pressure Signals with Sliding Time Window .......................................................................... 35 xi Figure 13. Simulated Velocity (left), Vorticity (middle) and VSS (right) Profiles in the XZ Plane at y=0. ......................................................................... 36 Figure 14. Simulated Velocity (left), Vorticity (middle), and VSS (right) Profiles in the YZ Plane at x=22mm ................................................................. 37 Figure 15. Simulated Velocity Profiles in the XY Plane at (a) y=0mm, (b) y=4mm, and (c) y=6.25mm ................................................................ 38 Figure 16. (a) XY Plane at 86\u00b5m Away from the Valve Hinge, (b) Visualization of Leakage Flow in the XZ Plane. ........................................................ 40 Figure 17. Fluid Flow Results at the Vessel Root Annulus for EOA Calculations. 41 Figure 18. Summary of Clinical Data Available on All of US Research for (a) Maximum Systolic Velocity, (b) Average Systolic Pressure Gradient 43 xii LIST OF ABBREVIATIONS FSI \u2013 Fluid-Structure Interaction CVD \u2013 Cardiovascular Diseases BHV \u2013 Bioprosthetic Heart Valves SVD \u2013 Structural Valve Degradation MHV \u2013 Mechanical Heart Valve MCL \u2013 Mock Circulation Loop FEM \u2013 Finite Element Method ALE \u2013 Arbitrary Lagrangian-Eulerian EOA \u2013 Effective Orifice Areas CFD \u2013 Computational Fluid Dynamics 2D \u2013 Two Dimensional 3D \u2013 Three Dimensional Pa(t) \u2013 Aortic Pressure Pv(t) \u2013 Ventricular Pressure PIV \u2013 Particle Image Velocimetry A5C \u2013 apical-5 chamber LES \u2013 Large Eddy Simulation RANS \u2013 Reynolds-Averaged Navier-Stokes 1 1. Introduction 1.1 Motivation Cardiovascular disease s (CVD) remain the leading cause of mortality worldwide. According to published global health data, CVD was responsible for approximately 17.9 million deaths in 2019, accounting for nearly one -third of all global deaths [1]. In the United States, one individual", "Two Dimensional 3D \u2013 Three Dimensional Pa(t) \u2013 Aortic Pressure Pv(t) \u2013 Ventricular Pressure PIV \u2013 Particle Image Velocimetry A5C \u2013 apical-5 chamber LES \u2013 Large Eddy Simulation RANS \u2013 Reynolds-Averaged Navier-Stokes 1 1. Introduction 1.1 Motivation Cardiovascular disease s (CVD) remain the leading cause of mortality worldwide. According to published global health data, CVD was responsible for approximately 17.9 million deaths in 2019, accounting for nearly one -third of all global deaths [1]. In the United States, one individual dies every 34 seconds from cardiovascular -related conditions [2]. Among these, heart valve failure is a significant contributor, often arising from a combination of factors such as aging, congenital heart defects, chronic kidney disease, genetic predispositions, and other underlying cardiovascular conditions [3]. Heart valve dysfunction typically results from two principal mechanical failures: stenosis and regurgitation. Stenosis refers to the narrowing of the valve opening, which restricts blood flow during the inflow phase. Regurgitation, on the other hand, occurs when the valve fails to close properly, allowing blood to leak backward after ejection. Both conditions are leading causes of surgical heart valve replacement [4]. 1.2 Overview of Prosthetic Heart Valves Bioprosthetic heart valves (BHV) are often favored for their biocompatibility, as their mechanical properties closely resemble those of native tissue. However, BHVs are limited by their reduced long -term durability and are prone to structural valve degrada tion (SVD), primarily due to calcification. This degradation often leads to the development of common valve dysfunctions, such as stenosis and regurgitation [5, 6]. In contrast, mechanical heart valves (MHVs) maintain their structural integrity over time; however, their use typically 2 requires long -term anticoagulant therapy to prevent thrombosis, which remains a primary concern associated with MHVs [7]. MHVs have been extensively studied for several decades starting from the 1950s to date for their superior durability relative to their bioprosthetic counterparts. Advancements of MHV designs have been focused on the use of proper materials, leaflets\u2019 shape, and the opening and closure mechanism of the valve aiming to reduce the risk of thrombosis formation. MHVs produce relatively high shear on the blood due to the huge gap bet ween their mechanical properties compared to the intrinsic mechanical properties of tissue and the surrounding blood [8]. Various designs have been implemented starting from the historic ball and cage design followed by the tilted disc design [9], and the bi -leaflet MHV which is the current clinical", "leaflets\u2019 shape, and the opening and closure mechanism of the valve aiming to reduce the risk of thrombosis formation. MHVs produce relatively high shear on the blood due to the huge gap bet ween their mechanical properties compared to the intrinsic mechanical properties of tissue and the surrounding blood [8]. Various designs have been implemented starting from the historic ball and cage design followed by the tilted disc design [9], and the bi -leaflet MHV which is the current clinical standard and is still being extensively researched [10]. Furthermore, tri-leaflet MHVs are the most recent design to be investigated and have not been clinically implemented yet [11]. Although the tri -leaflet MHV research is still underdeveloped, one group reported performance improvement of a tri-leaflet design compared to a bi-leaflet design, however, they still reported clotting deposition around the valve hinges [12]. Henceforth, extensive testing is highly needed for constant improvement and refinement of MHV designs. Mock Circulation Loop (MCL) is an in-vitro testing technique that applies circulatory flow with controlled flow rate to an MHV device emulating blood circulation. MCL is used extensively to test clotting formation on MHVs and is usually paired with a CAD file to further investigate a particular problem with specific conditions [13]. Feeding such a tool with a comprehensive numerical model for the unsteady flow through MHV is of great importance to imitate the conditions and 3 flow characteristics including the hemodynamics of the blood to further eliminate nuisance variables. 1.3 Significance of Numerical Simulation The lack of reliable and comprehensive yet relatively straightforward simulations of blood flow through an MHV significantly limits the ability to optimize their design and performance. In the context of systemic circulation, the heart functions by pumping oxygen ated blood through the mitral and aortic valves to supply the body\u2019s tissues and organs. Once the oxygen is delivered, the deoxygenated blood returns to the heart via the tricuspid valve and is then directed through the pulmonary valve to the lungs for reoxygenation. Accurately modeling this cycle, particularly in the presence of an MHV, poses a considerable computational challenge. This complexity arises from several physiological and mechanical factors, including the non - Newtonian rheology of blood , the unsteady and pulsatile nature of pressure fluctuations throughout the cardiac cycle, and the convoluted two-way fluid-structure interaction between the blood and the dynamic movement of the valve leaflets. Capturing these dynamics in a", "is then directed through the pulmonary valve to the lungs for reoxygenation. Accurately modeling this cycle, particularly in the presence of an MHV, poses a considerable computational challenge. This complexity arises from several physiological and mechanical factors, including the non - Newtonian rheology of blood , the unsteady and pulsatile nature of pressure fluctuations throughout the cardiac cycle, and the convoluted two-way fluid-structure interaction between the blood and the dynamic movement of the valve leaflets. Capturing these dynamics in a realistic computational framework is essential for improving MHV designs and ensuring long- term functionality and biocompatibility [14]. The complex interplay between the fluid (i.e., blood) and the solid structure (i.e., heart valve) is effectively captured through various FEM analysis techniques, which are widely used in computational biomechanics. Among the notable contributions in this area, S. Ha et al. [15] developed a numerical solver using custom FORTRAN code in 2022 to simulate this problem with a high level of accuracy. While this approach demonstrates considerable computational capability, its implementation complexity 4 and reliance on low -level programming languages such as FORTRAN present a barrier to widespread adoption\u2014particularly among designers and researchers who may not possess a background in numerical programming. To address this challenge and promote broader engagement with FSI modeling in heart valve studies, leveraging commercial FEM software becomes a practical and accessible alternative. Commercial platforms not only streamline the modeling workflow but also enable reproducibility, collaboration, and future extensions of the work by a broader range of users within the biomedical engineering community. COMSOL\u00a9 Multiphysics offers a robust, fully coupled FSI solver that facilitates the simulation of dynamic interactions between fluids and deformable solids. This solver employs the Arbitrary Lagrangian-Eulerian (ALE) method, a hybrid computational framework that combines the Eulerian formulation for fluid domains \u2014used to solve the Navier -Stokes equations governing blood flow\u2014and the Lagrangian formulation for the solid domains, which describes the deformation of the heart valve structure. This dual -domain approach allows for accurate and synchronized modeling of the mutual influence between fluid and solid behavior during the cardiac cycle. To provide a clearer understanding of how COMSOL integrates both fluid mechanics and solid dynamics within its FSI framework, a flow chart illustrating the computational sequence of operations is presented in Fig.1. 5 Figure 1. Summary of COMSOL\u2019s Implementation of FSI Solver Integrating Fluid Mechanics and Solid Dynamics. 1.4 Proposed Approach In this research, I focused", "structure. This dual -domain approach allows for accurate and synchronized modeling of the mutual influence between fluid and solid behavior during the cardiac cycle. To provide a clearer understanding of how COMSOL integrates both fluid mechanics and solid dynamics within its FSI framework, a flow chart illustrating the computational sequence of operations is presented in Fig.1. 5 Figure 1. Summary of COMSOL\u2019s Implementation of FSI Solver Integrating Fluid Mechanics and Solid Dynamics. 1.4 Proposed Approach In this research, I focused on simulating the unsteady flow through an MHV using the FSI model developed within the COMSOL Multiphysics environment. The FSI problem was addressed by dividing it into three main components: solving the fluid domain using the Eulerian formulation of the Navier \u2013Stokes equations (as shown in the left portion of Fig . 1), solving the solid domain using the Lagrangian formulation suitable for deformable structures (illustrated in the right portion of Fig. 1), and capturing the interaction between the fluid a nd solid domains by enforcing boundary conditions associated with the continuity equation and conservation of momentum at their shared interface (depicted in the bottom portion of Fig. 1). A fully coupled FSI solution required the use of deformable meshes in both the solid and fluid domains. In scenarios involving large deformations, the fluid exerted high pressure on 6 the solid structure, leading to significant displacement of the valve leaflets. These large deformations in turn compromised mesh quality, which negatively impacted the convergence of the iterative solution and limited the effective domain in which the Arb itrary Lagrangian- Eulerian (ALE) method could produce accurate results. The primary objective of this study was to develop a reliable, versatile, and computationally manageable FSI model for simulating blood flow through MHVs. This was achieved by implementing physics -controlled, manually assigned meshes as direct inputs to the solver, thereby enhancing control over mesh quality and solution accuracy. An additional strength of the model is its configurability; a wide range of variables\u2014such as inlet and outlet conditions, velocity profiles, pressure conditions, MHV geometry, an d material properties\u2014 which can be seamlessly configured within the simulation framework. Furthermore, the model was designed to be adaptable and not limited to a specific MHV configuration. Instead, it was intended to serve as a generalized platform for simulating various MHV designs and flow conditions, thus providing a robust foundation for future investigations and iterative design improvements. 7 2. Literature", "its configurability; a wide range of variables\u2014such as inlet and outlet conditions, velocity profiles, pressure conditions, MHV geometry, an d material properties\u2014 which can be seamlessly configured within the simulation framework. Furthermore, the model was designed to be adaptable and not limited to a specific MHV configuration. Instead, it was intended to serve as a generalized platform for simulating various MHV designs and flow conditions, thus providing a robust foundation for future investigations and iterative design improvements. 7 2. Literature Review 2.1 Heart Valve Functionality and Design Evolution: 2.1.1 Native Heart Valve Function The human heart contains four essential valves that regulate blood flow: the pulmonary and tricuspid valves on the right side, and the mitral and aortic valves on the left side. The pulmonary and tricuspid valves control the flow of deoxygenated blood retu rning from the body, directing it to the lungs for oxygenation. In contrast, the mitral and aortic valves manage the flow of oxygen -rich blood from the lungs, ensuring it is effectively pumped through the aorta and distributed to the rest of the body. Thes e valves function through precise opening and closing mechanisms that maintain unidirectional blood flow, thereby preventing backflow and ensuring efficient circulation throughout the cardiac cycle [16]. These valves open during the inflow phase as a result of the pressure exerted by the incoming blood, allowing it to pass through efficiently. Once the blood has moved to the next chamber or into circulation, the valves promptly close to prevent any backflo w, a condition known as regurgitation. This coordinated opening and closing mechanism is crucial for maintaining the unidirectional flow of blood and for ensuring optimal cardiac performance throughout the cardiac cycle [17]. The most common types of valve malfunction stem from either the valve's failure to open efficiently during the inflow phase or its inability to close properly after blood has passed through. These conditions\u2014commonly referred to as stenosis and regurgitation \u2014disrupt the efficiency of the heart\u2019s function, leading to increased cardiac wo rkload and decreased blood 8 flow. Over time, such dysfunction often necessitates surgical intervention in the form of heart valve replacement, which has driven the development and continual improvement of prosthetic heart valves. 2.1.2 Bioprosthetic Heart Valves Bioprosthetic heart valves were initially explored due to their close resemblance to native heart valves, particularly in terms of mechanical behavior and biocompatibility. These valves are typically fabricated from biologically", "regurgitation \u2014disrupt the efficiency of the heart\u2019s function, leading to increased cardiac wo rkload and decreased blood 8 flow. Over time, such dysfunction often necessitates surgical intervention in the form of heart valve replacement, which has driven the development and continual improvement of prosthetic heart valves. 2.1.2 Bioprosthetic Heart Valves Bioprosthetic heart valves were initially explored due to their close resemblance to native heart valves, particularly in terms of mechanical behavior and biocompatibility. These valves are typically fabricated from biologically derived materials, such as porcine or bovine pericardial tissue, or constructed using natural polymers that mimic the extracellular matrix. This biomimetic design allows bioprosthetic valves to integrate more naturally with the surrounding tissue, minimizing immune response and offe ring smoother hemodynamic performance compared to their mechanical counterparts [18]. However, despite their favorable biocompatibility, bioprosthetic valves are limited by their relatively short lifespan. Over time, these valves tend to lose their structural integrity primarily due to calcification\u2014a process in which calcium deposits accumulate on the valve tissue, leading to stiffness, reduced mobility, and eventual functional failure. This degradation significantly compromises the long -term performance of bioprosthetic valves, especia lly in younger patients. As a result, the need for more durable alternatives became evident, paving the way for the development and continuous evolution of mechanical heart valves, which are engineered to provide long -lasting structural strength and reliable function over extended periods [19]. 9 2.1.3 Bioprosthetic Heart Valves Market Size The global bioprosthetic heart valve market is experiencing significant growth, driven by the increasing prevalence of valvular heart diseases, advancements in bioprosthetic valve technology, and a rising elderly population. In 2022, the market was valued at approximately $5.8 billion and is projected to reach around $9.4 billion by 2027, growing at a compound annual growth rate (CAGR) of 10.0% during this period [20]. This growth is further supported by the preference for bioprosthetic valves due to their biocompatibility and reduced need for long-term anticoagulation therapy, making them particularly suitable for older patients. Major companies o perating in this market include Medtronic PLC, Edwards Lifesciences Corporation, Boston Scientific Corporation, LivaNova PLC, and Abbott Laboratories (St. Jude Medical Inc.) [21]. 2.2. Mechanical Heart Valve (MHV) Designs In contrast to bioprosthetic heart valves, MHVs are recognized for their exceptional strength and long -term durability, making them a preferred choice for patients requiring lifelong valve function. However, this advantage comes with certain drawbacks. The significant mismatch", "long-term anticoagulation therapy, making them particularly suitable for older patients. Major companies o perating in this market include Medtronic PLC, Edwards Lifesciences Corporation, Boston Scientific Corporation, LivaNova PLC, and Abbott Laboratories (St. Jude Medical Inc.) [21]. 2.2. Mechanical Heart Valve (MHV) Designs In contrast to bioprosthetic heart valves, MHVs are recognized for their exceptional strength and long -term durability, making them a preferred choice for patients requiring lifelong valve function. However, this advantage comes with certain drawbacks. The significant mismatch between the mechanical properties of MHVs and the native cardiovascular tissue leads to altered hemodynamics, particularly elevated shear stress on the surrounding blood. This increased shear can promote platelet activation, which con tributes to the formation of blood clots. As a result, MHVs are associated with a heightened risk of thrombosis, necessitating the use of lifelong anticoagulation therapy to mitigate these complications [7]. Addressing the issue of thrombogenicity has remained one of the primary 10 challenges for engineers and researchers throughout the evolution of MHV designs. Over the years, continuous efforts have been made to optimize valve geometry and material selection to minimize shear-induced blood damage while maintaining the valve\u2019s mecha nical durability. The development journey began with the ball -and-cage valve design, followed by more advanced configurations such as tilting disc valves and eventually le d to modern bi -leaflet MHVs. Additionally, tri-leaflet MHVs aim to better mimic the physiological behavior of native valves , but t ri-leaflet MHVs are still under research and not clinically available . A detailed overview of these existing MHV designs and their respective mechanical and hemodynamic characteristics is presented in the following subsections [22-24]. A depiction of the different MHV designs is shown in Fig. 2. 11 Figure 2. Available MHV Designs: (a) Ball and Cage, (b) Tilting Disk, (c) Bi-leaflet, and (d) Tri-leaflet 2.2.1 Ball and Cage MHV Design The ball -and-cage design , shown in Fig. 2(a), was the first MHV design investigated. During the forward flow, the blood pushes the ball away from the cage, causing circumstantial movement of the blood around the ball in 360\u00b0. The ball then moves back adjacent to the cage by the pressure difference caused by the blood on the ball [25]. However, it is worth mentioning that this design is now obsolete due to the advancement of the further MHV designs. 12 2.2.2 Tilting-Disk MHV Design Tilting-disk design , shown in Fig. 2(b), followed", "design investigated. During the forward flow, the blood pushes the ball away from the cage, causing circumstantial movement of the blood around the ball in 360\u00b0. The ball then moves back adjacent to the cage by the pressure difference caused by the blood on the ball [25]. However, it is worth mentioning that this design is now obsolete due to the advancement of the further MHV designs. 12 2.2.2 Tilting-Disk MHV Design Tilting-disk design , shown in Fig. 2(b), followed the ball -and-cage design with little changes in the opening and closing mechanisms of the valve during the inflow and outflow phases. As the name reflects, the tilted disc opens in an asymmetric manner to create large and small Effective Orifice Areas (EOA) , causing the blood to flow through the openings with different fluxes. A higher flow rate passes through the large EOA, and a relatively smaller flow rate passes through the smaller EOA [8]. 2.2.3 Bi-leaflet MHV Design Bi-leaflet MHVs , shown in Fig. 2(c), represent one of the most widely adopted and extensively studied designs in the history of mechanical heart valve development. As the second most recent innovation in MHV evolution, these valves have demonstrated a favorable balance between mechanical dur ability and hemodynamic performance. Due to their symmetrical leaflet configuration and relatively low resistance to forward flow, bi -leaflet valves remain a preferred choice in both clinical settings and ongoing research and development efforts. Their con tinued use highlights their reliability, while also serving as a foundation for further advancements in valve design and optimization. Various designs of bi-leaflet MHV were investigated with a focus on the angle of the tilted leaflets, hinge positioning, EOA, and material selection [9, 26]. The fundamental structure of the bi-leaflet MHV features two semicircular leaflets that pivot around central hinges. When the valve is in the open position, these leaflets rotate to allow blood flow through three primary 13 EOAs: two lateral passageways on either side of the valve and one central opening between the leaflets. This configuration promotes a more uniform flow distribution compared to earlier designs and contributes to improved hemodynamic performance by reducing flow resistance and pressure drop across the valve [9]. 2.2.4 Tri-leaflet MHV Design The tri-leaflet MHV, shown in Fig. 2(d), has been under investigation recently due to its anatomical similarity to the native heart valve in both structure and function. This design aims", "EOAs: two lateral passageways on either side of the valve and one central opening between the leaflets. This configuration promotes a more uniform flow distribution compared to earlier designs and contributes to improved hemodynamic performance by reducing flow resistance and pressure drop across the valve [9]. 2.2.4 Tri-leaflet MHV Design The tri-leaflet MHV, shown in Fig. 2(d), has been under investigation recently due to its anatomical similarity to the native heart valve in both structure and function. This design aims to more closely replicate the natural dynamics of valve operation. During the inflow phase, the hinges guide the three leaflets to open in a coordinated manner, creating a single central EOA through which blood can pass efficiently. Once the forward flow of blood decreases and the pressure gradient reverses, the leaflets ar e driven back toward the center, effectively closing the valve to prevent backflow. Research and development efforts have primarily concentrated on optimizing the position of the hinges, the contour and geometry of the leaflets, and the leaflet opening ang les to enhance flow efficiency and reduce the risk of thrombogenic complications [27]. The first tri -leaflet MHV design that showed promising in -vitro results was made by K. Schubert et al. [11]. They aimed to enhance the EOA by using six hinges, two hinges attached to each leaflet. They also used polymeric leaflets instead of conventional metallic leaflets to enhance the biocompatibility and reduce the probability of thrombosis formation. 14 2.3. Simulation Techniques of Flow Through MHVs Simulating blood flow through prosthetic heart valves is inherently complex, as it requires the integration of several interdependent components. These include a robust CFD solver to capture the unsteady fluid behavior, a solid mechanics solver to accurate ly model the deformation and motion of the valve structure, and a well -defined interface to couple the interaction between the two domains. The problem becomes even more challenging due to the non-Newtonian behavior of blood and the nonlinear mechanical response of the valve materials under physiological loading conditions. These factors demand careful consideration of both numerical stability and physiological relevance. An overview of the most adopted simulation strategies for modeling flow circulation through MHVs \u2014including solver types, coupling techniques, and boundary conditions\u2014is provided in the following subsections. 2.3.1 Two Fixed MHV States (Open/Shut) Based Simulation Model In the early stage of developing a simulation model of the blood flow through the MHV, researchers", "of blood and the nonlinear mechanical response of the valve materials under physiological loading conditions. These factors demand careful consideration of both numerical stability and physiological relevance. An overview of the most adopted simulation strategies for modeling flow circulation through MHVs \u2014including solver types, coupling techniques, and boundary conditions\u2014is provided in the following subsections. 