Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,149 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: cc-by-nc-4.0
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-nc-4.0
|
| 3 |
+
task_categories:
|
| 4 |
+
- image-segmentation
|
| 5 |
+
- surgical-workflow
|
| 6 |
+
tags:
|
| 7 |
+
- medical
|
| 8 |
+
- microsurgery
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
# MAVIS (Micro-surgical Artificial Vascular Anastomosis)
|
| 12 |
+
|
| 13 |
+
## Dataset Overview
|
| 14 |
+
|
| 15 |
+
MAVIS is a microsurgical dataset comprising 19 videos of artificial vascular anastomosis procedures performed by three expert microsurgeons at Korea University. For each video frame, it provides:
|
| 16 |
+
|
| 17 |
+
- **Pixel-level segmentation** of seven tool categories
|
| 18 |
+
- **Frame-level workflow annotations**: surgical stage, phase, and step
|
| 19 |
+
|
| 20 |
+
This dataset enables research on both surgical tool segmentation and surgical workflow recognition in ultra-fine microsurgical environments.
|
| 21 |
+
|
| 22 |
+
## Data Collection
|
| 23 |
+
|
| 24 |
+
1. **Subjects & Cases**
|
| 25 |
+
- 19 recorded anastomosis sessions on an artificial vessel simulator
|
| 26 |
+
- Surgeon assignments:
|
| 27 |
+
- CASE 01–07: Surgeon 1 (Jaemin Lee)
|
| 28 |
+
- CASE 08–14: Surgeon 2 (Yeongyun Ko)
|
| 29 |
+
- CASE 15–19: Surgeon 3 (Jaejun Nam)
|
| 30 |
+
|
| 31 |
+
2. **Acquisition Setup**
|
| 32 |
+
- Microscope/Camera model: ???
|
| 33 |
+
- Original resolution: ??? px
|
| 34 |
+
- Frame rate: ??? fps
|
| 35 |
+
- Cropped frame size: 1920 × 1072 px
|
| 36 |
+
|
| 37 |
+
3. **Annotation Tools & Process**
|
| 38 |
+
- **Segmentation**: ???
|
| 39 |
+
- **Workflow**: Manual tagging of stage/phase/step by the non-medical
|
| 40 |
+
|
| 41 |
+
4. **Annotators**
|
| 42 |
+
- (Three fellowship-trained microsurgeons)(확인 필요)
|
| 43 |
+
- Labels applied following a standardized workflow guideline
|
| 44 |
+
|
| 45 |
+
## Data Details
|
| 46 |
+
|
| 47 |
+
### Directory Structure
|
| 48 |
+
|
| 49 |
+
```
|
| 50 |
+
MAVIS
|
| 51 |
+
├── frames
|
| 52 |
+
│ ├── CASE01
|
| 53 |
+
│ │ ├── image_00001.jpg
|
| 54 |
+
│ │ └── ...
|
| 55 |
+
│ ├── ...
|
| 56 |
+
├── annotations
|
| 57 |
+
│ ├── long-term.json
|
| 58 |
+
│ ├── short-term.json
|
| 59 |
+
│ ├── segmentations
|
| 60 |
+
│ │ ├── CASE01
|
| 61 |
+
│ │ │ ├── image_00001.png
|
| 62 |
+
│ │ │ └── ...
|
| 63 |
+
│ │ ├── ...
|
| 64 |
+
│ └── segmentations_with_keypoint
|
| 65 |
+
│ ├── CASE01
|
| 66 |
+
│ │ ├── image_00001.png
|
| 67 |
+
│ │ └── ...
|
| 68 |
+
│ ├── ...
|
| 69 |
+
├── fig
|
| 70 |
+
└── README.md
|
| 71 |
+
```
|
| 72 |
+
|
| 73 |
+
### Annotation Formats
|
| 74 |
+
|
| 75 |
+
- **short-term.json**
|
| 76 |
+
- For each frame: polygon mask data for seven tool classes(CASE마다 약 64장 정도씩만 있는 것 언급하는게 좋을지?)
|
| 77 |
+
- **long-term.json**
|
| 78 |
+
- For each frame: stage, phase, and step labels
|
| 79 |
+
|
| 80 |
+
|
| 81 |
+
### Stage–Phase–Step Hierarchy
|
| 82 |
+
|
| 83 |
+
The workflow annotations are organized into six **Stages**, each containing one or more **Phases**, which in turn consist of individual **Steps**. Below is the full breakdown with descriptions:
|
| 84 |
+
|
| 85 |
+
1. **First tying** (forming the first knot)
|
| 86 |
+
1. **Phase: Suturing** – place and position the suture
|
| 87 |
+
- **Needle holding**: grasp the suture needle securely with the needle holder
|
| 88 |
+
- **Needle passing**: insert the needle through both edges of the vessel and pull it through
|
| 89 |
+
- **Needle dropping**: release the needle at the optimal position for tying (≈5 o’clock)
|
| 90 |
+
2. **Phase: Knot tying** – create the knot
|
| 91 |
+
- **1st knot**: wrap the free end of the suture around the instrument and tighten
|
| 92 |
+
- **2nd knot**: repeat wrapping in the opposite direction and tighten
|
| 93 |
+
- **3rd knot**: final wrap to secure the stitch
|
| 94 |
+
3. **Phase: Cutting** – trim excess suture
|
| 95 |
+
- **Cutting**: use scissors to sever both ends of the suture (can cut both at once or sequentially)
|
| 96 |
+
|
| 97 |
+
2. **Second tying** (forming the second knot at a rotated position) (6개 stage가 되려면 120°, 180° tying을 분리해야 하는데 설명을 추가할지?)
|
| 98 |
+
- *Subtasks*: identical to First tying (Suturing → Knot tying → Cutting)
|
| 99 |
+
|
| 100 |
+
3. **Front side tying** (additional knot on the front face between first and second)
|
| 101 |
+
- *Subtasks*: identical to First tying
|
| 102 |
+
|
| 103 |
+
4. **Flip** (reorient vessel for back‐side access)
|
| 104 |
+
- **Phase: Flip** – flip the vessel clamp
|
| 105 |
+
- **Flip clamp**: reposition the clamp so that the vessel’s backside faces the camera
|
| 106 |
+
|
| 107 |
+
5. **Back side tying** (forming knots on the backside between first and second)
|
| 108 |
+
- *Subtasks*: identical to First tying
|
| 109 |
+
|
| 110 |
+
|
| 111 |
+
### Tool Segmentation Classes
|
| 112 |
+
|
| 113 |
+
The dataset contains the following classes:
|
| 114 |
+
|
| 115 |
+
| Class ID | Class Name |RGB Color |
|
| 116 |
+
|----------|------------------|----------------|
|
| 117 |
+
| 0 | forceps |(253, 0, 26) |
|
| 118 |
+
| 1 | scissors |(43, 253, 62) |
|
| 119 |
+
| 2 | vascular_clamps |(0, 43, 249) |
|
| 120 |
+
| 3 | needle_holder |(255, 253, 66) |
|
| 121 |
+
| 4 | vessel |(253, 40, 250) |
|
| 122 |
+
| 5 | needle |(38, 255, 254) |
|
| 123 |
+
| 6 | thread |(198, 161, 251) |
|
| 124 |
+
|
| 125 |
+
## Examples of Labeled Data
|
| 126 |
+
|
| 127 |
+
Figure 1, 2, 3 shows some examples of labeled data.
|
| 128 |
+
|
| 129 |
+
<figure>
|
| 130 |
+
<img src="./fig/ex1.jpg" width="1920" height="1072"/>
|
| 131 |
+
<figcaption>Figure 1: Example of Segmentation Mask of Image 1</figcaption>
|
| 132 |
+
</figure>
|
| 133 |
+
|
| 134 |
+
<figure>
|
| 135 |
+
<img src="./fig/ex2.jpg" width="1920" height="1072"/>
|
| 136 |
+
<figcaption>Figure 2: Example of Segmentation Mask of Image 2</figcaption>
|
| 137 |
+
</figure>
|
| 138 |
+
|
| 139 |
+
<figure>
|
| 140 |
+
<img src="./fig/ex3.jpg" width="1920" height="1072"/>
|
| 141 |
+
<figcaption>Figure 3: Example of Segmentation Mask of Image 3</figcaption>
|
| 142 |
+
</figure>
|
| 143 |
+
|
| 144 |
+
|
| 145 |
+
## Citation
|
| 146 |
+
|
| 147 |
+
```bibtex
|
| 148 |
+
@misc{
|
| 149 |
+
}
|