2.3.1 Two Fixed MHV States (Open/Shut) Based Simulation Model In the early stage of developing a simulation model of the blood flow through the MHV, researchers started by assuming that the MHV has two fixed states: open and shut. The flow dynamics are then studied in a 2D environment for each of the cases separately [28]. In this case, there is a one-way interaction between the fluid and the solid (valve). This simulation methodology is an oversimplification of the actual complicated scenario , which involves full coupling between the fluid and the valve. 2.3.2 Fluid-Structure Interaction (FSI)-Based Model Using Numerical Solver The blood flow through the MHV is best described as a Fluid -Structure Interaction (FSI) problem. As described in the introduction chapter, a fully coupled FSI model is decomposed 15 into a fluid mechanics solver and a solid dynamics solver, and these two solvers are coupled through the FSI boundary conditions [29]. Throughout the solution, both the fluid and solid domains need to be discretized. The challenge arises from the fact that the solid domain mesh needs to have the freedom to move/deform based on the fluid forces, and similarly, the fluid domain mesh needs to change with every single movement of the solid structure. Furthermore, the fluid-structure coupling is enforced by applying certain boundary condition s to either one or both solid/fluid domains. Consequently, the FSI solvers could be categorized based on two main factors : the definition of the fluid/solid meshes and the fluid -structure coupling mechanism [30]. The different meshing techniques are discussed in section 2.3.3 , and the different ways to implement the fluid-structure coupling are reported in section 2.3.4. 2.3.3 Existing Meshing Techniques for FSI of MHV Various techniques have been explored to implement a meshing algorithm accurately and efficiently for the FSI solver for MHVs. The two commonly used meshing techniques are the fixed-grid [31-32] and moving-grid methods [33-35]. The term \u201cmoving\u201d or \u201cfixed\u201d here refers only to the fluid\u2019s mesh condition since the solid\u2019s mesh is moving in all cases. In the fixed grid situation, the fluid\u2019s mesh", "implement the fluid-structure coupling are reported in section 2.3.4. 2.3.3 Existing Meshing Techniques for FSI of MHV Various techniques have been explored to implement a meshing algorithm accurately and efficiently for the FSI solver for MHVs. The two commonly used meshing techniques are the fixed-grid [31-32] and moving-grid methods [33-35]. The term \u201cmoving\u201d or \u201cfixed\u201d here refers only to the fluid\u2019s mesh condition since the solid\u2019s mesh is moving in all cases. In the fixed grid situation, the fluid\u2019s mesh is fixed, and the solid\u2019s mesh is allowed to move along with the solid domain displacement. In the moving grid meth ods, both the fluid and the solid meshes are represented in a deformable mesh at each time instant. This formulation is commonly known as the Arbitrary-Langranien Eulerian (ALE) method. In the ALE, the fluid mesh elements near the FSI boundary are allowed to move in any arbitrary direction based on the 16 solid deformation. However, this moving grid needs updating every time instant in the numerical solution to reflect the current location of the solid/fluid particles [36]. The ALE is widely used compared to the fixed mesh; however, the ALE cannot provide sufficient numerical accuracy in large deformation scenarios. A comparison between the two meshing methods is summarized in Table 1. Table 1. Comparison Summary of Fixed and Moving Grid FSI Methods. Grid Type Solid Domain Mesh Fluid Domain Mesh Movement Control Fixed Grid Moving Fixed Solid movement transmitted to fluid domain through boundary conditions Moving Grid Moving Moving Fluid elements near the solid allowed to move in any arbitrary direction 2.3.4 Existing Fluid-Structure Coupling Computational Methods To demonstrate the importance of fluid -structure coupling in the combined solver, a comparison between the resultant wall shear stress on a Tri -leaflet textile heart valve was reported using fully-coupled and de -coupled solvers at different observation points [37] implemented using co -simulation of two commercial software packages, Flow Vision\u00a9 and Abaqus FEA\u00a9. The authors concluded that when the observation point is away from the leaflets\u2019 deformation area, there is no significant difference between the fu lly and de-coupled solvers. Furthermore, with progressing toward the deformation area, the fully coupled solver showed higher stress levels, inferring that some of the pressure points were considered in the 17 fully coupled method and were neglected in the decoupled solution. Given the highlighted importance of the fully coupled solution, the discussion shifts to its", "Abaqus FEA\u00a9. The authors concluded that when the observation point is away from the leaflets\u2019 deformation area, there is no significant difference between the fu lly and de-coupled solvers. Furthermore, with progressing toward the deformation area, the fully coupled solver showed higher stress levels, inferring that some of the pressure points were considered in the 17 fully coupled method and were neglected in the decoupled solution. Given the highlighted importance of the fully coupled solution, the discussion shifts to its implementation techniques. The fluid-structure coupling implementation can be decomposed into two main categories: partitioned coupling [38-40] and monolithic approach [41-43]. The partitioned coupled solvers are based on either solving the fluid differential equations first and then applying the numerically acquired velocity to the solid interface or vice versa [38]. The partitioned coupled solvers usually suffer from numerical instability issues, given that at each time instant, one of the domains is partially solved without taking the other domain into account [40]. To mitigate this issue, the monolithic approach combines both the fluid dynamic s and solid mechanics governing differential equations together. Such formulation is tremendously complex and requires extensive computational effort [42]. A comparison summary of the partitioned and monolithic approaches is shown below in Table 2. Table 2. Comparison Summary of Partitioned and Monolithic FSI Coupling Approaches Coupling Method Implementation Advantages Disadvantages Partitioned Approach \u2022 Each of the two PDEs is solved separately. \u2022 The solution from each is used as a boundary condition to the other at the FSI interface. Easy to implement. Numerical instability convergence issues. Monolithic Approach \u2022 Both PDEs are combined into one matrix and solved simultaneously. Improved accuracy. Complex implementation. 18 2.3.5 Overview of State-of-the-Art FSI Solvers Bornoff, et al developed a simulation environment to study the behavior of both the aortic and mitral valves in series in 2023 [43]. They used the Ansys Fluent 3D module to simulate the fluid flow. Moreover, they used pre-existing User Defined Functions (UDFs) to assign the motion properties to each of the moving zones. The structure had to be simplified assuming two symmetric boundary conditions, to only solve 25% of the computational domain. Another work worth mentioning work reported in [33], where Ansys Fluent 3D module in conjunction with a custom-built C program was used to simulate the Bi-leaflet MHV. The FSI model used ALE formulation (moving grid), and the partitioned coupling method was utilized. The coupling is", "pre-existing User Defined Functions (UDFs) to assign the motion properties to each of the moving zones. The structure had to be simplified assuming two symmetric boundary conditions, to only solve 25% of the computational domain. Another work worth mentioning work reported in [33], where Ansys Fluent 3D module in conjunction with a custom-built C program was used to simulate the Bi-leaflet MHV. The FSI model used ALE formulation (moving grid), and the partitioned coupling method was utilized. The coupling is done using the C program as they read the velocity profile on the leaflets from Ansys solution and feed it to the C program, which then calculates the estimated new location of the valve leaflets. The new anticipated valve position is then changing the Ansys deforming domain. A recent work [5] in 2023 demonstrated CFD simulation of a tri-leaflet valve using a custom code called CgLES-Y, which combines two solvers: CgLES for the fluid domain using the fixed grid method and Y -code for the solid domain. The partitioned coupling scheme is implemented to interface the two codes together. Furthermore, S. Ha et al [14] developed a CFD solver simulating the unsteady flow through an aortic valve using FORTRAN code on a textile tri-leaflet valve in 2022. Additionally, a native tricuspid heart valve was simulated in 2023 using SIMULIA software [44]. As discussed in the literature review, most of the existing CFD solvers for the fluid dynamics through MHV were implemented using a complicated computational analysis where researchers develop ed their codes implementing the FEM solution for both the fluid and 19 structure domains. In addition to the complexity involved in developing such codes, the solver is usually customized to certain valve designs and environments , which limits its reusability for testing other designs and fluid conditions. Alternatively, other researchers opt to use two separate FEM solvers (commercial software packages) to solve for the fluid and solid domain respectively and then dynamically link the two solvers using a script implementing the FSI module. All these challenges contributed to the heavy dependence on lab testing purely as a means for testing and enhancing the prosthetic heart valve designs. Such challenges lay the necessity for a simplified, yet reliable, FSI solver that is not limited to a specific valve design and implemented within a single Multiphysics solver. 20 3. Research Plan This chapter outlines the research plan that guided the simulation and", "then dynamically link the two solvers using a script implementing the FSI module. All these challenges contributed to the heavy dependence on lab testing purely as a means for testing and enhancing the prosthetic heart valve designs. Such challenges lay the necessity for a simplified, yet reliable, FSI solver that is not limited to a specific valve design and implemented within a single Multiphysics solver. 20 3. Research Plan This chapter outlines the research plan that guided the simulation and modeling efforts undertaken in this study. The primary objective of the research was to implement an accurate, reliable, and versatile FSI simulation environment to study unsteady flow through an MHV using COMSOL Multiphysics. The research aimed to develop a computational model capable of capturing the complex dynamics of MHV function under physiologically relevant conditions while maintaining flexibility for future adaptation to various valve geometries and fluid settings. 3.1 Simulation Framework The simulation environment was developed entirely within COMSOL Multiphysics, utilizing its built -in FSI module. The simulation process was structured into three interdependent components: (1) fluid flow modeling, (2) MHV design and structural modeling, and (3) FSI coupling settings. 3.2 Fluid Flow Modeling The unsteady nature of blood flow \u2014arising from the systolic -diastolic cycle \u2014was captured using a time-dependent solver. The inlet and outlet boundary conditions were defined by physiologically realistic pressure waveforms representing the ventricular and aortic pressures, respectively. These waveforms allowed the simulation to mimic real cardiac cycles, enhancing the clinical relevance of the computational results. Blood was modeled as a Newtonian fluid for the purposes of this simulation, consistent with prior literature that validated this simplification in large-vessel cardiovascular flow simulations. 21 3.3 MHV Design and Structural Modeling The MHV geometry was introduced in two stages. In the initial phase, a two -dimensional (2D) representation of a bi-leaflet MHV was constructed, where each leaflet was modeled as a rigid rod hinged at the base. This simplified geometry was chosen to reduce computational cost and allow for faster iterative refinement of the FSI model. The 2D simulation provided an effective platform to evaluate solver stability, mesh deformation behavior, and parameter sensitivity. Upon validating the 2D model, the simulation setup was extended to a fully three - dimensional (3D) representation of a bi -leaflet MHV. The 3D model incorporated more realistic geometric and boundary conditions, enabling a more comprehensive analysis of leaflet dynamics and flow distribution. Validation of", "geometry was chosen to reduce computational cost and allow for faster iterative refinement of the FSI model. The 2D simulation provided an effective platform to evaluate solver stability, mesh deformation behavior, and parameter sensitivity. Upon validating the 2D model, the simulation setup was extended to a fully three - dimensional (3D) representation of a bi -leaflet MHV. The 3D model incorporated more realistic geometric and boundary conditions, enabling a more comprehensive analysis of leaflet dynamics and flow distribution. Validation of the 3D model was conducted by comparing simulation results with previously published experimental data from laboratory studies. These comparisons focused on flow velocity profiles, pressure gradients, wall shear stress, and vorticity plots. A final validation phase was also implemented, wherein a statistical analysis was constructed based on available clinical data. Statistical Cohorts were created for the maximum systolic velocity and the average systolic pressure gradient based on available clinical ly measured lab data. The data was further processed in which 1) participants with heart failure were excluded, 2) duplicate data points for the same participant were also removed, and 3) the resultant dataset was analyzed using an R script to obt ain the statistical clinical average for 22 maximum velocity and mean pressure gradient during the systolic phase . The statistical data is then compared against the numerical simulation results for validation. 3.4 Fluid-Structure Interaction Configuration The FSI component of the simulation formed the core of the research methodology. COMSOL's Arbitrary Lagrangian -Eulerian (ALE) method was employed to resolve the interaction between the fluid and solid domains. The fluid dynamics were governed by the Eulerian form of the Navier\u2013Stokes equations, while the solid mechanics were modeled using the Lagrangian framework. The interaction between the two domains was enforced at their interface by ensuring continuity of velocity and conservation of momentum. A fully coupled FSI approach was adopted to enhance the fidelity of the simulation. In this configuration, both fluid and solid solvers iteratively exchanged boundary condition data at each time step, enabling accurate modeling of the mutual influence betw een flow forces and leaflet deformation. To address challenges related to mesh distortion , d eformable domains were restricted to the areas surrounding the valve leaflets to minimize convergence issues while still capturing the relevant structural responses. Additionally, manually defined physics-based meshes were implemented to improve solution stability in regions with fine details, like the sinus and around the hinges. The", "solid solvers iteratively exchanged boundary condition data at each time step, enabling accurate modeling of the mutual influence betw een flow forces and leaflet deformation. To address challenges related to mesh distortion , d eformable domains were restricted to the areas surrounding the valve leaflets to minimize convergence issues while still capturing the relevant structural responses. Additionally, manually defined physics-based meshes were implemented to improve solution stability in regions with fine details, like the sinus and around the hinges. The versatility of the simulation environment allowed for various parameters to be systematically altered and evaluated. These parameters included pressure waveforms, leaflet material properties, and MHV geometry. The model was intentionally designed to be 23 geometry-independent, making it a general -purpose tool for simulating different types of mechanical heart valves. 3.5 Summary of Research Goal and Specific Aims The final goal of this research was to develop a robust and adaptable simulation platform for analyzing unsteady FSI in MHVs. The specific aims included: 1. Developing a 2D FSI model of a bi-leaflet MHV to evaluate fundamental performance and refine simulation settings (Feasibility phase). 2. Extending the model to 3D with more realistic geometrical and boundary conditions (Design Phase). 3. Model validation using laboratory data and a c onstructed statistical model based on available data within the All-of-Us research database to assess the reliability and integrity of the simulation results (validation phase). A summary of my research goal and specific aims is summarized in Table 3. Table 3. Summary of Research Goal and Overview of Specific Research Aims and Their Corresponding Deliverables Specific Aims Aim 1 Aim 2 Aim 3 Actions Creating 2D FSI setup in COMSOL Transferring the 2D set up into a 3D environment. 3D Model Validation Implement the setup on 2D bi-leaflet MHV Implement the setup on 3D bi-leaflet MHV Deliverables Reporting FSI default model parameters and allowed options (i.e. max. Reporting FSI default parameters and allowed options (i.e. max. Comparison of Simulation results with lab data from literature 24 geometrical parameters, material constraints) geometrical parameters, material constraints) Reporting velocity, pressure, and shear profiles Reporting velocity, pressure, and shear profiles Statistical analysis of available clinical data and comparison with simulation Results 25 4. 2D FSI Bi-Leaflet MHV Simulation Model In this chapter, the 2D FSI model is discussed. This 2D simulation served as a foundational step toward the construction of a more complex three-dimensional 3D model. The development of", "max. Comparison of Simulation results with lab data from literature 24 geometrical parameters, material constraints) geometrical parameters, material constraints) Reporting velocity, pressure, and shear profiles Reporting velocity, pressure, and shear profiles Statistical analysis of available clinical data and comparison with simulation Results 25 4. 2D FSI Bi-Leaflet MHV Simulation Model In this chapter, the 2D FSI model is discussed. This 2D simulation served as a foundational step toward the construction of a more complex three-dimensional 3D model. The development of the 2D model allowed for refinement of the simulation methodology, reduction in computational demands, and early validation of the modeling approach. The setup of the simulation model, along with the resulting findings, is detailed in the sections that follow. 4.1 Materials and Methods 4.1.1 Geometrical Parameters Currently, the Cardio Lab at SJSU uses a St Jude Regent bi-leaflet MHV for clotting testing using the MCL equipment as shown in Fig 3. This study aimed to develop a simulation environment that models the aortic root, incorporating a simplified bi -leaflet MHV similar to the model in the SJSU Cardio Lab Figure 3. Mock Circulation Loop in the Cardio Lab at SJSU The overall 2D model is shown in Fig. 4. The geometrical parameters are listed in Table 4 and follow the work in [ 45, 46]. The blood vessel diameter is assumed to be 25mm, and the total vessel length is 75mm. Table 4 summarizes the 2D aortic root geometrical parameters. 26 Figure 4. 2D Geometry for Aortic Root with Bi-leaflet MHV Model Table 4. Simulated Aortic Root Geometrical Parameters Parameter Value Parameter Value Root base1 22 mm Valve length 12.8 mm Root base2 20 mm Valve thickness 0.65 mm Vessel Diameter 25 mm Hinges diameter 0.1625 mm Sinus distance 30 mm Sinus height 37 mm The sinus curvature in the aortic root was modeled using the 3 -point quadratic Bezier function. The general form of the rational Bezier polygon function is recorded in [47], whereas the quadratic 3-point form is reported in Equation 1. It is worth noting that COMSOL has a built-in Bezier model that was used with the specific parameters summarized in Table 5. (1) 27 Table 5. Summary of Quadratic Bezier Polygon Parameters Used in Sinus Curve (x, z) weight Value P1 (Root base1, Vessel diameter) w1 1 P2 (sinus distance, sinus height) w2 3/sqrt (2) P3 (Root base2, Vessel diameter) w3 1 4.1.2 Fluid and Solid", "is recorded in [47], whereas the quadratic 3-point form is reported in Equation 1. It is worth noting that COMSOL has a built-in Bezier model that was used with the specific parameters summarized in Table 5. (1) 27 Table 5. Summary of Quadratic Bezier Polygon Parameters Used in Sinus Curve (x, z) weight Value P1 (Root base1, Vessel diameter) w1 1 P2 (sinus distance, sinus height) w2 3/sqrt (2) P3 (Root base2, Vessel diameter) w3 1 4.1.2 Fluid and Solid Domain Materials Parameters The simulated 2D domain includes the fluid (blood) domain and the solid (MHV) domain. The blood is represented using a Newtonian model with a viscosity of 0.0035 \ud835\udc43\ud835\udc4e. \ud835\udc60 and density of 1060 \ud835\udc58\ud835\udc54/\ud835\udc5a3. The MHV material properties are summarized in Table 6, matching the material properties of the St Jude Bi-leaflet MHV [27] and depicted in Fig.3. The MHV hinges are spaced by 3mm [48] and are considered rigid connectors to the MHV leaflets. The angle between the two leaflets is assumed to be 130 degrees, as depicted in Fig.3. Table 6. Summary of MHV Parameters Density Young\u2019s Modulus Poisson Ratio Yield Strength MHV Leaflets 2116 \ud835\udc58\ud835\udc54/\ud835\udc5a3 30.5\ud835\udc65109 \ud835\udc41/\ud835\udc5a2 0.3 407.7\ud835\udc65106 \ud835\udc41/\ud835\udc5a2 MHV Hinges 7805 \ud835\udc58\ud835\udc54/\ud835\udc5a3 205\ud835\udc65109 \ud835\udc41/\ud835\udc5a2 0.28 - 4.1.3 FSI Solver Setup COMSOL 6.2 is solely used as the Multiphysics FEM platform for configuring the FSI solver. The primary reason for selecting COMSOL is its integration of the CFD module, solid mechanics solver, and fluid -structure interaction (FSI) physics within a single Multiphysics solver using the ALE (moving mesh) technique and partitioned coupling method [49]. This capability simplifies the development of FSI models and allows seamless navigation between 28 different physics modules while maintaining control of the entire system. Within the FSI module, fluid flow can be characterized as laminar or turbulent, depending on Reynold\u2019s number. Under normal physiological conditions, blood flow through a healthy heart valve predominantly exhibits laminar behavior. However, during the end-systolic phase, as the valve closes, transient turbulence occurs around the sinus cavity as the velocity increases [50]. Furthermore, numerous research studies have used laminar [ 51\u201353] or turbulent [54\u201356] flow solvers in their simulations. While turbulence models provide higher accuracy in capturing finer vorticities, they require finer meshes and smaller time steps, increasing computational cost. A physics-controlled laminar flow FSI solver can approximate turbulence effects while significantly reducing computational time and capturing the dominant flow. The implementation of physics", "end-systolic phase, as the valve closes, transient turbulence occurs around the sinus cavity as the velocity increases [50]. Furthermore, numerous research studies have used laminar [ 51\u201353] or turbulent [54\u201356] flow solvers in their simulations. While turbulence models provide higher accuracy in capturing finer vorticities, they require finer meshes and smaller time steps, increasing computational cost. A physics-controlled laminar flow FSI solver can approximate turbulence effects while significantly reducing computational time and capturing the dominant flow. The implementation of physics -controlled meshing enhances the reliability of simulations by ensuring adaptive refinement in regions with significant spatia l fluid dynamic variations. Consequently, if turbulent flow characteristics prevail, the physics -controlled laminar flow solver will fail to achieve convergence, indicating the inappropriate selection of the governing physics [57]. In this work, the single -phase physics-controlled laminar flow module was used to solve Navier-Stokes equations and to set up the fluid dynamics. The fluid is assumed to be incompressible and flowing in fixed walls under non -slip conditions. The inlet and outlet boundary conditions are set to the ventricular and aortic pressure waveforms from [ 46]. In which measured data points have been recorded and fitted to an equation implementing the ventricular and aortic pressure waveforms. The measured aortic (outlet) and ventricular (inlet) pressure points are entered as input for a cubic piecewise interpolating function. The resulting 29 pressure plots are shown in Fig . 5. Moreover, it is essential to highlight that, at each time instance, the pressure difference governs the fluid\u2019s exerted force on the valve leaflets, resulting in their deformation per the fully coupled configuration within the FSI framework. A custom mesh is used with finer element size in the domain where the valve moves near the MHV leaflets and the sinus curve, while a coarser mesh size was utilized for the rest of the 3D domain to reduce the computational time. Figure 5. Pressure Waveforms at (a) Inlet/Ventricular and (b) Outlet/Aortic. On the other hand, solid mechanics physics was used to define the characteristics of the leaflets and hinges of the MHV solving the linear-elastic form of the Lagrangian equations. A rigid material is utilized with free rotation boundary conditions for the leaflet around the center of the hinge for each leaflet. This way the leaflets are free to rotate based on the external forces exerted by the fluid. 30 4.2 Results 4.2.1 Mesh Sensitivity Analysis To ensure the", "(b) Outlet/Aortic. On the other hand, solid mechanics physics was used to define the characteristics of the leaflets and hinges of the MHV solving the linear-elastic form of the Lagrangian equations. A rigid material is utilized with free rotation boundary conditions for the leaflet around the center of the hinge for each leaflet. This way the leaflets are free to rotate based on the external forces exerted by the fluid. 30 4.2 Results 4.2.1 Mesh Sensitivity Analysis To ensure the accuracy of the simulation, a mesh sensitivity analysis was conducted. Four different meshes were generated using varying element sizes: fine, finer, extra -fine, and extremely fine. A summary of the overall mesh structure, total number of ele ments, and element quality is presented in Fig. 6. Although decreasing the mesh element size allows for better resolution of the computed velocity fields, it also results in longer simulation times and increased computational resource demands. Figure 6. Summary of Mesh Sensitivity Analysis for the 2D COMSOL BMHV Model The deformation angle of the MHV leaflets for the four mesh cases is shown in Fig. 7. The simulations using the Fine and Finer meshes exhibited convergence issues during the valve closure phase. In contrast, the Extra -Fine and Extremely -Fine meshes produced consistent results, with a maximum deviation of approximately 3 degrees. As a resu lt, the Extra -Fine mesh was selected for use throughout the model development to optimize simulation time, while the extremely fine mesh was reserved for final verification purposes. 31 Figure 7. Leaflet Deformation Angle with Different Mesh Settings 4.2.2 2D Simulation Results The velocity profiles at three different time points \u2014corresponding to mid -opening, maximum-opening, and mid -closing valve positions \u2014are illustrated in Fig. 8. A video showing the dynamic valve deformation is included in the supplemental material (S1). During the systolic phase, the ventricular pressure exceeded the aortic pressure, resulting in the opening of the valve . As the MHV opened, three distinct jets were observed, as shown in Fig.8(a). 32 Figure 8. Velocity Profiles at Mid-opening, Max-opening, and Mid-closing Positions. During the diastolic phase, the aortic pressure exceeded the ventricular pressure, reversing the flow direction relative to the systolic phase and causing the valve to close, as shown in Fig. 8(c). An additional notable observation during this phase was the formation of four distinct vortices, illustrated in Fig. 9. Two vortices appeared between the valve tips", "As the MHV opened, three distinct jets were observed, as shown in Fig.8(a). 32 Figure 8. Velocity Profiles at Mid-opening, Max-opening, and Mid-closing Positions. During the diastolic phase, the aortic pressure exceeded the ventricular pressure, reversing the flow direction relative to the systolic phase and causing the valve to close, as shown in Fig. 8(c). An additional notable observation during this phase was the formation of four distinct vortices, illustrated in Fig. 9. Two vortices appeared between the valve tips and the central region of the aortic root, while the other two formed between the valve tips and the sinus wall. 33 Figure 9. Vorticity Profile at Two Different Time Instants During the Valve Closing Phase. To illustrate how the vorticity magnitude varied over time, a cut line was defined, as shown in Fig. 10(a), along which the vorticity distribution was analyzed. Figure 10(b) presents the vorticity magnitude plotted against the arc length of the cut line at various time instants. It\u2019s important to note that the valve closure appears to take a longer time than expected, which is attributed to the inlet and outlet pressure waveform which is fixed in the 3D simulations by using spline interpolation. Figure 10. (a) Cut Line at the Middle of Aortic Sinus, (b) Vorticity Magnitude at Different Time Instants. 34 5. 3D FSI Bi-leaflet MHV Simulation Model This chapter includes a comprehensive description of the 3D FSI solver setup, including the required modifications between the 2D and 3D models, the 3D simulation results and the validation of the simulation results. The 3D simulation results are cross-compared against two methods: (1) available experimental data in literature, and (2) clinical data analyzed in this work based on All of Us research data base. 5.1 Materials and Methods 5.1.1 3D COMSOL Simulation Setup The sinus curve was drawn in the x -z plane then the sweep function was used around the x-axis to generate the sinus 3D model in Fig. 10. The aortic root are modeled with cylinders in 3D (instead of the rectangles used in 2D model). The same geometrical parameters of the St Jude bi-leaflet MHV is used to model the valve leaflets dimensions and material. The ove rall 3D model is depicted in Fig. 11. Figure 11. Simulated 3D Aortic Root with MHV Model in COMSOL. Similar to the 2D model, the inlet and outlet boundary conditions are set to the ventricular and", "3D model in Fig. 10. The aortic root are modeled with cylinders in 3D (instead of the rectangles used in 2D model). The same geometrical parameters of the St Jude bi-leaflet MHV is used to model the valve leaflets dimensions and material. The ove rall 3D model is depicted in Fig. 11. Figure 11. Simulated 3D Aortic Root with MHV Model in COMSOL. Similar to the 2D model, the inlet and outlet boundary conditions are set to the ventricular and aortic pressure waveforms as described in Chapter 4. All side walls are also set to non-slip 35 boundary conditions. Moreover, a custom mesh is employed for the 3D model with finer element size around the valve and sinus and a relatively coarse mesh elsewhere. 5.1.2 3D FSI Solver Setup A fully-coupled FSI module is used as the multi -physics solver for the 3D simulations. These simulations were carried on San Jose State University (SJSU) Computer Science Lab machines. Due to limited computing capabilities of the machine, attempting to simulate the entire time domain causes simulation challenges due to memory. To overcome th is issue, the entire time domain is decomposed into slots of 0.15 s and a time -domain sliding window simulation is implemented as shown in Fig. 12. In this work, the solution from each preceding window serves as the initial condition for the subsequent window. Overlapping between consecutive windows is introduced to mitigate artificial numerical discontinuities [58]. This sliding window mechanism enables the simulation to progress across time slots, ultimately covering the complete time domain of the pressure signals. Figure 12. Simulated Time-Domain Aortic/Ventricular Pressure Signals with Sliding Time Window 36 5.2 Results 5.2.1 3D COMSOL Simulation Results The profiles of the resultant velocity (left), vorticity (middle), and wall shear stress (right) are presented in Fig. 13 and 14, illustrating their evolution over time in the X-Z and Y-Z planes, respectively. Two distinct time instants are analyzed, capturing the dynamic behavior of the valve when it is fully-open (top) and fully-closed (bottom). Figure 13. Simulated Velocity (left), Vorticity (middle) and VSS (right) Profiles in the XZ Plane at y=0. The sequence begins with the valve in a fully closed position at the end -diastolic phase, followed by a gradual opening at the time period when the ventricular pressure exceeds the aortic pressure. Subsequently, the pressure gradient across the valve continues to increase, leading to the valve reaching its", "the dynamic behavior of the valve when it is fully-open (top) and fully-closed (bottom). Figure 13. Simulated Velocity (left), Vorticity (middle) and VSS (right) Profiles in the XZ Plane at y=0. The sequence begins with the valve in a fully closed position at the end -diastolic phase, followed by a gradual opening at the time period when the ventricular pressure exceeds the aortic pressure. Subsequently, the pressure gradient across the valve continues to increase, leading to the valve reaching its maximum opening, coinciding with the peak ventricular-aortic pressure gradient during the systolic phase. Subsequently, the valve undergoes a progressive closure, returning to its fully closed state by the end-systolic phase, marking the transition into the next diastolic phase. 37 Figure 14. Simulated Velocity (left), Vorticity (middle), and VSS (right) Profiles in the YZ Plane at x=22mm At peak systole when the the valve reaches its fully open state, the velocity decreases due to flowing through a larger cross-sectional area. Simultaneously, the fluid flow exhibits three distinct jet formations, as shown in Figures 13(a) and 14(a), two outer jets and one central jet. In the second half of the cycle, the aortic pressure is higher than the ventricular pressure , forcing the valve closure. The vorticity magnitude increases during the diastolic phase when the aortic valve is fully closed due to the enforced backflow and recirculation of the fluid imposed by the leaflet walls as shown in Fig. 13(e), concurrently, the wall shear stress reaches its highest value during the same time due to these dynamics as shown in Fig. 13(f). To demonstrate the velocity profile in the third plane (XY), three different cut planes are used as depicted in Fig. 15 left at distances equal to 0, 4mm, and 6.25mm away from the center of the aortic root. The velocity profiles are reported when the valve is fully open. The velocity profile of the central jet is shown in Fig. 15 (a) and the outer jet is shown in Fig. 15 (c). Both profiles seem similar with close maximum velocities. In Fig. 15 (b), the cut plane is placed as 38 close as possible to the top leaflet of the MHV, and it can show some interaction between the outer and central jets as visualized also in Fig13 (a). Figure 15. Simulated Velocity Profiles in the XY Plane at (a) y=0mm, (b) y=4mm, and (c) y=6.25mm 5.2.2 Validations of 3D", "15 (a) and the outer jet is shown in Fig. 15 (c). Both profiles seem similar with close maximum velocities. In Fig. 15 (b), the cut plane is placed as 38 close as possible to the top leaflet of the MHV, and it can show some interaction between the outer and central jets as visualized also in Fig13 (a). Figure 15. Simulated Velocity Profiles in the XY Plane at (a) y=0mm, (b) y=4mm, and (c) y=6.25mm 5.2.2 Validations of 3D Model CFD Simulation In this section, we will cross -compare our 3D model CFD simulation results against the experimental data available in the literature. Additionally, statistical analysis of clinical data 39 in the All of Us research database is carried out for further comparison against the simulation results of the proposed model. 5.2.2.1 Comparison of 3D Simulation Results with Experimental Data To validate the Fluid -Structure Interaction (FSI) model, the simulation results were systematically compared against experimental laboratory data corresponding to both the diastolic and systolic phases of the cardiac cycle. During the diastolic phase, when the valve is in a closed position, the time-averaged leakage backflow in the St. Jude bi -leaflet mechanical heart valve was experimentally measured using two -dimensional (2D) Particle Image Velocimetry (PIV), as reported in [59]. Klusak et al. [59] conducted detailed PIV measurements to quantify key hemodynamic parameters, including the time -averaged velocity and shear stress, at a spatial location 86 \u00b5m from the valve hinge, with a measurement resolution of 167 \u00b5m x 167\u00b5m. According to their experimental data, the maximum velocity recorded in the vicinity of the valve hinge was 5.03 m/s, while the maximum shear stress reached 87 Pa, and the maximum vorticity was reported as 29,300 1/s. In comparison, the computational results obtained from the Computational Fluid Dynamics (CFD) simulation under the same conditions predicted a maximum velocity of 4.37 m/s, a maximum shear stress of 115 Pa, and a maximum vorticity of 35,630 1/s, as illustrated in Fig. 16. These values indicate a reasonable agreement between the numerical model and experimental observations, demonstrating the model\u2019s capability in capturing critical flow 40 characteristics, albeit with some discrepancies likely attributable to computational assumptions and the under estimation of the turbulence due to using laminar flow physics. Figure 16. (a) XY Plane at 86\u00b5m Away from the Valve Hinge, (b) Visualization of Leakage Flow in the XZ Plane. Furthermore, the Effective Orifice", "maximum vorticity of 35,630 1/s, as illustrated in Fig. 16. These values indicate a reasonable agreement between the numerical model and experimental observations, demonstrating the model\u2019s capability in capturing critical flow 40 characteristics, albeit with some discrepancies likely attributable to computational assumptions and the under estimation of the turbulence due to using laminar flow physics. Figure 16. (a) XY Plane at 86\u00b5m Away from the Valve Hinge, (b) Visualization of Leakage Flow in the XZ Plane. Furthermore, the Effective Orifice Area (EOA) of the bi-leaflet mechanical heart valve was estimated using the Gorlin equation [60], as expressed in Equation 2. (2) This widely used empirical equation provides a quantitative assessment of the functional valve area based on pressure gradients and volumetric flow rates. To determine the outward mean flow rate, calculations were performed at the root annulus, as illustrated in Figure 17. Based on these computations, the resulting EOA was determined to be 2.34 cm\u00b2. 41 Figure 1717. Fluid Flow Results at the Vessel Root Annulus for EOA Calculations. For comparative validation, the standard clinically measured EOA of a 25 mm St. Jude mechanical heart valve (MHV) has been reported in the literature as 2.5 cm\u00b2 [61], based on clinical data collected from a cohort of 34 participants. This cross -comparison suggests that the computationally derived EOA closely aligns with clinically observed values, further reinforcing the applicability of the simulation methodology. 5.2.2.2 Statistical Analysis of Supporting Clinical Data The available clinical data in the All of US Research database [ 62] are used to cross - compare with the simulated CFD results. Clinical data for two parameters are extracted to be analyzed including: average max. systolic velocity, and mean systolic pressure gradient. These parameters are measured using standard medical t echniques. The max. systolic velocity is commonly measured using continuous wave Doppler ultrasound imaging, where the probe is placed at the apex of the heart using an apical -5 chamber(A5C) [63]. The exact measurement point varies based on patients, there fore clinicians manually place the probe parallel to the aortic blood jet flow until they observe the maximum velocity over the systolic period. 42 Moreover, the mean systolic pressure gradient is measured using full Bernoulli method using Equation (3): ) (3) where V1 is the velocity at proximal point (before the valve), typically 3-10mm away from the valve annulus [ 64] and V2 is the velocity at distal point (after", "[63]. The exact measurement point varies based on patients, there fore clinicians manually place the probe parallel to the aortic blood jet flow until they observe the maximum velocity over the systolic period. 42 Moreover, the mean systolic pressure gradient is measured using full Bernoulli method using Equation (3): ) (3) where V1 is the velocity at proximal point (before the valve), typically 3-10mm away from the valve annulus [ 64] and V2 is the velocity at distal point (after the valve) with max. systolic velocity downstream [65]. In this work, two datasets are created for the max. systolic velocity, and mean pressure gradient based on the available clinical data within All of US database. The summary of the cohort selected for each of the datasets is reported in Table 7. The statistical analysis of the clinical data is done using R code and analyzed within All of US research workbench cloud computing resources. Table 7. Cohort Summary for Clinical Datasets based on All of US Research Database Max velocity Cohort Pressure gradient Cohort Total count 358 2275 Gender Female Male 238 120 1410 865 Age 18-39 40-59 60-79 12 110 218 244 748 1137 +80 18 146 Race White Black Asian 313 14 2 1275 637 147 Other 31 216 43 5.2.2.3 Comparison of 3D Simulation Results and Clinical Data To further validate the CFD results, the available clinically measured data for the maximum systolic velocity from All of US research database were analyzed for total of 358 participants. The cohort of the maximum systolic velocity data is summarized in Ta ble 7. The scatter plot of the max. systolic velocity data is depicted in Fig. 18(a). The average max systolic velocity based on the clinical data was reported at 1.639 m/s as depicted with the blue line on Fig. 18(a). On the other hand, the CFD simulation estimated a maximum systolic velocity of 1.11 m/s at t=0.16s showing good level of correlation with the clinical data. Moreover, the simulated instantaneous systolic pressure gradient is calculated as \u2206 P(t) = P2(t)\u2212P1(t) where P1(t) measured at 5mm away from the valve annulus (upper stream) and P2(t) is measured at the cross-section wher e maximum velocity is observed. Afterwards, the time -average systolic pressure gradient is calculated and found to be 3.235 mmHg. Figure 1818. Summary of Clinical Data Available on All of US Research for (a) Maximum Systolic Velocity, (b)", "good level of correlation with the clinical data. Moreover, the simulated instantaneous systolic pressure gradient is calculated as \u2206 P(t) = P2(t)\u2212P1(t) where P1(t) measured at 5mm away from the valve annulus (upper stream) and P2(t) is measured at the cross-section wher e maximum velocity is observed. Afterwards, the time -average systolic pressure gradient is calculated and found to be 3.235 mmHg. Figure 1818. Summary of Clinical Data Available on All of US Research for (a) Maximum Systolic Velocity, (b) Average Systolic Pressure Gradient 44 Similar to the clinical velocity data analysis, the average systolic pressure gradient clinical data for 2275 participants were compared with the CFD results. The summary of the clinical pressure gradient data is shown in Fig. 18(b), where the average systolic pressure gradient is reported at 4.614 mmHg. A summary of the simulated CFD results and the clinical data is reported in Table 8. Overall, the simulation results show good agreement with the clinical data in terms of EOA, Max systolic velocity, and mea n pressure gradient as summarized in Table 8. Table 8. Comparison of CFD Simulation Results versus Clinical Data Max Systolic Velocity Avg Systolic \u2206P Orifice Area Clinical Data 1.639 m/s (this work) 4.614 mmHg (this work) 2.5cm2 [58] This work (CFD) 1.11 m/s 3.235 mmHg 2.34cm2 It is important to mention that the most relevant clinical data requires surveying healthy participants with St Jude Regent Bi -leaflet MHV. Given that this data is not specifically available, the primary benefit of using the clinical cohort data is that it provides a realistic physiological boundary for comparison, even if not all subjects had MHVs, the data reflects the boundary range of flow and pressure conditions observed in human aortic valves under systolic conditions. I excluded the population with any heart -valve related complication or heart failure. By comparing the 3D CFD simulation outcomes with these clinical measurements, we ensure that the predicted velocities and pressure gradients fall within clinically observed limits, thereby enhancing the physiological relevance and plausibility of the simulation results. This approach allows for a sanity check \u2014helping to confirm that the 45 simulated MHV behaves within the expected clinical spectrum, even if the dataset is not MHV- specific. It serves as a preliminary validation step, especially useful when direct MHV-specific in vivo data is limited or unavailable. 46 6. Limitations A significant limitation encountered in this study was the restricted availability", "pressure gradients fall within clinically observed limits, thereby enhancing the physiological relevance and plausibility of the simulation results. This approach allows for a sanity check \u2014helping to confirm that the 45 simulated MHV behaves within the expected clinical spectrum, even if the dataset is not MHV- specific. It serves as a preliminary validation step, especially useful when direct MHV-specific in vivo data is limited or unavailable. 46 6. Limitations A significant limitation encountered in this study was the restricted availability of computational resources. Specifically, the simulations were executed on a workstation located in the SJSU College of Engineering Laboratory, equipped with an Intel Xeon E 5-1603 processor operating at 2.8 GHz and 32 GB of RAM. In contrast, the study presented in [ 66] (2024), which also focused on BMHVs, utilized a high -performance computing (HPC) environment. Their simulations were performed on a supercomputing platform employing 200 to 640 parallel processing units, each node featuring Intel Xeon Platinum 8268 processors (2.9 GHz) and 960 GB of memory. This disparity in computational capacity significantly affects the achievable resolution, complexity, and efficiency of FSI simulations. Consequently, future work necessitates access to SJSU\u2019s HPC infrastructure, along with the installation of the COMSOL CFD module on HPC nodes, to enable high -fidelity modeling and improved simulation performance . Such limitation in the computational resources necessitated the segmentation of the simulation domain into multiple discrete sections. Instead of performing a fully coupled simulation across the entire domain simultaneously, each segment was simulated individually, with the output of one segment serving as the initial condition for the subsequent segment. While this approach enabled the successful execution of the simulations within available computational capabilities, it may have introduced minor discontinuities in flow characteristics between segments, potentially affecting the accuracy of transient flow predictions. 47 Furthermore, the restricted computational power also limited the implementation of high - fidelity turbulence models, which are essential for accurately capturing the complex flow dynamics, particularly the formation and dissipation of vortices , as we expect turbulent -like features developing in late systole, decelerating flow. As a result, certain lateral vortices within the flow field were likely underestimated, which may have led to deviations from fully resolved experimental or high -resolution numerical results. Incorporating more advanced turbulence modeling techniques, such as L arge Eddy Simulation (LES) or Reynolds -Averaged Navier- Stokes (RANS) models, in future studies would enhance the accuracy of the predicted", "flow dynamics, particularly the formation and dissipation of vortices , as we expect turbulent -like features developing in late systole, decelerating flow. As a result, certain lateral vortices within the flow field were likely underestimated, which may have led to deviations from fully resolved experimental or high -resolution numerical results. Incorporating more advanced turbulence modeling techniques, such as L arge Eddy Simulation (LES) or Reynolds -Averaged Navier- Stokes (RANS) models, in future studies would enhance the accuracy of the predicted flow structures. The computational ca pabilities also limited the implementation of modeling the blood as a Newtonian fluid. Non-Newtonian fluid models are available in COMOSL but require additional licenses. Additionally, while the primary objective of this study was to develop a robust and computationally efficient FSI solver, rather than to optimize or refine mechanical heart valve (MHV) design, the simulations were nonetheless conducted on a simplified valve geometry on an axisymmetric aortic model with a fixed hinge model . To demonstrate the applicability of our FSI solver with complex aortic root model, we simulated with three plugs aortic sinus for the opening phase only (0.15-0.18s) and an animated video is reported in the supplemental material SV2. More realistic available hinge designs usually contain cavities where flow tends to stagnate, and therefore, accounting for the hinge flow allows for thrombogenicity testing; such an effect is not considered in this study. Additionally, no-slip condition was used to model the aortic root walls ; in reality , the vessel wall need s to be modeled as viscoelastic walls allowing for wall deformation and fluid-wall interaction . Although this simplification 48 facilitated computational feasibility and model validation, future work should consider more anatomically and physiologically realistic valve geometries to further improve the applicability of the model to real-world cardiovascular scenarios. Lastly, the supporting clinical data utilized in this study were derived from the All of Us research database, which presented certain inherent limitations. The dataset included cohorts comprising both healthy and unhealthy participants, with the exclusion of indiv iduals diagnosed with heart failure. For future investigations, it is recommended to stratify healthy participants with bi -leaflet mechanical heart valves separately to facilitate more precise comparative analyses. 49 7. Conclusions This study successfully developed a three -dimensional (3D) Fluid -Structure Interaction (FSI) model using COMSOL to investigate the complex, unsteady flow dynamics within a bi- leaflet mechanical heart valve (MHV). The model effectively simulated and", "dataset included cohorts comprising both healthy and unhealthy participants, with the exclusion of indiv iduals diagnosed with heart failure. For future investigations, it is recommended to stratify healthy participants with bi -leaflet mechanical heart valves separately to facilitate more precise comparative analyses. 49 7. Conclusions This study successfully developed a three -dimensional (3D) Fluid -Structure Interaction (FSI) model using COMSOL to investigate the complex, unsteady flow dynamics within a bi- leaflet mechanical heart valve (MHV). The model effectively simulated and analyze d critical hemodynamic parameters, including velocity distributions, vorticity fields, shear stress variations, and pressure gradients, providing a comprehensive representation of the flow characteristics within the valve. The accuracy of the computational model was validated through comparisons with previously published experimental data, demonstrating strong agreement and reinforcing the model\u2019s reliability for capturing essential flow phenomena associated with MHV function. A key strength of this study lies in the balance between model complexity and computational efficiency. The framework was designed to incorporate sufficient physiological detail while maintaining a simplified and adaptable structure, facilitating ease of m odification and expansion for future studies. This adaptability makes the model a valuable tool for researchers and engineers seeking to explore different valve designs, assess material properties, and refine simulation parameters to optimize cardiovascular device performance. Despite these advancements, the study faced computational limitations that influenced certain aspects of the simulation. Due to constraints in computational power, the simulation domain was segmented into multiple subdomains, requiring each segment to be s imulated individually, with the results of one segment serving as the initial conditions for the next. While 50 this approach allowed for feasible simulation execution, it introduced potential numerical discontinuities. Additionally, computational constraints limited the implementation of high - fidelity turbulence models, potentially leading to underestimations of la teral vortices and smaller-scale turbulent structures. These limitations suggest that future work could benefit from enhanced computational resources to enable full -domain simulations and the incorporation of advanced turbulence modeling techniques, improving the accuracy of flow predictions. Nevertheless, this research represents a significant step forward in bridging the gap between experimental and computational analyses in the field of MHV design. By providing a validated and adaptable numerical framework, this study contributes to the ongoing advancements in computational cardiov ascular engineering, paving the way for future investigations into patient-specific valve designs, improved prosthetic heart valve performance, and more physiologically realistic simulations of cardiac flow dynamics. 51 References", "of advanced turbulence modeling techniques, improving the accuracy of flow predictions. Nevertheless, this research represents a significant step forward in bridging the gap between experimental and computational analyses in the field of MHV design. By providing a validated and adaptable numerical framework, this study contributes to the ongoing advancements in computational cardiov ascular engineering, paving the way for future investigations into patient-specific valve designs, improved prosthetic heart valve performance, and more physiologically realistic simulations of cardiac flow dynamics. 51 References 1. Roth, G. A., G. A. Mensah, C. O. Johnson, G. Addolorato, E. Ammirati, L. M. Baddour, N. C. Barengo, A. Z. Beaton, E. J. Benjamin, C. P. Benziger, A. Bonny, M. Brauer, M. Brodmann, T. J. Cahill, J. Carapetis, A. L. Catapano, S. S. Chugh, L. T. Cooper, J. Coresh, and M. Criqui, Global burden of cardiovascular diseases and risk factors, 1990 -2019: update from the GBD 2019 study. J Am Coll Cardiol , 76(25): 2982\u20133021, 2020 . https://doi.org/10.1016/j.jacc.2020.11.010 2. Virani, S. S., A. Alonso, H. J. Aparicio, E. J. Benjamin, M. S. Bittencourt, C. W. Callaway, A. P. Carson, A. M. Chamberlain, S. Cheng, F. N. Delling, M. S. Elkind, K. R. Evenson, J. F. Ferguson, D. K. Gupta, S. S. Khan, B. M. Kissela, K. L. Knutson, C. D. Lee, T. T. Lewis, and J. Liu, Heart disease and stroke statistics-2021 update: a report from the american heart association. Circulation, 143(8): e254\u2013e743, 2021. https://doi.org/10.1161/CIR.0000000000000950 3. Nkomo, V. T., J. M. Gardin, T. N. Skelton, J. S. Gottdiener, C. G. Scott, and M. Enriquez- Sarano, Burden of valvular heart diseases: a population -based study. Lancet, 368(9540): 1005\u20131011, 2006. https://doi.org/10.1016/S0140-6736(06)69208-8 4. Otto, C. M. and B. Prendergast. Aortic-valve stenosis--from patients at risk to severe valve obstruction. N Engl J Med , 371(8): 744\u2013756,2014. https://doi.org/10.1056/NEJMra1313875 5. Abramov, A., Y. Xue, A. Zakharchenko, M. Kurade, R. K. Soni, R. J. Levy, and G. Ferrari, Bioprosthetic heart valve structural degeneration associated with metabolic syndrome: Mitigation with polyoxazoline modification, Proc Natl Acad Sci USA; 120(1), 2023. doi: 10.1073/pnas.2219054120. 6. Runyu X., H. Xun, X. Tianyuan, C. Mao, J. R. Williams, and L. Pengzhi, Numerical investigation of calcification effects on aortic valve motions and ambient flow characteristics, J Fluids Struct , 124(104014):0889-9746, 2024. https://doi.org/10.1016/j.jfluidstructs.2023.104014. 7. Lawrence N. S., S. Rolland, D. J. Blundon, M. A. Deutsch, T. R. P. Martin, W. J. Dutton, E. M. Kolahdouz, and B. E. Griffith, Canary in the cardiac-valve", "metabolic syndrome: Mitigation with polyoxazoline modification, Proc Natl Acad Sci USA; 120(1), 2023. doi: 10.1073/pnas.2219054120. 6. Runyu X., H. Xun, X. Tianyuan, C. Mao, J. R. Williams, and L. Pengzhi, Numerical investigation of calcification effects on aortic valve motions and ambient flow characteristics, J Fluids Struct , 124(104014):0889-9746, 2024. https://doi.org/10.1016/j.jfluidstructs.2023.104014. 7. Lawrence N. S., S. Rolland, D. J. Blundon, M. A. Deutsch, T. R. P. Martin, W. J. Dutton, E. M. Kolahdouz, and B. E. Griffith, Canary in the cardiac-valve coal mine: Flow velocity and inferred shear during prosthetic valve closure \u2013 predictors of blood damage and clotting, bioRxiv 2022; doi: https://doi.org/10.1101/2022.06.23.497372 8. Mohd Salleh, N., M. S. Zakaria, and M. J. Abd Latif, Reducing of thrombosis in mechanical heart valve through the computational method: a review. J. Adv. Res. Fluid Mech. Therm. Sci., 65(2) :178\u2013200, (2020). Retrieved from https://akademiabaru.com/submit/index.php/arfmts/article/view/2787 52 9. Andrea D. M., D. Aldo, T. Gaetano, and B. Uberto, Diamond anniversary of mechanical cardiac valve prostheses: a tale of cages, balls, and discs, Ann Thorac Surg, 110(4): 1427- 1433, 2020. https://doi.org/10.1016/j.athoracsur.2020.04.143 10. Pawlikowski M., and A. Nieroda. Comparative analyses of blood flow through mechanical trileaflet and bileaflet aortic valves, Research Square, PREPRINT (Version 1) , 2021 https://doi.org/10.21203/rs.3.rs-1121085/v1 11. Li C. P., and P. C. Lu, Numerical comparison of the closing dynamics of a new trileaflet and a bileaflet mechanical aortic heart valve. J Artif Organs 15 : 364\u2013374, 2012. https://doi.org/10.1007/s10047-012-0650-8 12. Schubert K., T. Schaller, E. Stojenthin, C. Stephan, H. Sievers, and M. Scharfschwerdt, A novel trileaflet mechanical heart valve: first in vitro results, Interact Cardiovasc Thorac Surg, 28(5): 689-694, May 2019. https://doi.org/10.1093/icvts/ivy337 13. Xu K. W., Q. Gao, M. Wan, and K. Zhang K. Mock circulatory loop applications for testing cardiovascular assist devices and in vitro studies. Front Physiol. 14:1175919, 2023. doi:10.3389/fphys.2023.1175919. PMID: 37123281; PMCID: PMC10133581. 14. Chen A ., A. A. Basri, N. B. Ismail, K. A. Ahmad. The numerical analysis of non- Newtonian blood flow in a mechanical heart valve. Processes. 11(1):37, 2023. https://doi.org/10.3390/pr11010037 15. Ha, S.T., T. D. Nguyen, V. C. Vu, M. H. Nguyen, and M. D. Nguyen. A study of fluid- structure interaction of unsteady flow in the blood vessel using finite element method. In: Tien Khiem, N., Van Lien, T., Xuan Hung, N. (eds) Modern Mechanics and Applications. Lecture Notes in Mechanical Engineering. Springer, 2022. https://doi.org/10.1007/978981- 16-3239-6_85 16. Toma M., S. Singh-Gryzbon, E. Frankini, Z. Wei, and A. P. Yoganathan.", "flow in a mechanical heart valve. Processes. 11(1):37, 2023. https://doi.org/10.3390/pr11010037 15. Ha, S.T., T. D. Nguyen, V. C. Vu, M. H. Nguyen, and M. D. Nguyen. A study of fluid- structure interaction of unsteady flow in the blood vessel using finite element method. In: Tien Khiem, N., Van Lien, T., Xuan Hung, N. (eds) Modern Mechanics and Applications. Lecture Notes in Mechanical Engineering. Springer, 2022. https://doi.org/10.1007/978981- 16-3239-6_85 16. Toma M., S. Singh-Gryzbon, E. Frankini, Z. Wei, and A. P. Yoganathan. Clinical impact of computational heart valve models. Materials. 15(9):3302 , 2022. https://doi.org/10.3390/ma15093302 17. Kodigepalli K. M., K. Thatcher, T. West, D. P. Howsmon, F. J. Schoen, M. S. Sacks, C. K. Breuer, J. Lincoln, Biology and biomechanics of the heart valve extracellular matrix. J Cardiovasc Dev Dis. 7(4):57,2020. https://doi.org/10.3390/jcdd7040057 18. Fioretta E.S., S. E. Motta, V. Lintas, et al. Next-generation tissue-engineered heart valves with repair, remodelling, and regeneration capacity. Nat Rev Cardiol 18 : 92\u2013116, 2021. https://doi.org/10.1038/s41569-020-0422-8 19. Faggioli M., A. Moro, S. Butt, M. Todesco, D. Sandrin, G. Borile, A. Bagno, A. Fabozzo, F. Romanato, M. Marchesan, et al. A new decellularization protocol of porcine aortic valves using tergitol to characterize the scaffold with the biocompatibility profile using 53 human bone marrow mesenchymal stem cells. Polymers. 14(6):1226 , 2022. https://doi.org/10.3390/polym14061226 20. Gotadki R., Prosthetic Heart Valve Market Research Report Forecast to 2035. ID: MRFR/HC/41650-HCR. April 2025 Source: https://www.marketresearchfuture.com/reports/prosthetic-heart-valve-market-43316 21. Mordor Intelligence . (2025). Prosthetic Heart Valve Market Size & Share Analysis - Growth Trends & Forecasts (2025 - 2030) Source: https://www.mordorintelligence.com/industry-reports/prosthetic-heart-valve- market 22. Qian J. Y., Z. X. Gao, C. W. Hou, et al. A comprehensive review of cavitation in valves: mechanical heart valves and control valves. Bio -des. Manuf. 2 : 119\u2013136, 2019.https://doi.org/10.1007/s42242-019-00040-z 23. Carrel T., P. R. Vogt, D. Obrist, and H. Schaff. Evolving technology: the TRIFLO tri - leaflet mechanical valve without oral anticoagulation: a potential major innovation in valve surgery. Front Cardiovasc Med.10:1220633 , 2023. doi : 10.3389/fcvm.2023.1220633. PMID: 37840955; PMCID: PMC10570810. 24. Kumar T., A. Singh, S. Thakre, S. Acharya, S. Shukla, and S. Kumar. Scientific evolution of artificial heart valves: a narrative review. Cureus. 15(7):e42131 , 2023. doi: 10.7759/cureus.42131. PMID: 37602004; PMCID: PMC10438674. 25. De Martino A., A. D. Milano, M. D. Barbera, G. Thiene, and U. Bortolotti; The caged-ball prosthesis 60 years later: a historical review of a cardiac surgery milestone. Tex Heart Inst J, 49 (2): e207267, 2022. doi: https://doi.org/10.14503/THIJ-20-7267 26.", "doi : 10.3389/fcvm.2023.1220633. PMID: 37840955; PMCID: PMC10570810. 24. Kumar T., A. Singh, S. Thakre, S. Acharya, S. Shukla, and S. Kumar. Scientific evolution of artificial heart valves: a narrative review. Cureus. 15(7):e42131 , 2023. doi: 10.7759/cureus.42131. PMID: 37602004; PMCID: PMC10438674. 25. De Martino A., A. D. Milano, M. D. Barbera, G. Thiene, and U. Bortolotti; The caged-ball prosthesis 60 years later: a historical review of a cardiac surgery milestone. Tex Heart Inst J, 49 (2): e207267, 2022. doi: https://doi.org/10.14503/THIJ-20-7267 26. Khellaf B., and B. Boussad, Computational hemodynamic investigation of a new bileaflet mechanical heart valve. SIMULATION , 96(5):459-469, 2020. doi:10.1177/0037549719886364 27. Wium E., Design, manufacture and hydrodynamic evaluation of a prototype trileaflet mechanical heart valve for aortic valve replacement (Doctoral dissertation, Stellenbosch: Stellenbosch University), 2020. 28. Myers M. R., and J. M. Porter, Impulsive-motion model for computing the closing motion of mechanical heart -valve leaflets. Ann Biomed Eng. 31(9):1031-9, 2003. doi: 10.1114/1.1603750. PMID: 14582606. 29. Borazjani I., A review of fluid-structure interaction simulations of prosthetic heart valves. J Long Term Eff Med Implants. 25(1-2):75-93, 2015. doi: 10.1615/jlongtermeffmedimplants.2015011791. PMID: 25955008. 54 30. Abbas S . S., M. S. Nasif, and R. Al-Waked, State-of-the-art numerical fluid \u2013structure interaction methods for aortic and mitral heart valves simulations: A review. SIMULATION. 98(1):3-34, 2022. doi:10.1177/00375497211023573 31. De Vita F., M. D. De Tullio and R. Verzicco, Numerical simulation of the non-Newtonian blood flow through a mechanical aortic valve. Theor Comput Fluid Dyn 30(1 \u20132): 129\u2013 138, 2016. 32. Hedayat M, H. Asgharzadeh and I. Borazjani, Platelet activation of mechanical versus bioprosthetic heart valves during systole. J Biomech 56: 111\u2013116, 2017. 33. Abbas S. S., M. S.Nasif MS, R. Al-Waked R and M. A Said, Numerical investigation on the relationship of cavitation initiation in bileaflet mechanical heart valves (BMHVs) with the aortic root geometry and valve\u2019s implantation rotation angle. J Braz Soc Mech Sci 42(1): 1\u20137, 2020. 34. Li W. Q., Z. X. Gao, Z. J. Jin and J. Y. Qian, Transient study of flow and cavitation inside a bileaflet mechanical heart valve. Appl Sci 10(7): 2548, 2020. 35. Kadhim S . K., M. S. Nasif, H. H. Al-Kayiem and R. Al-Waked, Computational fluid dynamics simulation of blood flow profile and shear stresses in bileaflet mechanical heart valve by using monolithic approach. Simulation 94(2): 93\u2013104, 2018. 36. Sakamura Y. , K. Nakayama, and M. Oshima, Numerical simulation of shock -induced motion of a cuboidal solid body", "Jin and J. Y. Qian, Transient study of flow and cavitation inside a bileaflet mechanical heart valve. Appl Sci 10(7): 2548, 2020. 35. Kadhim S . K., M. S. Nasif, H. H. Al-Kayiem and R. Al-Waked, Computational fluid dynamics simulation of blood flow profile and shear stresses in bileaflet mechanical heart valve by using monolithic approach. Simulation 94(2): 93\u2013104, 2018. 36. Sakamura Y. , K. Nakayama, and M. Oshima, Numerical simulation of shock -induced motion of a cuboidal solid body using the overset grid functionality of Open-FOAM. Shock Waves. 31: 1-13, 2021. 10.1007/s00193-021-01035-5 37. Sodhani S. R., A. Aksenov, S. So\u011fanci, S. Jockenh\u00f6vel, P. Mela, and S. E. Stapleton, Fluid- structure interaction simulation of artificial textile reinforced aortic heart valve: Validation with an in -vitro test, J Biomech , 78 : 52-69, 2018. https://doi.org/10.1016/j.jbiomech.2018.07.018. 38. Lim W. Z., R. Y. Xiao, T. Hong and C. S. Chin, Partitioned methods in computational modelling on fluid \u2013structure interactions of concrete gravity -dam. Comput Informat Sci 6(4): 154, 2013. 39. Wald S., A. Liberzon and I. Avrahami, A numerical study of the hemodynamic effect of the aortic valve on coronary flow. Biomech Mod Mechanobiol 17(2): 319\u2013338, 2018. 40. Lavon K., R. Halevi, G. Marom, et al. Fluid\u2013structure interaction models of bicuspid aortic valves: The effects of nonfused cusp angles. J Biomech Eng, 140(3): 1\u20137, 2018. 41. Stijnen J. M., J. De Hart, P. H. Bovendeerd and F. N. Vosse, Evaluation of a fictitious domain method for predicting dynamic response of mechanical heart valves. J Fluids Struct 19(6): 835\u2013850, 2004. 55 42. P. B. Ryzhakov, R. Rossi, S. R. Idelsohn and E. Onate, A monolithic Lagrangian approach for fluid\u2013structure interaction problems. Comput Mech, 46(6): 883\u2013899, 2010. 43. Bornoff J., H.S. Gill, A. Najar, I.L. Perkins, A.N. Cookson and K.H. Fraser , Overset meshing in combination with novel blended weak -strong fluid-structure interactions for simulations of a translating valve in series with a second valve, Comput Methods Biomech Biomed Engin, 2023. DOI: 10.1080/10255842.2023.2199903 44. Morany A., K. Lavon, R. G. Bardon, et al. Fluid \u2013structure interaction modeling of compliant aortic valves using the lattice Boltzmann CFD and FEM methods. Biomech Model Mechanobiol 22: 837\u2013850, 2023. https://doi.org/10.1007/s10237-022-01684-0 45. Fei X., L. J. Emily, W. Chenglong, J. Arian, Y. Cheng-Hau, S. S. Michael, K. Adarsh, and H. MingChen, Computational investigation of left ventricular hemo -dynamics following bioprosthetic aortic and mitral valve replacement. Mech Res Commun 112, 2021.", "Comput Methods Biomech Biomed Engin, 2023. DOI: 10.1080/10255842.2023.2199903 44. Morany A., K. Lavon, R. G. Bardon, et al. Fluid \u2013structure interaction modeling of compliant aortic valves using the lattice Boltzmann CFD and FEM methods. Biomech Model Mechanobiol 22: 837\u2013850, 2023. https://doi.org/10.1007/s10237-022-01684-0 45. Fei X., L. J. Emily, W. Chenglong, J. Arian, Y. Cheng-Hau, S. S. Michael, K. Adarsh, and H. MingChen, Computational investigation of left ventricular hemo -dynamics following bioprosthetic aortic and mitral valve replacement. Mech Res Commun 112, 2021. https://doi.org/10.1016/j.mechrescom.2020.103604 46. Yeh H.H., Computational analysis of fluid structure interaction in artificial heart valves. Master\u2019s thesis, The University of British Columbia, Vancouver , 2013. Available online: https://open.library.ubc.ca/ 47. Sanchez-Reyes J., The uniqueness of the rational Bezier polygon is unique. Comput. Aided Geom. Des. 96, 2022. https://doi.org/10.1016/j.cagd.2022.102118 48. Choi C., and C. Kim, C., Numerical analysis on the hemodynamics and leaflet dynamics in a bileaflet mechanical heart valve using a fluid -structure interaction method. ASAIO J 55: 428\u2013437, 2009. 49. COMSOLAB: COMSOL Multiphysics Fluid-Structure Interaction Module User\u2019s Guide. Stockholm, Sweden , 2023. COMSOL AB. Available at: https://www.comsol.com/documentation 50. Manchester E., S. Pirola, M. Salmasi, D. O\u2019Regan, T. Athanasiou, and X. Xu, Analysis of turbulence effects in a patient -specific aorta with aortic valve stenosis. Cardiovasc Eng Technol 12(4): 438\u2013453, 2021. https://doi.org/10.1007/s13239-021-00536-9 51. Abbas S., M. Nasif, R. Al-Waked, M. Said, Numerical investigation on the effect of bileaflet mechanical heart valve\u2019s implantation tilting angle and aortic root geometry on intermittent regurgitation and platelet activation. Artif Organs, 44 : 20\u201339, 2020. https://doi.org/10.1111/aor.13536 52. Nowak M., E. Divo, W. P. Adamczyk, Multiscale model for blood flow after a bileaflet artificial aortic valve implantation. Comput Biol Med 158 , 2023. https://doi.org/10.1016/j.compbiomed.2023.106805 56 53. Yeh H. H., O. Barannyk, D. Grecov, and P. Oshkai, The influence of hematocrit on the hemodynamics of artificial heart valve using fluid -structure interaction analysis. Comput Biol Med 110: 79\u201392, 2019. https://doi.org/10.1016/j.compbiomed.2019.05.003 54. Nitti A., G. D. Cillis, M. D. de Tullio, Numerical investigation of turbulent features past different mechanical aortic valves. J Fluid Mech 256 : 940\u201343, 2022. https://doi.org/10.1017/jfm.2022.256 55. Gallo D., U. Morbiducci, M. D. de Tullio, On the unexplored relationship between kinetic energy and helicity in prosthetic heart valves hemodynamics. Int J Eng Sci 177 , 2022. https://doi.org/10.1016/j.ijengsci.2022.10370 56. Li S., C. He, and Y. Liu, Unsteady flow enhancement on an airfoil using sliding window weak-constraint four-dimensional variational data assimilation. Physics of Fluids , 35 (6): 065122, 2023.", "Tullio, Numerical investigation of turbulent features past different mechanical aortic valves. J Fluid Mech 256 : 940\u201343, 2022. https://doi.org/10.1017/jfm.2022.256 55. Gallo D., U. Morbiducci, M. D. de Tullio, On the unexplored relationship between kinetic energy and helicity in prosthetic heart valves hemodynamics. Int J Eng Sci 177 , 2022. https://doi.org/10.1016/j.ijengsci.2022.10370 56. Li S., C. He, and Y. Liu, Unsteady flow enhancement on an airfoil using sliding window weak-constraint four-dimensional variational data assimilation. Physics of Fluids , 35 (6): 065122, 2023. https://doi.org/10.1063/5.0152348 57. Ahmed M., N. Gupta, R. Jana, M. K. Das, and K. K. Kar, Ramifications of vorticity on aggregation and activation of platelets in bi-leaflet mechanical heart valve: fluidstructure- interaction study. J Biomech Eng. 144, 2022. https://doi.org/10.1115/1.4053665 58. Derazgisoo S.M., P. Akbarzadeh, and A. A. Lehdarboni, Accelerate the convergence of turbulent flows simulation: A novel progressive locally power -law preconditioning method. Computers & Fluids 241 : 105483, 2022. https://doi.org/10.1016/j.compfluid.2022.105483 59. Klusak E., A. Bellofiore, S. Loughnane, S., and N. J. Quinlan, High -resolution measurements of velocity and shear stress in leakage jets from bileaflet mechanical heart valve hinge models. J Biomech Eng 137(11):111008, 2015. https://doi.org/10.1115/1.4031350 60. Hui S., M. Reisman, N. Kumar, Q. Orlandi, D. Lim, and A. Azhari, Aortic valve area \u2014 technical communication: Continuity and Gorlin equations revisited. J Cardiothorac Vasc Anesth 32(6): 2599\u20132606, 2018. https://doi.org/10.1053/j.jvca.2018.03.010 61. Kim H., S. Park, H. Koo, J. Kang, D. Yang, S. Jung, S. Choo, C. Chung, J. Lee, and J. Kim, Determinants of effective orifice area in aortic valve replacement: anatomic and clinical factors. J. Thorac. Dis. 12(5): 1942\u2013195, 2020. https://doi.org/10.21037/jtd-20-188 62. All of Us Research Program: All of Us Researcher Workbench. National Institutes of Health (2024). https://www.researchallofus.org/ 63. Borowiec K., M. Kowalski, M. Kumor, and K. Biernacka, Prolonged left ventricular contraction duration in apical segments as a marker of arrhythmic risk in patients with long QT syndrome. Cardiol. J. 22, 2020. https://doi.org/10.1093/europace/euaa107 57 64. Manzo R., F. Ilardi, D. Nappa, A. Mariani, D. Angellotti, M. Immobile Molaro, G. Sgherzi, D. S. Castiello, F. Simonetti, C. Santoro, M. E. Canonico, M. Avvedimento, R. Piccolo, A. Franzone, and G. Esposito, Echocardiographic evaluation of aortic stenosis: A comprehensive review. Diagnostics (Basel) 13(15) : 2527, 2023. https://doi.org/10.3390/diagnostics13152527 65. Lemaire G., and D. Vancraeynest, Echocardiography of the aortic root: a practical approach for aortic valve-sparing surgery. Ann. Cardiothorac. Surg. 12(3): 194\u2013212, 2023. https://doi.org/10.21037/acs-2022-avs1-14. Epub 2023 May 11 66. Chauhan A., and C. Sasmal, The", "Nappa, A. Mariani, D. Angellotti, M. Immobile Molaro, G. Sgherzi, D. S. Castiello, F. Simonetti, C. Santoro, M. E. Canonico, M. Avvedimento, R. Piccolo, A. Franzone, and G. Esposito, Echocardiographic evaluation of aortic stenosis: A comprehensive review. Diagnostics (Basel) 13(15) : 2527, 2023. https://doi.org/10.3390/diagnostics13152527 65. Lemaire G., and D. Vancraeynest, Echocardiography of the aortic root: a practical approach for aortic valve-sparing surgery. Ann. Cardiothorac. Surg. 12(3): 194\u2013212, 2023. https://doi.org/10.21037/acs-2022-avs1-14. Epub 2023 May 11 66. Chauhan A., and C. Sasmal, The influence of non -Newtonian behaviors of blood on the hemodynamics past a bileaflet mechanical heart valve. Phys. Fluids , 36 (10): 101918 , 2024. https://doi.org/10.1063/5.0231839", "COMPARING LATERAL FLOW ASSAY AND PAPER-BASED MICROFLUIDIC POINT OF CARE DEVICES FOR THE PURPOSE OF INDICATING CHRONIC KIDNEY DISEASE By: Morgan Mah, Shivani Konduru, and Shannon Cao Technical Advisor: Dr. Alessandro Bellofiore Course Instructors: Kevin Maguire and Kathy Le BME 198B - Senior Project May 16, 2025 SIGNATURE PAGE Researcher 1 Researcher 2 Researcher 3 Shivani Konduru Morgan Mah Shannon Cao SID: 015674308 SID: 014740622 SID: 014395095 Researcher Signatures Project Advisor Dr. Alessandro Bellofiore TABLE OF CONTENTS FIGURE LIST ................................................................................................................................ 3 TABLE LIST .................................................................................................................................. 3 KEY WORDS ................................................................................................................................. 4 EXECUTIVE SUMMARY ............................................................................................................ 4 INTRODUCTION ......................................................................................................................... 5 LITERATURE REVIEW .............................................................................................................. 6 BIOMEDICAL MOTIVATION .................................................................................................... 8 FDA Considerations ................................................................................................................ 9 STATEMENT OF NEED .............................................................................................................. 9 MATERIALS AND METHODS .................................................................................................. 9 1. Lateral Flow Assay .............................................................................................................. 9 1.1. Component Preparation ............................................................................................. 9 1.2. Assembly .................................................................................................................... 10 2. Microfluidic Paper-Based Analytical Device ................................................................... 11 2.1. Component Preparation ........................................................................................... 11 2.2. Assembly .................................................................................................................... 11 3. Diagnostic Pad Pre-Treatment .......................................................................................... 12 4. Creatinine Solutions ........................................................................................................... 12 5. Experimental Methods ...................................................................................................... 12 6. Data Analysis ...................................................................................................................... 13 7. Statistical Analysis ............................................................................................................. 13 RESULTS ...................................................................................................................................... 15 1. 2 Way ANOVA .................................................................................................................... 15 2. 2 sample t-Test For LFA Device ........................................................................................ 17 3. 2 sample t-Test For mPAD Device .................................................................................... 19 DISCUSSION ............................................................................................................................... 21 1. 2 Way ANOVA .................................................................................................................... 21 2. 2 Sample t-Test for LFA Device ........................................................................................ 21 3. 2 Sample t-Test for mPAD Device ..................................................................................... 22 4. Summary of Findings ........................................................................................................ 22 CONCLUSION ............................................................................................................................ 22 FUTURE WORK ......................................................................................................................... 23 SAFETY ........................................................................................................................................ 23 ACKNOWLEDGEMENTS ........................................................................................................ 25 REFERENCES ............................................................................................................................. 26 COST ANALYSIS ........................................................................................................................ 27 1 APPENDIX I .................................................................................................................................. 1 BME 198 Technical Memos ..................................................................................................... 1 APPENDIX II ................................................................................................................................. 1 BME 198A NSF Proposal ........................................................................................................ 1 ADDITIONAL APPENDICES ..................................................................................................... 1 LFA Device With Porcine Blood of Unknown Creatinine Concentration .......................... 1 Discussion for Table C ............................................................................................................. 1 2 FIGURE LIST Figure 1 Lateral Flow Assay Device Assembly Figure 2 Dimensions of Filter Paper Design Figure 3 Mean RGB at 9 Minutes Post Sample for LFA Device Figure 4 Mean RGB at 9 Minutes Post Sample for mPAD Appendix Figure A Mean RGB Across 4 Time Points for LFA Device Appendix Figure B Mean RGB Across 4 Time Points for mPAD TABLE LIST Table 1 Dimensions for LFA Components Table 2 2 way ANOVA and Corresponding Tukey Test to compare LFA against mPAD Table 3 2 sample t-Test for LFA device to analyze its colorimetric change over time Table 4 2", "Minutes Post Sample for LFA Device Figure 4 Mean RGB at 9 Minutes Post Sample for mPAD Appendix Figure A Mean RGB Across 4 Time Points for LFA Device Appendix Figure B Mean RGB Across 4 Time Points for mPAD TABLE LIST Table 1 Dimensions for LFA Components Table 2 2 way ANOVA and Corresponding Tukey Test to compare LFA against mPAD Table 3 2 sample t-Test for LFA device to analyze its colorimetric change over time Table 4 2 sample t-Test for mPAD device to analyze its colorimetric change over time Appendix Table A Raw LFA Data with mean RGB and p value for each concentration and time point Appendix Table B Raw mPAD Data with mean RGB and p value for each concentration and time point Appendix Table C 2 sample t-Test for LFA device tested with porcine blood 3 KEY WORDS Point-of-Care at home test - a medical diagnostic test used to indicate the diagnosis of a disease or condition Lateral Flow Assay (LFA) - diagnostic testing method that detects the presence of a specific pathogen or antibody through antibody-antigen interaction via capillary action Microfluidic Paper-Based Analytical Device (mPAD) - diagnostic testing method which utilizes paper and chemical reagents to produce a colorimetric result via capillary action Colorimetric detection - measures the concentration of a substance in a sample through the intensity of the result\u2019s color, after the sample interacts with specific reagents Creatinine processing efficiency - how much creatinine can flow through the point of care device and react (show colorimetric change) over a given period of time EXECUTIVE SUMMARY Creatinine is a waste product produced by the body originating from the digestion of protein and natural breakdown of muscle tissue 1 . Creatinine leaves the body via blood and kidney filtration 1 . When kidneys are damaged, they are unable to effectively remove creatinine from blood leading to a diagnosis of Chronic Kidney Disease (CKD) 1 . There are six stages of Chronic Kidney Disease indicating severity of kidney damage and processing ability, with the final stage being kidney failure 2 . There are about 37 million people in the US who have CKD, and due to varying symptoms amongst individuals, approximately 9 in 10 adults are unaware they have the disease 2 . Populations most likely to be diagnosed with Chronic Kidney Disease include those 65 years old or older, women, and non-hispanic Black", "1 . There are six stages of Chronic Kidney Disease indicating severity of kidney damage and processing ability, with the final stage being kidney failure 2 . There are about 37 million people in the US who have CKD, and due to varying symptoms amongst individuals, approximately 9 in 10 adults are unaware they have the disease 2 . Populations most likely to be diagnosed with Chronic Kidney Disease include those 65 years old or older, women, and non-hispanic Black adults 2 . The current diagnosis procedure involves a blood test to calculate concentration of creatinine, which is used to calculate glomerular filtration rate and determine the stage of CKD 1 . Therefore, there is a need for a point of care device as it would provide a quick evaluation of creatinine concentration without need for the patient to travel to a clinic. 4 The research aims to compare two point-of-care at-home tests, the Microfluidic Paper-Based Analytical (mPAD) and Lateral Flow Assay (LFA). The comparison is made by manufacturing both devices and measuring their ability to detect creatinine. The goal of our research is to prove that mPADs are more efficient at detecting creatinine than LFA. Through manufacturing and testing both devices, we will be able to statistically analyze how well the two devices are able to measure creatinine. Stronger ability to measure creatinine will be evident in stronger colorimetric change. Overall, this research will aid in determining which point-of-care at home test is the more capable of measuring creatinine. INTRODUCTION On average, 1 in 7 adults deal with Chronic Kidney Disease (CKD) in the United States, impacting a total of 37 million people 4 . The key to detecting the severity of CKD, is to measure the level of creatinine within the subject\u2019s body. Since CKD causes the kidneys to gradually lose function, they become unable to filter waste properly leading to an abnormal buildup of creatinine 3 . As a persistent disease that commonly progresses to critical stages without notice, constant monitoring and early detection is critical to effectively manage the subject\u2019s condition. However, traditional diagnostic methods use lab tests to measure the patient\u2019s creatinine concentration. The measurement is made using colorimetric detection which works via a Jaffe\u2019s reaction, a chemical reaction between creatinine and picric acid 5 . The result is then used to calculate estimated glomerular filtration rate (eGFR) values to assess patient kidney performance. A", "As a persistent disease that commonly progresses to critical stages without notice, constant monitoring and early detection is critical to effectively manage the subject\u2019s condition. However, traditional diagnostic methods use lab tests to measure the patient\u2019s creatinine concentration. The measurement is made using colorimetric detection which works via a Jaffe\u2019s reaction, a chemical reaction between creatinine and picric acid 5 . The result is then used to calculate estimated glomerular filtration rate (eGFR) values to assess patient kidney performance. A major issue with the traditional methods is that they must be done clinically and in a medical laboratory environment. As such, they remain inaccessible devices for many subjects of CKD. To solve this gap in knowledge, we want to research which point-of-care at home testing method can provide early indication of CKD. Such a device allows early indication to be more accessible, provides quick results, and allows individuals to be better informed on their kidney\u2019s functional state. Two such possible devices are LFA and mPAD testing. Our main goal is to prove that mPADs are better suited for CKD point-of-care at home testing than LFA technology due to higher detection efficiency. Higher detection efficiency refers to how much creatinine can flow through the point of care device and react after a given amount of time. The primary differences between the devices are the device designs. LFA is an offset strip component design whereas 5 mPAD is a 3-layer component device 7 . Their general processes are similar as they are both point-of-care detection devices that use paper microfluidic technology to measure creatinine concentration of a sample. The device\u2019s reaction zone is pre-treated with alkaline picrate which produces a colorimetry change after creatinine saline solution is applied and flows through paper using capillary action due to the occurrence of a Jaffe\u2019s reaction. Then, image capture is done of the colorimetric changes at various time points. These images are analyzed through imaging software to extract RGB values. Lower RGB values indicate darker color and suggest that more creatine was able to flow through the device. These RGB values will be compared and assessed to determine which technology has a higher detection efficiency depending on which detection method\u2019s RGB values are lower at each testing time interval. LITERATURE REVIEW Point of care diagnostic tests are user-friendly devices that allow patients to determine if they are infected with a particular disease without sending the", "software to extract RGB values. Lower RGB values indicate darker color and suggest that more creatine was able to flow through the device. These RGB values will be compared and assessed to determine which technology has a higher detection efficiency depending on which detection method\u2019s RGB values are lower at each testing time interval. LITERATURE REVIEW Point of care diagnostic tests are user-friendly devices that allow patients to determine if they are infected with a particular disease without sending the results to a lab 1 . Point of care tests are beneficial as they provide a relatively quick result, solve the issue of sample degradation, and don\u2019t require the patient to return to the testing facility 1 . The application of microfluidics in point of care tests has enabled these tests to be effective in limited resource settings 2 . Point of care tests can be manufactured to determine a variety of diagnoses depending on the methods, process, and materials used in the test\u2019s design 1 . More than half of the world\u2019s population live in rural areas, therefore such places lack technology and commodities. Having a low cost diagnosis is important as it is the first obstacle when developing a diagnostic test 5 . Many factors have increased an interest in point of care diagnostics such as infrastructure, less medical professionals, and increased infections 5 . Medical diagnostics in rural areas are often ineffective and not caught up to current world technologies, which results in the tests to be through plain observation in symptoms 5 . When developing a low cost yet effective diagnostic test, several factors must be taken into account. These factors are: affordability, sensitivity, specificity, and must be user-friendly 5 . Lateral flow technology is the traditional method used in point of care testing devices, which have been used to test infectious diseases all over the world 2 . The lateral flow method works when the sample interacts with a conjugate pad, containing pre-loaded antibodies that will interact with the tested antigen 2 . The antibody-antigen pair will then pass through a membrane 6 designed to capture it, enough captured pairs will provide a positive test result 2 . The remaining sample goes through a second membrane to present a control line 2 . Membranes are typically made of hydrophobic nitrocellulose due to its high affinity for proteins, relatively low cost, commercial availability, and capillary flow", "sample interacts with a conjugate pad, containing pre-loaded antibodies that will interact with the tested antigen 2 . The antibody-antigen pair will then pass through a membrane 6 designed to capture it, enough captured pairs will provide a positive test result 2 . The remaining sample goes through a second membrane to present a control line 2 . Membranes are typically made of hydrophobic nitrocellulose due to its high affinity for proteins, relatively low cost, commercial availability, and capillary flow characteristics 2 . Although the lateral flow method in point of care tests has proven successful among transferable diseases, the simplicity of the process limits the ability to test a wider range of diagnoses 2 . A primary goal when improving any form of medical testing or technology is to ensure cost-effectivity and efficiency. When dealing with long-term diseases such as chronic kidney disease (CKD), constant monitoring is needed in order to observe the presence of, or lack of, certain chemicals within a patient. Medical care of such length that requires constant visits to the hospital is neither efficient nor cost-effective, making it not ideal for patients. As such, many choose to avoid such check-ups, which is not safe and can lead to damaging side effects due to a lack of knowledge of the patient's current state 3 . As a more affordable and convenient alternative, recent developments have allowed for the creation of a paper-based microfluidic assay. Paper-based microfluidic assay allows for point-of-care, or POC, testing. For instance, a developing form of testing, transaminase testing, which focuses on observing the presence of transaminase in the liver to maintain the health of patients with drug-induced liver injury, uses paper-based microfluidic technology. The assay allows for fast separation and measurement of the amount of the desired blood sample contents which includes unlike lateral flow assays that require running multiple tests in series to avoid cross-reactivity 3 . As such, paper-based microfluidic assays allow for more rapid testing due to being able to split liquid samples in parallel. Additionally, unlike lateral flow assays, paper-based microfluidic assays are lower in cost due to the methods by which they are created by. Many paper-based microfluidic assays can be fabricated using inkjet printing that can unload multiple reagents simultaneously 4 . The multitasking fabrication method with the use of minimal material allows this testing to remain low-cost and efficient. Inkjet printing is a process of", "assays allow for more rapid testing due to being able to split liquid samples in parallel. Additionally, unlike lateral flow assays, paper-based microfluidic assays are lower in cost due to the methods by which they are created by. Many paper-based microfluidic assays can be fabricated using inkjet printing that can unload multiple reagents simultaneously 4 . The multitasking fabrication method with the use of minimal material allows this testing to remain low-cost and efficient. Inkjet printing is a process of microfluidic assay. A requires a certain filter paper to be cured and after drying, the process is complete. Different types of reagents can be deposited at the same time, making it efficient and simple. Limitations may occur as the inks need to be designed to not block printing nozzles 4 . Origami was also found to be another process of 7 microfluidic assay. Patterns are created on the paper device using photolithography then are folded in certain ways. An aluminum clamp is then used to press the layers together and the process is complete. This process makes the microfluidic assay simple. Various types of microfluidic assays have been found to be fabricated to reduce costs as tests and equipment are pricey. Wax and inkjet printing are the most popular methods of fabrication in terms of cost and efficiency. Wax printing is able to be fabricated on a mass scale. However, it is not resistant to high temperatures. Inkjet printing uses cheap reagents and can also be performed on a mass scale but needs a printer of quality 4 . BIOMEDICAL MOTIVATION Assessing and comparing the RGB values extracted from the Jaffe\u2019s reaction on the LFA compared to the mPAD has several biomedical motivations. Chronic Kidney Disease is a condition that requires constant monitoring in order to ensure the severity of the condition is prevented as well as to detect the disease at an earlier state. As such, it is significant to develop and ensure a point-of-care diagnostic device such as the mPAD is the highest standard. In order to do so, comparing the performance of the LFA to the mPAD in regards to how efficiently creatinine is detected is essential. The LFA is a device that has been on the market for a lot longer than the mPAD and is used for diagnostic purposes for several medical conditions. It is one of the established standards when it comes to diagnostic", "to develop and ensure a point-of-care diagnostic device such as the mPAD is the highest standard. In order to do so, comparing the performance of the LFA to the mPAD in regards to how efficiently creatinine is detected is essential. The LFA is a device that has been on the market for a lot longer than the mPAD and is used for diagnostic purposes for several medical conditions. It is one of the established standards when it comes to diagnostic point-of-care devices. As such, when it comes to diagnostic capabilities the LFA device is the proper standard to compare the mPAD device to conclude whether the mPAD is able to detect creatinine more efficiently. Accurate detection in the timely manner of creatinine levels is important in staging CKD. This allows physicians to adjust medication doses and initiate intervention steps to slow disease progression, reducing complications such as end stage renal failure. Therefore it is important to optimize sensitivity in creatinine detection to improve patient outcomes. Additionally, low cost materials and fabrication of the mPAD would make a feasible solution in low resource settings where laboratory assessments are lackluster. With low costs and simple operation with little personnel needed, it can increase the widespread use for community assessments. 8 FDA Considerations Since the LFA device is already used in diagnostic testing and has been cleared by the FDA as a Class II medical device, it serves as a strong predicate for regulatory comparison. If the mPAD were to be approved for clinical or at-home use, it would likely fall under the same Class II classification and require submission through the FDA\u2019s 510(k) premarket notification process. This would involve demonstrating substantial equivalence to the current LFA device in terms of safety and effectiveness. Because the mPAD features a unique sample distribution and fluid flow structure, additional bench testing, reproducibility studies, and possibly usability testing would likely be required to meet FDA expectations, especially if it is intended for home use. In addition, the chemical agents used in the diagnostic process, including alkaline picrate, would need to be evaluated for safety and proper handling in accordance with FDA and ISO standards, such as ISO 14971 for risk management. Ensuring that the device meets FDA performance criteria would be a critical step in validating the mPAD as a viable and reliable diagnostic alternative. STATEMENT OF NEED mPADs are not as widely available on the", "especially if it is intended for home use. In addition, the chemical agents used in the diagnostic process, including alkaline picrate, would need to be evaluated for safety and proper handling in accordance with FDA and ISO standards, such as ISO 14971 for risk management. Ensuring that the device meets FDA performance criteria would be a critical step in validating the mPAD as a viable and reliable diagnostic alternative. STATEMENT OF NEED mPADs are not as widely available on the market as standardized diagnostic devices such as LFAs. As such, there is a need to draw a comparison in the performance of the mPAD with an already established reliable device. This evaluation could support future adoption of mPADs in clinical or at-home diagnostic workflows. Evaluating whether the mPAD performs as well, or better, than the LFA device, enhances its clinical dependability towards the device. MATERIALS AND METHODS 1. Lateral Flow Assay 1.1. Component Preparation There are four components in the lateral flow assay device design, which consist of pressure sensitive adhesive, cellulose fiber paper, glass fiber paper, and opaque plastic sheet. All four components were cut by SJSU\u2019s machine shop with a \u00b1 0.5 mm tolerance. 9 Component Dimensions Pressure sensitive adhesive 0.5 cm x 6.4 cm Cellulose fiber paper 0.5 cm x 0.5 cm Glass fiber paper 0.3 cm x 1.0 cm Opaque plastic sheet 0.5 cm x 3.4 cm Table 1 - dimensions for LFA components 1.2. Assembly To assemble the LFA device, the pressure sensitive adhesive is first placed onto the lab bench with the adhesive on top. The protective layer on the adhesive is removed, and the opaque plastic sheet is placed at the edge of one end of the adhesive. The glass fiber paper is placed ~1 cm from the other end of the adhesive. Finally, the cellulose fiber paper is placed so that 0.25 cm is above the glass fiber paper, and 0.25 cm is overlapping the glass fiber paper. The assembly of the LFA device allows creatinine solutions to be deposited onto the sample pad. By capillary action, the solution will travel onto the diagnostic pad, which was treated with alkaline picrate solution, to display colorimetric change and measure creatinine. Figure 1 - Lateral Flow Assay Device Assembly 10 2. Microfluidic Paper-Based Analytical Device 2.1. Component Preparation There are three components in the microfluidic paper-based analytical device, which consists of filter paper, membrane", "is overlapping the glass fiber paper. The assembly of the LFA device allows creatinine solutions to be deposited onto the sample pad. By capillary action, the solution will travel onto the diagnostic pad, which was treated with alkaline picrate solution, to display colorimetric change and measure creatinine. Figure 1 - Lateral Flow Assay Device Assembly 10 2. Microfluidic Paper-Based Analytical Device 2.1. Component Preparation There are three components in the microfluidic paper-based analytical device, which consists of filter paper, membrane paper, and a plastic sheet. All three components were cut to a dimension of 26mm x 14mm. The filter paper and membrane paper\u2019s designs were printed using a thermal printer and baked to make the borders hydrophobic. The hydrophobic borders on both the filter and membrane paper allow the sample to flow through the device via capillary action. The baking procedure for the filter paper was 30 minutes at 100 C, and 10 minutes at 100 C for the membrane paper. The filter paper\u2019s design is shown in figure 2, while the membrane paper\u2019s design is identical to the filter paper design, except without the line in the middle which separates the sample region from the diagnostic region. The diagnostic region on the filter paper is pre-treated with 10\u03bc L of alkaline picrate. The mPAD device, baking, and pre-treatment procedures were adapted from previous CardioLab teams. Figure 2 - Dimensions of filter paper design 2.2. Assembly After the components were prepared, they were stacked on top of one another, with the plastic sheet being the bottom layer, the membrane paper being the middle layer, and the filter paper being the top layer. The components are taped together with double-sided tape. The design\u2019s purpose is to direct flow of 11 the samples through the hydrophobic channels and into the diagnostic region, which will react with alkaline picrate solution for creatinine detection. 3. Diagnostic Pad Pre-Treatment In both devices, the diagnostic pads, glass fiber paper in LFA device and diagnostic region in mPAD are pretreated with an alkaline picrate solution. The solution consists of a 1:1 ratio of 0.04M picric acid and 2M NaOH. The 0.04M picric acid was created by diluting 1.3% stock picric acid with distilled water. The dilution requires 70.5mL of 1.3% stock picric acid and 29.5mL distilled water to be combined into a 100mL beaker. The 2M NaOH solution was created by measuring 7.9 grams of sodium hydroxide", "glass fiber paper in LFA device and diagnostic region in mPAD are pretreated with an alkaline picrate solution. The solution consists of a 1:1 ratio of 0.04M picric acid and 2M NaOH. The 0.04M picric acid was created by diluting 1.3% stock picric acid with distilled water. The dilution requires 70.5mL of 1.3% stock picric acid and 29.5mL distilled water to be combined into a 100mL beaker. The 2M NaOH solution was created by measuring 7.9 grams of sodium hydroxide pellets into an Erlenmeyer flask, 50mL of distilled water was added and the solution was stirred until the pellets dissolved. Once the pellets were dissolved, 50mL were added to get a final volume of 100mL. Finally, the alkaline picrate solution was created by combining the volumes of both solutions, 0.04M picric acid and 2M NaOH into a 500mL beaker. 4. Creatinine Solutions The creatinine solutions were created using Phosphate-buffered saline (PBS) with varying amounts of creatinine sourced by a creatinine kit from the Sigma Aldrich Company. We tested the devices with 8 different concentrations; two were below normal creatinine levels (0.4685 mg/dL, 0.9357 mg/dL), three within normal creatinine levels (1.875 mg/dL, 3.75 mg/dL, 7.5 mg/dL), and three above normal creatinine levels (15 mg/dL, 30 mg/dL, 60 mg/dL). These concentrations were chosen based on current literature for normal creatinine levels in adults; the normal range of creatinine in blood for adult men is between 0.74 and 1.35 mg/dL, and for adult women is between 0.59 to 1.04 mg/dL 7 . 5. Experimental Methods Since we needed a direct comparison between the two devices, mPAD and LFA, we manufactured 54 LFA devices and 54 mPAD devices; we also tested the devices using the same creatinine solutions. The assembled devices were secured onto standard 12 A4-sized paper using double-sided tape. The A4-sized paper with the devices attached to them was then placed into a lightbox in order to keep consistency when analyzing RGB values. The tests were grouped based on the type of device and the concentration that was used to test the device. In total we had 18 groups, with each group representing one of the 9 creatinine concentrations for one of the devices. We also had 6 replicates for each group to get a more average RGB value for our statistical tests. An image of both devices, each in a group of 6 of their respective devices, was captured under a", "values. The tests were grouped based on the type of device and the concentration that was used to test the device. In total we had 18 groups, with each group representing one of the 9 creatinine concentrations for one of the devices. We also had 6 replicates for each group to get a more average RGB value for our statistical tests. An image of both devices, each in a group of 6 of their respective devices, was captured under a lightbox at four time intervals: 0 minutes, 3 minutes, 6 minutes, and 9 minutes. An iPhone 14 Pro was used for LFA image capture and an iPhone 13 was used for mPAD image capture. For the LFA device, using a P20 pipette, 3\u03bc L of the tested creatinine concentration was pipetted onto the device\u2019s sample pad and left under the lightbox with the timer started immediately. For the mPAD, 30\u03bc L of the creatinine concentration was pipetted onto its sample region using a P200 pipette. The images that accounted for 0 minutes were taken before the creatinine solutions were pipetted. 4 images of the group of 6 devices for each device type were captured per creatinine concentration group. 6. Data Analysis The data was analyzed by measuring the RGB value for each data point using python. The images captured for each tested LFA and mPAD device were cropped to isolate the diagnostic pad of the analyzed device. The diagnostic pad of the device is the area where the Jaff reaction, in the case of LFA, or capillary action in the case of the mPAD, occurs which results in colorimetric change. As such, the diagnostic region for each test is isolated by cropping it into separate images. The images are organized into folders separating the devices by creatinine concentration, then further separating it by the time interval the image was captured at. The python code looped through each folder to extract the RGB values for each device at each time. 7. Statistical Analysis The goal of our statistical analysis was to compare RGB values of the LFA and mPAD devices against each other. To achieve this goal we first performed multiple 2-way 13 ANOVA tests, to demonstrate that the two devices don\u2019t process creatinine the same way for a given concentration and RGB channel. Since we had 9 different concentrations and 3 RGB channels, we ended up with 27 individual 2-way", "the RGB values for each device at each time. 7. Statistical Analysis The goal of our statistical analysis was to compare RGB values of the LFA and mPAD devices against each other. To achieve this goal we first performed multiple 2-way 13 ANOVA tests, to demonstrate that the two devices don\u2019t process creatinine the same way for a given concentration and RGB channel. Since we had 9 different concentrations and 3 RGB channels, we ended up with 27 individual 2-way ANOVA tests. In all 27 tests, the null hypothesis was that there is not a statistically significant difference in RGB between the mPAD and LFA device; while the alternative hypothesis was that there is a statistically significant difference in RGB between the mPAD and LFA device. Since we wanted to focus on colorimetric change, to prove that the devices process creatinine differently, we set the response as the analyzed color channel and the factors as device and time. In this set up, all 6 replicates of the LFA device for the RGB channel at a given concentration and at all 4 time points, was compared in one 2-way ANOVA test against all 6 replicates of the mPAD device for the RGB channel at the same concentration and at all 4 time points. We reported the F and p values of these tests, as the p value would allow us to either accept or reject the null hypothesis, and the F value would confirm that device type has a strong effect on RGB; with a larger F value meaning that the relationship between device type and RGB is strong, and not due to probability. Finally, to demonstrate further proof of statistical differences in RGB, we performed Tukey tests on each 2-way ANOVA test, as a Tukey test takes into consideration false positives. After we prove that the two devices process creatinine differently, we need to determine which device allows for a better flow of creatinine. To make that conclusion, we will perform multiple 2 sample t-Tests. For both devices, the 0 mg/dL is treated as a control, where the tested concentration is compared against it for a given RGB channel. The comparison is made against the 0 mg/dL concentration, in order to isolate creatinine in the analysis, and to take into consideration any colorimetric change which may occur from PBS. Considering that we are testing with 8 concentrations and 3 RGB", "better flow of creatinine. To make that conclusion, we will perform multiple 2 sample t-Tests. For both devices, the 0 mg/dL is treated as a control, where the tested concentration is compared against it for a given RGB channel. The comparison is made against the 0 mg/dL concentration, in order to isolate creatinine in the analysis, and to take into consideration any colorimetric change which may occur from PBS. Considering that we are testing with 8 concentrations and 3 RGB channels, we will have 24 individual 2 sample t-Tests for the LFA device and 24 individual 2 sample t-Tests for the mPAD device. We focused our data on 9 minutes after the sample was dropped, in order to get a better summary of the data, as the sample would still be developing at 3 and 6 minutes. The null hypothesis for these tests would be that there is not a statistically significant difference between the control and the tested concentration. While the alternative hypothesis would be that there is a statistically significant difference between 14 the control and the tested concentration. After performing all of the 2 sample t-Tests, we expect to see more p values in the mPAD than the LFA which are \u2264 0.05, or statistically significant. This would allow us to reject more null hypotheses in the mPAD, in order to conclude that the mPAD is more efficient in its ability to measure creatinine than the LFA 9 minutes after the sample was dropped. The last point in our statistical analysis, would be to graph the mean RGB point values, in order to see if there are any differences in how color is showing up in the devices. This observation is important to our experiment, as we are using two different phones to image our devices, therefore we would need to account for any differences in color tendencies between the phones. This would be an important observation when evaluating our 2 way ANOVA data. RESULTS 1. 2 Way ANOVA Concentration Color Channel F value p-value Adjusted p-value from Tukey 0 mg/dL red 219.92 0.000 0.000 green 46.98 0.000 0.000 blue 213.01 0.000 0.000 0.4685 mg/dL red 200.86 0.000 0.000 green 0.76 0.387 0.387 blue 151.69 0.000 0.000 0.9375 mg/dL red 43.31 0.000 0.000 green 42.90 0.000 0.000 blue 239.50 0.000 0.000 1.875 mg/dL red 1.82 0.185 0.185 green 53.94 0.000 0.000 blue 374.56 0.000 0.000 15 3.75", "when evaluating our 2 way ANOVA data. RESULTS 1. 2 Way ANOVA Concentration Color Channel F value p-value Adjusted p-value from Tukey 0 mg/dL red 219.92 0.000 0.000 green 46.98 0.000 0.000 blue 213.01 0.000 0.000 0.4685 mg/dL red 200.86 0.000 0.000 green 0.76 0.387 0.387 blue 151.69 0.000 0.000 0.9375 mg/dL red 43.31 0.000 0.000 green 42.90 0.000 0.000 blue 239.50 0.000 0.000 1.875 mg/dL red 1.82 0.185 0.185 green 53.94 0.000 0.000 blue 374.56 0.000 0.000 15 3.75 mg/dL red 75.50 0.000 0.000 green 6.19 0.017 0.017 blue 56.93 0.000 0.000 7.5 mg/dL red 28.33 0.000 0.000 green 403.07 0.000 0.000 blue 282.78 0.000 0.000 15 mg/dL red 40.27 0.000 0.000 green 142.78 0.000 0.000 blue 394.88 0.000 0.000 30 mg/dL red 2.05 0.160 0.160 green 30.17 0.000 0.000 blue 136.95 0.000 0.000 60 mg/dL red 8.48 0.006 0.006 green 89.67 0.000 0.000 blue 269.95 0.000 0.000 Table 2 - 2 way ANOVA and Corresponding Tukey Test to compare LFA against mPAD Table 2 highlights the three 2-way ANOVA tests which failed to reject the null hypothesis. Those tests prove that there was no statistically significant difference in RGB between LFA and mPAD for the specified concentration and color channel over time. This conclusion is due to those three tests having p values larger than 0.05, for both the ANOVA and Tukey test. It can be observed that most of the tests were determined to be statistically significant, which supports the conclusion that the LFA and mPAD devices process creatinine differently. Additionally, the statistically significant tests had a large F value, between ~40 and 300, which demonstrates a strong relationship between device type and RGB for those tests. 16 2. 2 sample t-Test For LFA Device Concentration Channel color P-value 0.4685 mg/dL @ 9 minutes red 0.930 green 0.565 blue 0.614 0.9375 mg/dL @ 9 minutes red 0.803 green 0.492 blue 0.282 1.875 mg/dL @ 9 minutes red 0.334 green 0.798 blue 0.863 3.75 mg/dL @ 9 minutes red 0.161 green 0.999 blue 0.077 7.5 mg/dL @ 9 minutes red 0.300 green 0.528 blue 0.672 15 mg/dL @ 9 minutes red 0.016 green 0.115 blue 0.011 30 mg/dL @ 9 minutes red 0.027 green 0.051 blue 0.014 60 mg/dL @ 9 minutes red 0.464 green 0.228 blue 0.067 17 Table 3 - 2 sample t-Test for LFA device to analyze its colorimetric change over time Table 3 displays", "0.334 green 0.798 blue 0.863 3.75 mg/dL @ 9 minutes red 0.161 green 0.999 blue 0.077 7.5 mg/dL @ 9 minutes red 0.300 green 0.528 blue 0.672 15 mg/dL @ 9 minutes red 0.016 green 0.115 blue 0.011 30 mg/dL @ 9 minutes red 0.027 green 0.051 blue 0.014 60 mg/dL @ 9 minutes red 0.464 green 0.228 blue 0.067 17 Table 3 - 2 sample t-Test for LFA device to analyze its colorimetric change over time Table 3 displays the colorimetric change, using multiple 2 sample t-Tests to evaluate the change at different creatinine concentrations. Similarly to table 1, the highlighted tests were those that failed to reject the null hypothesis. Each test compares a creatinine concentration of 0 mg/dL to a given creatinine concentration, to evaluate the effect of increased creatinine over time. According to table 3, there were 20 tests in which we failed to reject the null hypothesis, meaning that the majority of the tests were not statistically significant in the LFA device 9 minutes after the sample was dropped. For the LFA device, we were able to observe statistical significance as concentration increased in the device, as 15 mg/dL and 30 mg/dL had all 4 of the statistically significant tests. Figure 3 - Mean RGB at 9 Minutes Post Sample for LFA Device Figure 3 represents the trend in color change at 9 minutes after the sample was dropped onto the LFA device, analyzed with 2 sample t-Tests. It can be observed that the blue and red channels had the largest point change. Specifically, the blue channel had the highest point change at 3.75 mg/dL, 15 mg/dL, 30 mg/dL, and 60 mg/dL. 18 3. 2 sample t-Test For mPAD Device Concentration Channel color P-value 0.4685 mg/dL @ 9 minutes red 0.001 green 0.006 blue 0.002 0.9375 mg/dL @ 9 minutes red 0.003 green 0.041 blue 0.193 1.875 mg/dL @ 9 minutes red 0.341 green 0.167 blue 0.826 3.75 mg/dL @ 9 minutes red 0.235 green 0.856 blue 0.180 7.5 mg/dL @ 9 minutes red 0.000 green 0.000 blue 0.294 15 mg/dL @ 9 minutes red 0.002 green 0.005 blue 0.188 30 mg/dL @ 9 minutes red 0.855 green 0.879 blue 0.017 60 mg/dL @ 9 minutes red 0.000 green 0.001 blue 0.527 19 Table 4 - 2 sample t-Test for mPAD device to analyze its colorimetric change over time Table 4 displays the colorimetric change,", "blue 0.826 3.75 mg/dL @ 9 minutes red 0.235 green 0.856 blue 0.180 7.5 mg/dL @ 9 minutes red 0.000 green 0.000 blue 0.294 15 mg/dL @ 9 minutes red 0.002 green 0.005 blue 0.188 30 mg/dL @ 9 minutes red 0.855 green 0.879 blue 0.017 60 mg/dL @ 9 minutes red 0.000 green 0.001 blue 0.527 19 Table 4 - 2 sample t-Test for mPAD device to analyze its colorimetric change over time Table 4 displays the colorimetric change, using multiple 2 sample t-Tests to evaluate the change at different creatinine concentrations. The results were produced with the same procedure as table 3. According to table 4, there were 12 tests in which we failed to reject the null hypothesis, meaning that half of the tests were not statistically significant 9 minutes after the sample was dropped. Figure 4 - Mean RGB at 9 Minutes Post Sample for mPAD Figure 4 presents the point changes in RGB over time for the mPAD device, using 2 sample t-Tests, evaluated at 9 minutes. It can be observed that the red and green channels had the most point difference. There were some inconsistencies with the blue channel at the 7.5 mg/dL and the 15 mg/dL concentrations, as the blue channel points were visually much lower than the red and green channels at those concentrations. 20 DISCUSSION 1. 2 Way ANOVA The 2-way ANOVA tests (table 2) demonstrates statistical significance between the LFA and mPAD device. These results fell in line with what we expected, the two device designs have differences in how samples flow through them. The mPAD has a longer flow path as it has to go from the sample region, down to the membrane paper, and upwards to the diagnostic region where it reacts with alkaline picrate. The LFA has a shorter flow path, as it seeps through the sample pad, and directly onto the diagnostic pad, again reacting with alkaline picrate. The complexity of the flow path in the two devices, allowed us to make a logical hypothesis, that the two devices would produce different colorimetric change. A significant majority of the 2 way ANOVA tests, 89% of the tests had a p value less than 0.05, allowing us to reject many of the null hypotheses, and conclude that the devices process creatinine differently. Since the LFA data (figure 3) had more color change in the blue and red", "alkaline picrate. The complexity of the flow path in the two devices, allowed us to make a logical hypothesis, that the two devices would produce different colorimetric change. A significant majority of the 2 way ANOVA tests, 89% of the tests had a p value less than 0.05, allowing us to reject many of the null hypotheses, and conclude that the devices process creatinine differently. Since the LFA data (figure 3) had more color change in the blue and red channels, the phone used could have had a preference or tendency towards blue and red. Similarly, the phone used to image the mPAD data (figure 4) could have had a preference towards red and green. These differences could have affected our 2 way ANOVA tests, although we suspect the error to be very minimal since many of those tests were statistically significant. Therefore proving that the two devices in fact process creatinine differently. 2. 2 Sample t-Test for LFA Device The 2 sample t-tests for the LFA device (table 3), analyzed how the LFA device was able to process creatinine via colorimetric change on the diagnostic pad. Based on the table, we observed only 4 statistically significant tests. This informed us that the LFA device had difficulty allowing creatinine to flow through the LFA device. The LFA device was therefore unable to produce statistically significant colorimetric change, using a PBS with creatinine solution. Additionally, we were able to observe the 4 statistically significant tests that appeared at 15 mg/dL and 30 mg/dL, demonstrating that colorimetric change increases with a higher creatinine concentration at 9 minutes in the LFA device. 21 3. 2 Sample t-Test for mPAD Device The 2 sample t-Tests for the mPAD device (table 4), analyzed how the mPAD device was able to process creatinine. Based on the table, we observed 12 statistically significant tests. These results informed us that the mPAD was sufficient at allowing creatinine to flow through the device, using a PBS with creatinine solution. Unlike the LFA device, we were unable to observe a trend of statistical significance with increased creatinine concentration. This could have been sources of error such as lighting inconsistencies, or accuracy of the imaging. 4. Summary of Findings The results from our research demonstrate that our hypothesis, the mPAD device having a stronger ability to process creatinine than the LFA device, to be true. Our results proved that the mPAD", "flow through the device, using a PBS with creatinine solution. Unlike the LFA device, we were unable to observe a trend of statistical significance with increased creatinine concentration. This could have been sources of error such as lighting inconsistencies, or accuracy of the imaging. 4. Summary of Findings The results from our research demonstrate that our hypothesis, the mPAD device having a stronger ability to process creatinine than the LFA device, to be true. Our results proved that the mPAD device is better at processing creatinine when tested with phosphate-buffered saline (PBS) and creatinine. This conclusion was based on the mPAD device having 12 statistically significant 2 sample t-Tests, while the LFA device only had 4 statistically significant tests. Although device design differences are most likely the reason for our result, a potential source of error could have been due to blurriness of the images, how perfectly the images were cropped, or how even the alkaline picrate was pipetted onto the diagnostic regions. These three factors could have contributed to how the python code was able to measure average RGB, which would have affected the accuracy of the 2 sample t-Tests. CONCLUSION Our research analyzed the mPAD and LFA device, in order to determine which device was able to process creatinine the best. Prior to the research, we predicted that the mPAD would be better at processing creatinine, due to the device\u2019s ability to filter creatinine in blood. Based on our results and statistical analysis, we concluded that our hypothesis is true and that the mPAD device is better at processing creatinine via stronger colorimetric change on the diagnostic pad. The 2 way ANOVA tests, compared RGB across time points 0, 3, 6, and 9 minutes to establish the basis of our hypothesis. The 2 sample t-Tests then compared the RGB of PBS, to 22 the RGB of PBS with known amounts of creatinine over time, to determine which device could produce stronger colorimetric change 9 minutes after the sample was dropped. Since the mPAD device had more statistically significant tests, that meant that the mPAD device had a better flow of creatinine through the device. We validated our 2 way ANOVA test by graphing the mean RGB point changes for the two devices at 9 minutes using the 2 sample t-Test data, since the imaging of the devices used different phones. We were able to prove that although the", "could produce stronger colorimetric change 9 minutes after the sample was dropped. Since the mPAD device had more statistically significant tests, that meant that the mPAD device had a better flow of creatinine through the device. We validated our 2 way ANOVA test by graphing the mean RGB point changes for the two devices at 9 minutes using the 2 sample t-Test data, since the imaging of the devices used different phones. We were able to prove that although the phones had different color tendencies, the results still proved that the devices process creatinine differently, and that the mPAD device performed better than the LFA, via stronger colorimetric change. FUTURE WORK For future work, our lab could choose to pursue the mPAD device design in order to make a point of care device which can measure creatinine in blood. Our data suggests that the mAPD is better at processing creatinine, through stronger colorimetric change than the LFA device at 9 minutes after the sample is placed. The mPAD also has more research supporting its ability to detect biomarkers in blood, which is another good reason to use such device design. Another choice could be to try to make the LFA device design work, as perhaps modifications to the amount of overlap between the sample and diagnostic pad, could result in stronger color change. SAFETY The experiment involves the use of blood and two hazardous reagents, Picric acid and NaOH. The hazardous reagents require careful handling when being used in the experiment in order to prevent the risk of skin or eye irritation, burns, or various other potential health hazards. To handle the reagents with care while performing the experiments it is important to wear the proper personal protective equipment (PPE) at all times. The proper PPE attire involves wearing gloves, a lab coat, and goggles at all times while using these two solutions. Ensure that all contaminated materials and areas of the lab are thoroughly washed. It is also proper practice to ensure all solution containers are labelled to avoid misuse of the chemical. Picric acid is an explosive hazardous chemical that poses several hazards. Picric acid\u2019s explosive properties make it highly sensitive to heat shock and friction when it is dry. As such, it 23 is important to always ensure it is sealed when not in use. If in use, only leave the acid open under the hood during", "of the lab are thoroughly washed. It is also proper practice to ensure all solution containers are labelled to avoid misuse of the chemical. Picric acid is an explosive hazardous chemical that poses several hazards. Picric acid\u2019s explosive properties make it highly sensitive to heat shock and friction when it is dry. As such, it 23 is important to always ensure it is sealed when not in use. If in use, only leave the acid open under the hood during the experiment to ensure proper ventilation to avoid a build up of flammable vapors. Ingesting or inhaling the acid can be toxic and cause adverse health conditions. Skin or eye contact with the acid without proper PPE can result in skin or eye irritation and burns if in contact with skin. To minimize risks of burns and irritation, it is important to wear gloves and goggles. If contact with skin or eyes occurs, rinse thoroughly with water immediately. The Picric acid used in this experiment was of an extremely diluted solution of 0.04M. However, the acid still poses the same hazardous risks and must be used with the same precautions and standards. NaOH is a corrosive hazardous reagent that can result in various hazards if used incorrectly. It is a corrosive chemical that severely damages skin and eyes if it comes into contact or is inhaled or ingested. As such, it also requires the use of PPE at all times and must be used under the hood when performing experiments. Once done with the experiment, ensure any materials and areas of the lab contaminated with the chemical are washed thoroughly and immediately. Blood was also used in this experiment. Since blood is a biological material, it must be treated as a biohazard. It is standard to wear proper PPE while handling blood at all times. All lab surfaces and materials that were in contact with blood must be disinfected or properly disposed of along with any waste contaminated by it in the biohazard bin immediately after use or once the experiment is done. Blood used in the experiments followed standard biosafety guidelines. Waste disposal of Picric acid and NaOH must be done in an official labelled waste bucket requested. Ensure reagents are not poured down the drain and in a proper waste bucket issued to these specific chemicals and solutions. Do not use unlabelled containers to contain waste. The issued,", "properly disposed of along with any waste contaminated by it in the biohazard bin immediately after use or once the experiment is done. Blood used in the experiments followed standard biosafety guidelines. Waste disposal of Picric acid and NaOH must be done in an official labelled waste bucket requested. Ensure reagents are not poured down the drain and in a proper waste bucket issued to these specific chemicals and solutions. Do not use unlabelled containers to contain waste. The issued, labelled waste bucket must be used for its corresponding chemicals. All hazardous chemicals the experiments followed standard hazardous safety guidelines. 24 ACKNOWLEDGEMENTS Morgan Materials and Methods, Results, Discussion, Conclusion, Future Work Shivani Biomedical Motivation, Statement of Need, Materials and Methods, Safety Shannon Cost Analysis, Biomedical Motivation, References Collective Contribution Executive Summary, Introduction, Literature Review, References, Appendix 25 REFERENCES [1] \u201cCreatinine\u201d. National Kidney Foundation. 1 Jun. 2023. https://www.kidney.org/kidney-topics/creatinine#:~:text=More%20resources-,About%20Creatini ne,of%20a%20possible%20kidney%20problem. Accessed 6 Dec. 2024 [2] \u201cWhat is Chronic Kidney Disease?\u201d. Texas Health and Human Services. https://www.hhs.texas.gov/services/health/chronic-kidney-disease/what-chronic-kidney-disease#: ~:text=Chronic%20kidney%20disease%20(CKD)%2C,work%20as%20well%20as%20normal. Accessed 6 Dec. 2024 [3] \u201cChronic Kidney Disease (CKD)\u201d. National Kidney Foundation. 11 Sept. 2023. https://www.kidney.org/kidney-topics/chronic-kidney-disease-ckd. Accessed 6 Dec. 2024 [4] Centers for Disease Control and Prevention. Chronic Kidney Disease in the United States, 2021. Centers for Disease Control and Prevention, US Department of Health and Human Services; 2021. [5] Elsayed, R., \u201cPatient-Friendly Kidney Function Screening\u201d. The Faculty of the Department of Biomedical Engineering, San Jose State University, May 2018. pdf. [6] \u201cCreatinine Test\u201d. Mayo Clinic. 9 Feb. 2023. https://www.mayoclinic.org/tests-procedures/creatinine-test/about/pac-20384646#:~:text=Serum %20creatinine%20is%20reported%20as,urine%20samples%20may%20be%20used. Accessed 1 Dec. 2024 [7] Kovac, I., \u201cCharacterization of the Sensitivity and Limitations of Serum Creatinine Detection of Paper-Based Microfluidic Devices\u201d. The Faculty of the Department of Biomedical Engineering, San Jose State University. pdf. 26 COST ANALYSIS The cost analysis of the lateral flow assay focuses on low-cost production while maintaining functionality. The device consists of four primary components: pressure-sensitive adhesive (0.5 \u00d7 6.4 cm), cellulose fiber paper (0.5 \u00d7 0.5 cm), glass fiber paper (0.3 \u00d7 1.0 cm), and an opaque plastic sheet (0.5 \u00d7 3.4 cm). All of the components were cut to specification by San Jos\u00e9 State University\u2019s machine shop, with a \u00b10.5 mm dimensional tolerance. For bulk production estimates, the pressure-sensitive adhesive, which is commonly sourced from vendors like 3M, costs approximately $50 per 100 cm\u00b2 sheet, yielding around 100 strips per sheet. This resulted in a per-unit cost of about $0.05. Cellulose fiber paper is priced at around $60", "paper (0.3 \u00d7 1.0 cm), and an opaque plastic sheet (0.5 \u00d7 3.4 cm). All of the components were cut to specification by San Jos\u00e9 State University\u2019s machine shop, with a \u00b10.5 mm dimensional tolerance. For bulk production estimates, the pressure-sensitive adhesive, which is commonly sourced from vendors like 3M, costs approximately $50 per 100 cm\u00b2 sheet, yielding around 100 strips per sheet. This resulted in a per-unit cost of about $0.05. Cellulose fiber paper is priced at around $60 for 100 sheets measuring 20 \u00d7 20 cm, which provides approximately 1,600 usable squares, yielding about $0.04 per device. Glass fiber paper, used as a conjugate or sample pad, is slightly more expensive at roughly $80 per 100 sheets, offering around 6,600 segments per sheet. This translates to a cost of just $0.012 per unit. The opaque plastic backing sheet adds about $0.01 per device at a rate of $25 per 100 sheets. In terms of fabrication, the cutting process at the machine shop was free but can be estimated at about $0.05 per device, while manual assembly, including alignment and layering, would be an additional $0.08. Quality control and simple packaging contribute an additional $0.02. Altogether, the material cost per unit would be approximately $0.112, and the total fabrication and labor cost amounts to $0.15, resulting in a final estimated cost of $0.26 per device. The pressure-sensitive adhesive is the largest material cost, which suggests that cost reductions could be achieved by limiting the use of adhesive or selecting a more affordable alternative. More savings could be made possible through mass production techniques such as roll-to-roll fabrication as it increases efficiency and production speeds. Overall, the analysis demonstrates that a cost-effective, low complexity lateral flow device can be produced for under $0.50 per test, making it suitable for widened diagnostic use in resource-limited and cost-effective settings. 27 APPENDIX I BME 198 Technical Memos The following link contains a Google Drive folder with all of the technical memos submitted to BME 198A and BME 198B. SMS Technical Memos APPENDIX II BME 198A NSF Proposal The following pages contain the original NSF project proposal submitted for BME 198A included here for reference. NSF Proposal.docx.pdf ADDITIONAL APPENDICES LFA Mean RGB Concentration Channel color Mean RGB point value P-value 0.4685 mg/dL @ 0 minutes red 28.42 0.000 green 27.53 0.000 blue 19.42 0.000 0.4685 mg/dL @ 3 minutes red 5.13 0.004 green 5.60", "Drive folder with all of the technical memos submitted to BME 198A and BME 198B. SMS Technical Memos APPENDIX II BME 198A NSF Proposal The following pages contain the original NSF project proposal submitted for BME 198A included here for reference. NSF Proposal.docx.pdf ADDITIONAL APPENDICES LFA Mean RGB Concentration Channel color Mean RGB point value P-value 0.4685 mg/dL @ 0 minutes red 28.42 0.000 green 27.53 0.000 blue 19.42 0.000 0.4685 mg/dL @ 3 minutes red 5.13 0.004 green 5.60 0.008 blue 5.23 0.076 0.4685 mg/dL @ 6 minutes red 1.55 0.281 green 3.43 0.058 blue 2.05 0.506 0.4685 mg/dL @ 9 minutes red 0.18 0.930 green 1.40 0.565 blue 1.33 0.614 28 0.9375 mg/dL @ 0 minutes red 19.98 0.000 green 17.49 0.000 blue 13.40 0.000 0.9375 mg/dL @ 3 minutes red 13.69 0.001 green 16.72 0.000 blue 14.67 0.000 0.9375 mg/dL @ 6 minutes red 7.41 0.018 green 9.71 0.005 blue 6.83 0.006 0.9375 mg/dL @ 9 minutes red 0.77 0.803 green 2.42 0.492 blue 2.45 0.282 1.875 mg/dL @ 0 minutes red 26.20 0.000 green 24.51 0.000 blue 17.39 0.000 1.875 mg/dL @ 3 minutes red 7.43 0.004 green 8.83 0.003 blue 5.89 0.029 1.875 mg/dL @ 6 minutes red 3.67 0.028 green 5.79 0.007 blue 4.07 0.039 1.875 mg/dL @ 9 minutes red 2.70 0.334 green 0.77 0.798 blue 0.35 0.863 3.75 mg/dL @ 0 minutes red 28.36 0.000 29 green 25.27 0.000 blue 21.98 0.001 3.75 mg/dL @ 3 minutes red 6.64 0.003 green 10.24 0.000 blue 5.64 0.162 3.75 mg/dL @ 6 minutes red 5.19 0.009 green 8.73 0.001 blue 2.05 0.521 3.75 mg/dL @ 9 minutes red 3.37 0.161 green 0.00 0.999 blue 4.95 0.077 7.5 mg/dL @ 0 minutes red 29.04 0.000 green 29.86 0.000 blue 23.72 0.000 7.5 mg/dL @ 3 minutes red 12.57 0.002 green 11.47 0.001 blue 7.13 0.071 7.5 mg/dL @ 6 minutes red 3.70 0.306 green 3.13 0.275 blue 4.08 0.215 7.5 mg/dL @ 9 minutes red 3.33 0.300 green 1.94 0.528 blue 1.38 0.672 15 mg/dL @ 0 minutes red 30.22 0.000 green 27.20 0.000 30 blue 23.62 0.000 15 mg/dL @ 3 minutes red 5.39 0.025 green 8.29 0.003 blue 0.93 0.758 15 mg/dL @ 6 minutes red 1.77 0.320 green 4.31 0.072 blue 1.86 0.491 15 mg/dL @ 9 minutes red 7.18 0.016 green 5.27 0.115 blue 8.73 0.011 30 mg/dL @ 0 minutes", "0.275 blue 4.08 0.215 7.5 mg/dL @ 9 minutes red 3.33 0.300 green 1.94 0.528 blue 1.38 0.672 15 mg/dL @ 0 minutes red 30.22 0.000 green 27.20 0.000 30 blue 23.62 0.000 15 mg/dL @ 3 minutes red 5.39 0.025 green 8.29 0.003 blue 0.93 0.758 15 mg/dL @ 6 minutes red 1.77 0.320 green 4.31 0.072 blue 1.86 0.491 15 mg/dL @ 9 minutes red 7.18 0.016 green 5.27 0.115 blue 8.73 0.011 30 mg/dL @ 0 minutes red 27.48 0.000 green 26.93 0.000 blue 23.62 0.000 30 mg/dL @ 3 minutes red 3.37 0.135 green 3.91 0.110 blue 7.04 0.097 30 mg/dL @ 6 minutes red 0.00 0.998 green 0.23 0.905 blue 2.37 0.448 30 mg/dL @ 9 minutes red 5.76 0.027 green 5.88 0.051 blue 9.23 0.014 60 mg/dL @ 0 minutes red 17.26 0.000 green 15.52 0.000 blue 19.99 0.001 31 60 mg/dL @ 3 minutes red 11.97 0.000 green 14.71 0.000 blue 6.60 0.075 60 mg/dL @ 6 minutes red 10.04 0.000 green 11.96 0.000 blue 0.30 0.918 60 mg/dL @ 9 minutes red 1.67 0.464 green 3.47 0.228 blue 6.41 0.067 Table A - Raw LFA Data with mean RGB and p value for each concentration and time point Figure A - Mean RGB Across 4 Time Points for LFA Device 32 mPAD Mean RGB Concentration Channel color Mean RGB point value P-value 0.4685 mg/dL @ 0 minutes red 13.43 0.002 green 13.91 0.015 blue 4.95 0.250 0.4685 mg/dL @ 3 minutes red 6.27 0.188 green 9.27 0.210 blue 10.43 0.076 0.4685 mg/dL @ 6 minutes red 19.46 0.002 green 21.57 0.020 blue 17.24 0.010 0.4685 mg/dL @ 9 minutes red 18.22 0.001 green 21.57 0.006 blue 20.82 0.002 0.9375 mg/dL @ 0 minutes red 0.54 0.797 green 1.72 0.524 blue 3.18 0.483 0.9375 mg/dL @ 3 minutes red 0.15 0.950 green 2.17 0.550 blue 4.86 0.381 0.9375 mg/dL @ 6 minutes red 7.81 0.013 green 4.59 0.182 blue 4.16 0.452 0.9375 mg/dL @ 9 minutes red 12.59 0.003 green 7.46 0.041 33 blue 6.00 0.193 1.875 mg/dL @ 0 minutes red 6.54 0.257 green 12.42 0.152 blue 8.16 0.153 1.875 mg/dL @ 3 minutes red 8.60 0.159 green 15.81 0.086 blue 7.68 0.223 1.875 mg/dL @ 6 minutes red 5.88 0.253 green 12.32 0.112 blue 1.71 0.747 1.875 mg/dL @ 9 minutes red 5.33 0.341 green 11.61 0.167 blue 1.04 0.826 3.75", "7.81 0.013 green 4.59 0.182 blue 4.16 0.452 0.9375 mg/dL @ 9 minutes red 12.59 0.003 green 7.46 0.041 33 blue 6.00 0.193 1.875 mg/dL @ 0 minutes red 6.54 0.257 green 12.42 0.152 blue 8.16 0.153 1.875 mg/dL @ 3 minutes red 8.60 0.159 green 15.81 0.086 blue 7.68 0.223 1.875 mg/dL @ 6 minutes red 5.88 0.253 green 12.32 0.112 blue 1.71 0.747 1.875 mg/dL @ 9 minutes red 5.33 0.341 green 11.61 0.167 blue 1.04 0.826 3.75 mg/dL @ 0 minutes red 0.12 0.987 green 5.89 0.444 blue 2.46 0.759 3.75 mg/dL @ 3 minutes red 13.88 0.011 green 9.44 0.270 blue 14.39 0.108 3.75 mg/dL @ 6 minutes red 9.96 0.037 green 5.05 0.528 blue 10.37 0.201 3.75 mg/dL @ 9 minutes red 5.33 0.235 green 1.45 0.856 blue 10.42 0.180 34 7.5 mg/dL @ 0 minutes red 3.17 0.220 green 5.79 0.095 blue 3.02 0.552 7.5 mg/dL @ 3 minutes red 29.89 0.000 green 31.41 0.000 blue 19.69 0.028 7.5 mg/dL @ 6 minutes red 29.04 0.000 green 28.78 0.000 blue 10.30 0.075 7.5 mg/dL @ 9 minutes red 25.03 0.000 green 25.91 0.000 blue 5.46 0.294 15 mg/dL @ 0 minutes red 21.46 0.021 green 30.38 0.015 blue 15.22 0.028 15 mg/dL @ 3 minutes red 29.22 0.004 green 35.66 0.005 blue 13.28 0.074 15 mg/dL @ 6 minutes red 19.94 0.023 green 25.93 0.023 blue 6.40 0.292 15 mg/dL @ 9 minutes red 34.09 0.002 green 35.86 0.005 blue 7.26 0.188 30 mg/dL @ 0 minutes red 11.07 0.223 35 green 14.5 0.213 blue 3.18 0.645 30 mg/dL @ 3 minutes red 2.27 0.751 green 2.69 0.773 blue 5.80 0.345 30 mg/dL @ 6 minutes red 0.93 0.881 green 5.30 0.512 blue 7.15 0.207 30 mg/dL @ 9 minutes red 1.15 0.855 green 1.22 0.879 blue 13.02 0.017 60 mg/dL @ 0 minutes red 7.68 0.072 green 6.84 0.143 blue 2.65 0.549 60 mg/dL @ 3 minutes red 5.65 0.188 green 8.14 0.108 blue 2.71 0.622 60 mg/dL @ 6 minutes red 13.86 0.002 green 12.62 0.013 blue 3.17 0.608 60 mg/dL @ 9 minutes red 18.75 0.000 green 16.19 0.001 blue 3.10 0.527 Table B - Raw mPAD Data with mean RGB and p value for each concentration and time point 36 Figure B - Mean RGB Across 4 Time Points for mPAD LFA Device With Porcine Blood of Unknown Creatinine Concentration", "mg/dL @ 3 minutes red 5.65 0.188 green 8.14 0.108 blue 2.71 0.622 60 mg/dL @ 6 minutes red 13.86 0.002 green 12.62 0.013 blue 3.17 0.608 60 mg/dL @ 9 minutes red 18.75 0.000 green 16.19 0.001 blue 3.10 0.527 Table B - Raw mPAD Data with mean RGB and p value for each concentration and time point 36 Figure B - Mean RGB Across 4 Time Points for mPAD LFA Device With Porcine Blood of Unknown Creatinine Concentration Time Channel Color Mean RGB p-value 0 minutes red 37.77 0.000 green 54.13 0.000 blue 40.11 0.002 3 minutes red 45.18 0.000 green 68.15 0.000 blue 63.88 0.000 6 minutes red 53.81 0.000 green 73.38 0.000 blue 64.15 0.000 9 minutes red 40.28 0.000 37 green 62.30 0.000 blue 57.13 0.001 Table C - 2 sample t-Test for LFA device tested with porcine blood Based on the results in table C, we can conclude that testing the LFA test with porcine blood produced statistically significant results at all 3 RGB channels and at all 4 time points. This result was not what we expected, but it was likely the result of multiple sources of error. Discussion for Table C Testing the LFA device with porcine blood of an unknown concentration of creatinine, was based on curiosity, to see how the LFA device would perform. To get our blood results, we manufactured 6 LFA devices which were hand cut to the approximate size of the predetermined LFA components. We then performed multiple 2 sample t-Tests in the same fashion as the previous LFA and mPAD tests, comparing it to the RGB values of LFA device for 0 mg/dL. The results were statistically significant for all RGB channels and time points. We suspect that the colorimetric change was due to hemoglobin in the blood, and not due to creatinine reacting with alkaline picrate. Current literature suggests that LFA devices are better suited for measuring the presence of certain biomarkers in non-blood samples, compared to blood samples. This is due to the LFA devices lacking the ability to properly separate hemoglobin from blood. Since the reaction of creatinine and alkaline picrate produces a yellow-orange color, and we were unable to test the composition of the diagnostic pad after the test was complete, we can conclude that the color change could have been due to hemoglobin reacting with the alkaline picrate. One source of", "for measuring the presence of certain biomarkers in non-blood samples, compared to blood samples. This is due to the LFA devices lacking the ability to properly separate hemoglobin from blood. Since the reaction of creatinine and alkaline picrate produces a yellow-orange color, and we were unable to test the composition of the diagnostic pad after the test was complete, we can conclude that the color change could have been due to hemoglobin reacting with the alkaline picrate. One source of error which could have contributed to these results, could have been the quality or freshness of the porcine blood. The porcine blood we used was leftover from another Cardiolab team who tests with blood, and was left in the fridge for an unknown amount of time. Prior to the test, we noticed that the blood had separated and had to gently shake the bottle it was in to resuspend the blood. This could have been a source of error, as there could have been excess blood compounds in the samples we tested with, which could have greatly contributed to colorimetric change. Additionally, we statistically analyzed the data by comparing it to colorimetric change of the LFA 38 device when tested with PBS, a solution which does not directly minic the non-creatinine composition of blood. 39", "Detecting Thrombosis Onset in MHVs Through Changes in Valve Closing Sound Department of Biomedical Engineering, San Jose State University BME 198B: Senior Project Design II Joey Arey, Daniela Vivanco-Valdespino May 16, 2025 1 Table of Contents 1. Executive Summary 4 2. Literature Review 4 3. Biomedical Motivation 5 4. Statement of Need 6 5. Materials and Methods 7 I. Thrombogenicity Tester (TGT) Setup 8 II. Microphone and Audio Capture 8 III. Blood Trials 9 IV. Data Processing and Analysis 6. Results 10 I. Detection Metrics 10 II. Sound Signature Classification 10 III. Representative Figures 10 7. Discussion 12 I. Validity of Sound-Based Detection 13 II. Detection Across Thrombosis Stages 13 III. Limitations 13 IV. Statistical Significance 13 8. Conclusions 14 9. Future Work 14 I. Simulated Clot Modeling 15 II. Software Improvements 15 2 III. Signal Processing Enhancement 15 IV. Expanded Testing and Human Blood Trials 15 V. Clinical Translation 16 10. Safety 16 I. Biological Safety - Blood Handling 16 II. Thermal and Environmental Controls 16 III. Electrical and Equipment Safety 17 IV. Emergency Protocols 17 11. FDA Considerations 18 12. Acknowledgments 18 13. References 19 14. Cost Analysis 20 15. Appendix 22 3 Figure List Figure 1 - Successful Detection of Clot Formation This waveform shows a consistent drop in noise and sharpness once a clot has formed, compared to the baseline segment. Figure 2 - Progressive Thrombosis Signature Demonstrates clear transitions between the three clotting stages within a single 45-minute trial. Figure 3 - Audio Waveform vs. Flowmeter A waveform from a successful detection trial aligned with simultaneous flowmeter data, both showing a signal drop at the exact time of clot formation. Figure 4 - Unsuccessful Waveform The waveform generated is what we consider an unsuccessful detection. Thrombosis occurred during this trial but our audio detection method was unable to detect it Table List Table 1 The results of the our successful detections and our 95% confidence interval that aligns with our alternative hypothesis 4 Executive Summary Thrombosis in mechanical heart valves (MHVs) presents a significant clinical risk, potentially leading to impaired valve function and life-threatening complications. This project explores a novel, non-invasive approach to detect clot formation in MHVs by analyzing changes in valve closing sounds within a controlled simulation environment. Utilizing a Thrombogenicity Tester (TGT), we conducted 26 blood trials using porcine samples, of which 22 resulted in successful detections, yielding a clot detection success", "interval that aligns with our alternative hypothesis 4 Executive Summary Thrombosis in mechanical heart valves (MHVs) presents a significant clinical risk, potentially leading to impaired valve function and life-threatening complications. This project explores a novel, non-invasive approach to detect clot formation in MHVs by analyzing changes in valve closing sounds within a controlled simulation environment. Utilizing a Thrombogenicity Tester (TGT), we conducted 26 blood trials using porcine samples, of which 22 resulted in successful detections, yielding a clot detection success rate of 84.61%. The methodology centered on identifying acoustic signatures corresponding to different clotting stages\u2014no clot, partial clot, and full thrombosis\u2014through real-time sound monitoring and waveform analysis. Signal processing improvements were applied to enhance detection accuracy, and comparisons were made between pre- and post-clot audio signals. Statistical analysis, including confidence intervals and hypothesis testing, was used to validate the effectiveness of the sound-based detection system. This research demonstrates the feasibility of using audio data as a diagnostic tool to detect thrombosis onset in MHVs. The findings provide foundational data for future expansion into clinical applications, potentially improving early diagnosis and patient outcomes for individuals with mechanical heart valves. Literature Review Mechanical heart valves (MHVs) are commonly used prosthetic devices in patients requiring valve replacement; however, they carry a known risk of thrombosis formation. The incidence of thrombosis ranges from 0.5% to 8% in left-sided valves and can reach up to 20% in 5 right-sided valves, posing a serious health concern due to the potential for embolic events and valve obstruction [1,5]. Thrombosis in MHVs is primarily attributed to flow disruptions, shear stress variations, and surface interactions between blood components and the valve material. Research has shown that the complex hemodynamics around bileaflet MHVs, including flow separation and turbulence, can initiate platelet activation and thrombus formation [3,4]. Computational studies and in-vitro flow visualizations have been conducted to characterize these dynamics, but detecting clot formation in real time remains a significant challenge [2,8]. Traditional detection methods rely on imaging and flow-based techniques, which are either invasive, expensive, or unsuitable for continuous monitoring. This has led to an interest in non-invasive alternatives, including the use of acoustic signals. Prior studies have demonstrated that MHVs produce distinct sounds during opening and closing, and that the presence of a clot can alter the frequency, amplitude, and waveform of these signals [7]. Building on these findings, this project explores whether audio recordings from MHVs can be", "[2,8]. Traditional detection methods rely on imaging and flow-based techniques, which are either invasive, expensive, or unsuitable for continuous monitoring. This has led to an interest in non-invasive alternatives, including the use of acoustic signals. Prior studies have demonstrated that MHVs produce distinct sounds during opening and closing, and that the presence of a clot can alter the frequency, amplitude, and waveform of these signals [7]. Building on these findings, this project explores whether audio recordings from MHVs can be leveraged to identify clot formation events. Our approach uses a TGT simulation with porcine blood and real-time audio capture to analyze sound changes corresponding to different thrombosis stages. This method aims to provide a low-cost, non-invasive, and real-time tool for thrombosis detection, improving early intervention potential. Biomedical Motivation Mechanical heart valves (MHVs) play a critical role in restoring function in patients with valvular heart disease. However, one of their most serious and frequent complications is thrombosis, which can lead to catastrophic embolic events or obstruct valve function altogether. Patients with MHVs typically require lifelong anticoagulation therapy to mitigate this risk, but 6 this treatment increases the chances of bleeding and still does not eliminate the possibility of clot formation. Early and reliable detection of thrombosis is, therefore, essential for timely intervention and improved patient outcomes. Despite significant advancements in prosthetic valve design, a dependable method for continuous, non-invasive thrombosis monitoring does not exist. Existing diagnostic tools such as echocardiography or CT imaging are costly, operator-dependent, and impractical for real-time, routine monitoring. A low-cost, real-time monitoring system that can detect clot formation in its early stages would be a game-changing innovation in prosthetic heart valve management. Sound-based diagnostic approaches have emerged as a promising alternative. Mechanical heart valves inherently produce distinct acoustic profiles with every opening and closing cycle. These profiles may change predictably in response to clot formation, due to altered leaflet motion, dampened closure forces, or disrupted flow acoustics. By identifying these changes with signal processing techniques, we can transform heart valve sounds into biomarkers of thrombotic events. This project aims to bridge this gap by evaluating whether thrombosis onset can be reliably detected through changes in MHV closing sound. Such a tool would support earlier detection, reduce reliance on imaging, and potentially reduce patient morbidity and healthcare costs. Ultimately, this research supports the long-term goal of improving quality of care for patients living with mechanical heart valves. Statement", "By identifying these changes with signal processing techniques, we can transform heart valve sounds into biomarkers of thrombotic events. This project aims to bridge this gap by evaluating whether thrombosis onset can be reliably detected through changes in MHV closing sound. Such a tool would support earlier detection, reduce reliance on imaging, and potentially reduce patient morbidity and healthcare costs. Ultimately, this research supports the long-term goal of improving quality of care for patients living with mechanical heart valves. Statement of Need Thrombosis in mechanical heart valves (MHVs) represents a persistent and life-threatening complication for patients worldwide. Although the implantation of MHVs restores critical hemodynamic function, the risk of clot formation mandates the use of 7 anticoagulants such as warfarin. However, these treatments require continuous monitoring, are prone to patient noncompliance and elevate the risk of hemorrhagic events. More importantly, current clinical protocols offer no method for continuous, non-invasive, and real-time monitoring of clot development within MHVs. Present diagnostic techniques\u2014such as transesophageal echocardiography, CT scans, or Doppler ultrasound\u2014while effective, are limited by their cost, accessibility, and inability to provide continuous clot surveillance. As a result, clot formation may go undetected until symptoms appear or a thromboembolic event occurs, by which point the patient\u2019s risk has significantly escalated. Given this unmet clinical need, a real-time and non-invasive clot detection system would provide significant value. A solution based on sound waveform analysis could be embedded into a portable or implantable monitoring device and alert clinicians to early thrombosis formation, allowing for more timely intervention. This detection method could also be used in other in vitro simulations that work to improve the MHV safety. This project seeks to develop and validate such a detection method using signal analysis of MHV closing sounds. By confirming that clot formation alters valve acoustics in predictable and statistically significant ways, this study lays the groundwork for future diagnostic tools that can be deployed at the bedside or integrated into remote patient monitoring systems. Materials and Methods I. Thrombogenicity Tester (TGT) Setup The core experimental platform for this study was the Thrombogenicity Tester (TGT), a benchtop system designed to simulate blood flow through a bileaflet mechanical heart valve 8 (MHV). The TGT, which consists of two primary components: the tygon tubing setup and the TGT motor. The tygon tubing setup includes two tygon tubes and two valves\u2014one housing the MHV and another serving as a fill valve\u2014which was", "integrated into remote patient monitoring systems. Materials and Methods I. Thrombogenicity Tester (TGT) Setup The core experimental platform for this study was the Thrombogenicity Tester (TGT), a benchtop system designed to simulate blood flow through a bileaflet mechanical heart valve 8 (MHV). The TGT, which consists of two primary components: the tygon tubing setup and the TGT motor. The tygon tubing setup includes two tygon tubes and two valves\u2014one housing the MHV and another serving as a fill valve\u2014which was filled with porcine blood. The tygon setup is mounted on the TGT motor, a disk plate attached to a motor mechanism programmed to simulate systolic and diastolic cycles through back-and-forth rotations at adjustable speeds. As the motor operates, the MHV opens and closes in response to the inertia of the flowing liquid, replicating the mechanical behavior of heart valves under realistic circulatory conditions. The fully set up TGT simulation was then put into an incubator to replicate a human physiological temperature of 37\u00b0C. The system was enclosed in an incubator to maintain stable thermal conditions and minimize environmental noise. The four different MHV models were tested across multiple trials. A high-sensitivity microphone was fixed directly above the valve chamber to record the closing sounds of the MHV during simulated cardiac cycles. For additional information about how each trial was set up, see Appendix A. II. Microphone and Audio Capture Sound data was captured using a studio-grade condenser microphone with directional sensitivity optimized for high-frequency valve clicks. The microphone was connected to a digital audio interface, and recordings were made using Adobe Audition and Audacity. Due to software limitations with autoscaling on these platforms, alternate audio visualization tools (e.g., Raven Lite) were considered but not implemented during the initial testing phase. III. Blood Trials 9 A total of 26 porcine blood trials were conducted to simulate real clot formation. Each trial followed one of two protocols: \u25cf 25-Minute Trials: Used for rapid detection tests. These trials were stopped immediately upon detection of a suspected clot based on waveform changes. \u25cf 45-Minute Trials: Allowed full progression of clot formation to observe distinct audio signatures of three thrombosis stages: no clot, partial clot, and full clot (Figure 2). In both protocols, blood was freshly sourced, and samples were agitated under flow to promote thrombus development on the valve surface. Trials were terminated if significant changes were observed in amplitude, frequency, or waveform profile.", "rapid detection tests. These trials were stopped immediately upon detection of a suspected clot based on waveform changes. \u25cf 45-Minute Trials: Allowed full progression of clot formation to observe distinct audio signatures of three thrombosis stages: no clot, partial clot, and full clot (Figure 2). In both protocols, blood was freshly sourced, and samples were agitated under flow to promote thrombus development on the valve surface. Trials were terminated if significant changes were observed in amplitude, frequency, or waveform profile. IV. Data Processing and Analysis All recorded audio was exported in WAV format and segmented by time intervals. Waveform analysis was conducted by comparing valve sounds across three stages: baseline (no clot), during partial clotting, and full clot obstruction. Key metrics included: \u25cf Peak amplitude shifts \u25cf Frequency profile distortion \u25cf Loss or attenuation of closure sounds Success was defined as a distinct and consistent waveform deviation from baseline when a clot formed; see figure 1 for reference. An unsuccessful trial would consist of a waveform with no drop in sound level for a consistent period of time; see figure 4 for a visual. To evaluate statistical reliability, the success rate of clot detection was tested against a null hypothesis using a one-proportion z-test. A 95% confidence interval was computed for the proportion of successful detections. 10 Results A total of 26 porcine blood trials were conducted using the TGT system to evaluate the feasibility of detecting thrombus formation in MHVs through unfiltered sound waveform analysis. Of these, 22 trials resulted in successful clot detection based on clearly identifiable changes in the valve closing sound waveform. This corresponds to a success rate of 84.61%, with a 95% confidence interval of (70.7%, 98.5%) and a margin of error of \u00b113.9%. I. Detection Metrics A confidence interval was conducted to validate the significance of this detection rate against the null hypothesis that the sound-based method performs no better than an 80% success rate (see Appendix B for table). The observed proportion of successful trials supports rejection of the null hypothesis, suggesting the method is statistically significant for clot detection. II. Sound Signature Classification Two distinct experimental protocols were used to characterize the stages of thrombosis: \u25cf 25-Minute Tests (Early Detection): These were designed to confirm whether a sound change could be used as a prompt to terminate a test. Most of the successful trials showed an abrupt amplitude drop or altered waveform", "rate (see Appendix B for table). The observed proportion of successful trials supports rejection of the null hypothesis, suggesting the method is statistically significant for clot detection. II. Sound Signature Classification Two distinct experimental protocols were used to characterize the stages of thrombosis: \u25cf 25-Minute Tests (Early Detection): These were designed to confirm whether a sound change could be used as a prompt to terminate a test. Most of the successful trials showed an abrupt amplitude drop or altered waveform envelope corresponding with clot initiation. \u25cf 45-Minute Tests (Progression Analysis): These longer trials captured the full transition from no clot to full clot. In 5 out of 10 extended tests, all three clot stages were observed: 1. No Clot: Sharp, high-frequency valve click with consistent amplitude. 2. Partial Clot: Decreased amplitude and slight waveform smearing. 3. Full Clot: Abrupt loss of valve click or highly dampened signal. III. Representative Figures 11 Figure 1: Successful Detection of Clot Formation - This waveform shows a consistent drop in noise and sharpness once a clot has formed, compared to the baseline segment. Figure 2: Progressive Thrombosis Signature - Demonstrates clear transitions between the three clotting stages within a single 45-minute trial. 12 Figure 3: Audio Waveform vs. Flowmeter - A waveform from a successful detection trial aligned with simultaneous flowmeter data, both showing a signal drop at the exact time of clot formation. Discussion The results of this study demonstrate the potential of using audio signal analysis to detect thrombosis in mechanical heart valves (MHVs) under simulated physiological conditions. With a successful detection rate of 84.61%, the proposed method offers a promising foundation for developing a real-time, non-invasive monitoring system. I. Validity of Sound-Based Detection The detection of clot formation was consistently correlated with changes in the valve\u2019s closing sound, particularly in amplitude, waveform shape, and frequency content. The valve\u2019s characteristic \u201cclick\u201d diminished or vanished in the presence of full clot occlusion, while partial thrombosis typically caused gradual attenuation and distortion. These acoustic patterns were visible in the waveform plots and confirmed via comparison to flowmeter readings where 13 available. The comparison with the flowmeter readings, shown in figure 3, validates that our sound detection method was able to detect the exact moment the thrombosis began to form. The high success rate, supported by a narrow confidence interval, suggests that this method is robust under the controlled conditions used in the TGT. Moreover,", "partial thrombosis typically caused gradual attenuation and distortion. These acoustic patterns were visible in the waveform plots and confirmed via comparison to flowmeter readings where 13 available. The comparison with the flowmeter readings, shown in figure 3, validates that our sound detection method was able to detect the exact moment the thrombosis began to form. The high success rate, supported by a narrow confidence interval, suggests that this method is robust under the controlled conditions used in the TGT. Moreover, waveform changes were observable without needing invasive imaging or direct visualization of the clot, meeting the criteria for a minimally intrusive detection mechanism. II. Detection Across Thrombosis Stages The ability to distinguish between no clot, partial clot, and full clot conditions in 50% of the 45-minute trials demonstrates another key strength of this approach. Capturing the full progression of clot development may be crucial for clinical decision-making, enabling interventions before full obstruction occurs. However, improvements in consistency are still needed, as not all trials captured the intermediate clotting stage. III. Limitations There were several limitations encountered during the study: \u25cf Software Constraints: The inability to autoscale in Adobe Audition and Audacity reduced real-time visibility of waveform shifts. This limited the responsiveness of early detection and necessitated manual post-analysis. \u25cf Environmental Noise: Despite incubator insulation, external noise sources occasionally introduced artifacts into the waveform data. \u25cf Clot Confirmation: Because clot formation was inferred from acoustic changes, there was no direct imaging or microscopic validation of thrombus presence during the trials. While the porcine blood model is biologically relevant, the results in human blood may differ due to clotting kinetics, viscosity, or valve biointeractions. 14 IV. Statistical Significance The statistical strength of the results was validated using a one-proportion z-test and confidence interval analysis. The observed success rate exceeded the null hypothesis threshold of 80%, suggesting the method offers a measurable improvement over non-detection or subjective observation approaches. Conclusions This study successfully demonstrated that acoustic analysis of valve closing sounds can be used to detect thrombosis formation in bileaflet mechanical heart valves (MHVs) with a high degree of accuracy. Through the use of a Thrombogenicity Tester (TGT) and porcine blood trials, we showed that consistent and measurable changes in waveform features\u2014such as amplitude reduction and signal disruption\u2014occur when a clot forms on or near the valve. Out of 26 total trials, 22 exhibited clearly detectable audio changes correlating with thrombus presence, yielding", "demonstrated that acoustic analysis of valve closing sounds can be used to detect thrombosis formation in bileaflet mechanical heart valves (MHVs) with a high degree of accuracy. Through the use of a Thrombogenicity Tester (TGT) and porcine blood trials, we showed that consistent and measurable changes in waveform features\u2014such as amplitude reduction and signal disruption\u2014occur when a clot forms on or near the valve. Out of 26 total trials, 22 exhibited clearly detectable audio changes correlating with thrombus presence, yielding a success rate of 84.61%. These findings support the potential of sound-based thrombosis detection as a non-invasive, real-time monitoring method. In longer trials, distinct sound patterns corresponding to the three stages of thrombosis (no clot, partial clot, and full clot) were observed, further emphasizing the diagnostic potential of this approach. While technical limitations in real-time waveform visualization and environmental noise remain areas for refinement, the data collected during this project validates the central hypothesis: that thrombosis onset can be reliably identified via sound signature analysis. This project establishes the groundwork for the future creation of an in vitro diagnostic system that may support MHV experiments, ultimately leading to better patient outcomes and safety. 15 Future Work While this study confirmed the feasibility of detecting thrombus formation through audio analysis in a controlled environment, further development is required to refine and validate this technology for real-world applications. I. Simulated Clot Modeling Future experiments will use synthetic or pre-formed clots to simulate the three thrombosis stages\u2014no clot, partial obstruction, and full occlusion\u2014in a more controlled and reproducible manner. This will allow for better characterization of audio signatures associated with each stage and reduce the variability associated with spontaneous clot formation in blood. II. Software Improvements To improve real-time monitoring capabilities, alternative audio analysis software platforms will be explored. Specifically, software that allows autoscaling and real-time frequency domain analysis will be integrated into the testing pipeline. Raven Lite and custom MATLAB scripts are potential candidates. III. Signal Processing Enhancement Additional digital signal processing techniques such as Fourier transform, wavelet analysis, and machine learning-based pattern recognition could significantly improve classification accuracy. These tools could enable automated detection and alert systems capable of distinguishing between normal function and various stages of clotting with high specificity. IV. Expanded Testing & Human Blood Trials Further testing with larger sample sizes and under varied physiological conditions\u2014including varying flow rates, temperatures, and valve models\u2014is necessary to 16 generalize the", "potential candidates. III. Signal Processing Enhancement Additional digital signal processing techniques such as Fourier transform, wavelet analysis, and machine learning-based pattern recognition could significantly improve classification accuracy. These tools could enable automated detection and alert systems capable of distinguishing between normal function and various stages of clotting with high specificity. IV. Expanded Testing & Human Blood Trials Further testing with larger sample sizes and under varied physiological conditions\u2014including varying flow rates, temperatures, and valve models\u2014is necessary to 16 generalize the findings. Longer-term, transitioning to in vitro trials with human blood or ex vivo heart-lung circuits could provide critical validation of the system\u2019s translational potential. V. Clinical Translation The ultimate goal is to develop an acoustic monitoring device that could provide continuous surveillance of MHV function in in vitro settings to help improve MHV safety. Integrating this system into simulated monitoring frameworks could reduce dependency on utilizing random chance and improve MHV to lower hospitalization rates and improve long-term patient management. Safety Safety considerations were an integral part of this project due to the handling of biological fluids and use of electronic equipment in a controlled environment. I. Biological Safety\u2014Blood Handling All porcine blood samples were treated as biohazardous materials and were handled according to institutional biosafety guidelines. The following procedures were followed: \u25cf Personal Protective Equipment (PPE) including lab coats, nitrile gloves, and eye protection, was worn at all times. \u25cf Blood samples were stored in clearly labeled, sealed containers and handled within a biological safety cabinet when possible. \u25cf All surfaces and tools in contact with blood were decontaminated using 10% bleach solution before and after use. \u25cf Contaminated consumables (e.g., pipette tips, tubing, gloves) were disposed of in biohazard waste containers in compliance with lab safety regulations. 17 II. Thermal and Environmental Controls The TGT simulation chamber operated continuously at 37\u00b0C to simulate physiological temperature. Precautions were taken to avoid burns or overheating: \u25cf All incubators were regularly monitored with external thermometers. \u25cf Power cords and heated components were inspected before each use. \u25cf Ventilation was maintained to prevent overheating of surrounding electronic devices. III. Electrical and Equipment Safety Audio capture equipment, including microphones and interfaces, operated in close proximity to moist or potentially conductive surfaces. To reduce the risk of electrical hazards: \u25cf All electronic devices were elevated or isolated from fluid-containing components. \u25cf Only low-voltage, properly grounded equipment was used inside the incubator environment. \u25cf Extension", "monitored with external thermometers. \u25cf Power cords and heated components were inspected before each use. \u25cf Ventilation was maintained to prevent overheating of surrounding electronic devices. III. Electrical and Equipment Safety Audio capture equipment, including microphones and interfaces, operated in close proximity to moist or potentially conductive surfaces. To reduce the risk of electrical hazards: \u25cf All electronic devices were elevated or isolated from fluid-containing components. \u25cf Only low-voltage, properly grounded equipment was used inside the incubator environment. \u25cf Extension cords and surge protectors were kept away from direct contact with blood or tubing. IV. Emergency Protocols In case of spills or exposure, emergency eyewash stations and spill kits were located within immediate reach of the testing area. A spill response procedure was reviewed with all team members before beginning trials. By adhering to these safety protocols, the research was conducted without incident, ensuring the health and safety of all personnel involved. FDA Considerations Any future use of a sound-based thrombosis detection system for mechanical heart valves (MHVs) would most likely fall under the FDA's Center for Devices and Radiological Health (CDRH) category of medical devices in the context of clinical translation. The 18 system\u2014particularly if integrated into a real-time monitoring device\u2014may fall under Class II devices due to its diagnostic function and moderate risk profile. If the system progresses toward in vivo application or patient monitoring, an Investigational Device Exemption (IDE) approval would be necessary for clinical trials. These trials would provide critical safety and efficacy data for FDA submission. Acknowledgments We would like to express our sincere gratitude to the following individuals and institutions whose support made this project possible: Faculty and Research Advisors Dr. Alessandro Bellofiore Project Advisor, Department of Biomedical Engineering San Jos\u00e9 State University Dr. Sreyashi Chakraborty Consulting Faculty, Department of Biomedical Engineering San Jos\u00e9 State University Technical Support and Lab Assistance Santosh Dasari and Yhanira Amaro For their assistance with porcine blood handling and experimental setup. Sound Detection Team Daniela Vivanco Valdespino Audio data analysis and experimental planning. Joey Arey TGT simulation management, experimental coordination, and statistical validation. 19 Special Thanks: To the San Jos\u00e9 State University Biomedical Engineering Department for providing access to equipment, laboratory space, and software resources essential to the success of this project. We also thank the Institutional Biosafety Office for guidance on handling biological materials. Finally, we are grateful to our peers and families for their support and encouragement", "experimental setup. Sound Detection Team Daniela Vivanco Valdespino Audio data analysis and experimental planning. Joey Arey TGT simulation management, experimental coordination, and statistical validation. 19 Special Thanks: To the San Jos\u00e9 State University Biomedical Engineering Department for providing access to equipment, laboratory space, and software resources essential to the success of this project. We also thank the Institutional Biosafety Office for guidance on handling biological materials. Finally, we are grateful to our peers and families for their support and encouragement throughout the year-long effort behind this senior design project. References 1. Chambers, J. B., Pomar, J. L., Mestres, C. A., & Palatianos, G. M. (2013). Clinical event rates with the On-X bileaflet mechanical heart valve: A multicenter experience with follow-up to 12 years. The Journal of Thoracic and Cardiovascular Surgery , 145(2), 420\u2013424. https://doi.org/10.1016/j.jtcvs.2011.12.059 2. Department of Cardiac Surgery. (2009). In-vitro localization of initial flow-induced thrombus formation. ASAIO Journal . https://journals.lww.com/asaiojournal/Abstract/2009/01000/In_vitro_localization_of_initi al_flow_induced.6.aspx 3. Govindarajan, V., Udaykumar, H. S., Herbertson, L. H., Deutsch, S., Manning, K. B., & Chandran, K. B. (2009). Impact of design parameters on bileaflet mechanical heart valve flow dynamics. Journal of Heart Valve Disease , 18(5), 535\u2013545. 4. Hatoum, H., Maureira, P., & Dasi, L. P. (2020). A turbulence in vitro assessment of On-X and St. Jude Medical prostheses. The Journal of Thoracic and Cardiovascular Surgery . https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6703975/ 20 5. Mayo Foundation for Medical Education and Research. (2015). Prosthetic valve thrombosis: Time is critical. Mayo Clinic . https://www.mayoclinic.org/medical-professionals/cardiovascular-diseases/news/prostheti c-valve-thrombosis-time-is-critical/mac-20430866 6. Strong, E. B., Knutsen, C., Wells, J. T., Jangid, A. R., Mitchell, M. L., Martinez, N. W., & Martinez, A. W. (2019). Wax-printed fluidic time delays for automating multi-step assays in paper-based microfluidic devices. Micromachines , 4(1), 20. https://www.mdpi.com/2411-5134/4/1/20 7. Tosoni, S., Tarzia, V., Colli, A., Gerosa, G., & Bagno, A. (2017). Phonographic detection of mechanical heart valve thrombosis. Journal of Artificial Organs , 20(4), 394\u2013398. https://doi.org/10.1007/s10047-017-1002-5 8. Zakaria, M. S., Ismail, F., Tamagawa, M., Aziz, A. F. A., Wiriadidjaja, S., Basri, A. A., & Ahmad, K. A. (n.d.). Review of numerical methods for simulation of mechanical heart valves and the potential for blood clotting. Medical & Biological Engineering & Computing . https://pubmed.ncbi.nlm.nih.gov/28744828/ 9. Cancer Therapy Advisor. (n.d.). Prosthetic heart valves overview. https://www.cancertherapyadvisor.com/home/decision-support-in-medicine/hospital-medi cine/-heart-valves/ Cost Analysis The total estimated cost of materials, equipment, and software used throughout this project is outlined below. These estimates include both reusable equipment and consumable 21 materials required for each blood trial.", "F. A., Wiriadidjaja, S., Basri, A. A., & Ahmad, K. A. (n.d.). Review of numerical methods for simulation of mechanical heart valves and the potential for blood clotting. Medical & Biological Engineering & Computing . https://pubmed.ncbi.nlm.nih.gov/28744828/ 9. Cancer Therapy Advisor. (n.d.). Prosthetic heart valves overview. https://www.cancertherapyadvisor.com/home/decision-support-in-medicine/hospital-medi cine/-heart-valves/ Cost Analysis The total estimated cost of materials, equipment, and software used throughout this project is outlined below. These estimates include both reusable equipment and consumable 21 materials required for each blood trial. Most equipment was sourced through the San Jos\u00e9 State University Biomedical Engineering Department. Category Item Cost (USD) Simulation Hardware Thrombogenicity Tester (TGT) chamber (x2) $1,000 (est.) Incubators (x2) $1,500 Flowmeter $1,200 MHV prototypes (4 types) $800 Tubing, pipettes, containers (per trial) ~$15 per trial \u00d7 26 = $390 Audio Equipment Wireless GO II Microphone (2x) $300 Audio interface (USB-powered) $150 Cables and accessories $100 Software Adobe Audition (academic license) Institutional license Audacity (free) $0 22 Microsoft Excel / MATLAB (university provided) Institutional license Biological Materials Porcine blood (per 500 mL) ~$30 per trial \u00d7 26 = $780 Safety Supplies Gloves, bleach, disinfectants $100 Biohazard disposal containers $50 Total Estimated Cost ~$6,370 Appendix I. Appendix A - Lab procedures A. MHV Mass Measurement 1. Gather all four Mechanical Heart Valves (MHVs). 2. Use a precision balance to measure and record the initial mass of each MHV before exposure to blood. 3. Label each MHV clearly with a unique ID number. B. Protamine Cocktail Preparation 1. Prepare four protamine cocktails as follows: - Mix 200 mg of protamine with 2 mL of buffer solution. 2. Stir until the solution is fully dissolved. 3. Label each vial and store at room temperature until use. C. Porcine Blood Setup and Storage 1. Start Blood Cytometer: - Power on and calibrate the cytometer using the porcine blood sample from Lampire. - Record initial hematocrit readings. 2. Combine Blood: - Pour the contents of all four porcine blood bottles into a large measuring cup. - Mix thoroughly using a clean glass stirring rod. - Clean and rinse all four original bottles for reuse. 23 3. Adjust Hematocrit: - Target hematocrit range: 36% \u2013 42%. - If below range, add 100 mL of NaCl, mix, and remeasure. - Repeat until within the desired range. 4. Re-bottle and Store: - Divide the blood evenly (~450 mL per bottle) into the cleaned bottles. - Store at room temperature until", "porcine blood bottles into a large measuring cup. - Mix thoroughly using a clean glass stirring rod. - Clean and rinse all four original bottles for reuse. 23 3. Adjust Hematocrit: - Target hematocrit range: 36% \u2013 42%. - If below range, add 100 mL of NaCl, mix, and remeasure. - Repeat until within the desired range. 4. Re-bottle and Store: - Divide the blood evenly (~450 mL per bottle) into the cleaned bottles. - Store at room temperature until testing. D. TGT Motor and Arduino Setup 1. Power and Connection: - Plug in the TGT motor to the power source. - Confirm that all Arduino Uno pins are properly inserted. 2. Program Arduino: - Set the experiment run time (25\u201345 minutes) in the Arduino code. - Upload the code to the Arduino board. 3. Finalize TGT Setup: - Ensure the TGT disk is clean and accessible. - Prepare rod supports with velcro straps to later secure tubing in place. - Ensure the USB cable from the Arduino is connected and ready for power-up. E. Porcine Blood Trial Setup 1. Valve Housing and Tygon Tubing Setup 1. Insert and Secure MHV: - Place one MHV into a valve housing. - Secure the valve using screws to lock it into place. 2. Label the Housing: - Apply a strip of tape to the valve housing. - Clearly label the MHV number and mark the flow direction with an arrow. 3. Attach Tygon Tubes to Housing: - Connect two Tygon tubes to the inlet and outlet ports of the valve housing. - Secure all connections with PVC clamps. 4. Connect Tubes to Fill Port: - Take the open ends of the two Tygon tubes and connect them to the fill port assembly. - Use additional PVC clamps to ensure a sealed system. 2. Fill Tubing with Porcine Blood 1. Use a funnel at the fill valve to pour porcine blood into the tubing system. 2. Continue filling until all air bubbles are removed from the setup. 3. Pre-Incubation Measurements 1. Place the filled Tygon setup on top of the TGT (Thrombogenicity Tester) disk. 2. Use the flow meter to collect 10 seconds of baseline flow data. 24 4. Add Protamine Cocktail 1. Inject one vial of protamine cocktail into the filled tubing system after the initial flow readings. 5. Incubation and Recording Setup ( Set to 37C @ Start of Day) 1.", "tubing system. 2. Continue filling until all air bubbles are removed from the setup. 3. Pre-Incubation Measurements 1. Place the filled Tygon setup on top of the TGT (Thrombogenicity Tester) disk. 2. Use the flow meter to collect 10 seconds of baseline flow data. 24 4. Add Protamine Cocktail 1. Inject one vial of protamine cocktail into the filled tubing system after the initial flow readings. 5. Incubation and Recording Setup ( Set to 37C @ Start of Day) 1. Transfer the filled and sealed Tygon setup to the incubator, maintaining orientation on the TGT disk and within 5 minutes after adding the protamine. 2. Place the microphone above the valve housing and secure it using a velcro strap. 3. Close the incubator door and begin audio recording. 6. Post-Trial Steps 1. After incubation, stop the TGT motor. 2. Remove TGT from the incubator and place it back on the shelf. 3. Use the flow meter to collect the post-experiment flow profile. 4. Extract the MHV from the housing and weigh it to assess clot formation. 5. Analyze the 10 blood samples using the blood cytometer. II. Appendix B - Statistical Analysis Results (Table 1) Metric Value Total Trials 26 Successful Detections 22 Detection Rate 84.61% Confidence Level 95% Margin of Error \u00b113.9% Confidence Interval (0.707, 0.985) Table 1: The results of the our successful detections and our 95% confidence interval that aligns with our alternative hypothesis III. Appendix C - Unsuccessful Waveform 25 Figure 4: Unsuccessful Waveform - The waveform generated is what we consider an unsuccessful detection. Thrombosis occurred during this trial but our audio detection method was unable to detect it"]