File size: 2,831 Bytes
8f2f0a6
 
 
 
 
 
 
 
 
 
 
 
 
 
960f6b3
 
 
 
8f2f0a6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
---
tags:
- image-classification
- microbiology
- gram-stain
- medical-imaging
- research-use-only
- urine
- blood-culture
pipeline_tag: image-classification
---

# BiTTE

<p align="center">
  <img src="./icon.jpg" alt="BiTTE icon" width="220">
</p>

BiTTE is an application within the CarbConnect platform. It is designed for the simple and efficient classification of microorganisms from Gram-stained microscopy images.

The app classifies findings into seven primary groups:

1. Gram-negative rods (GNR)
2. Gram-negative cocci (GNC)
3. Gram-positive rods (GPR)
4. Gram-positive cocci (GPC)
5. Yeast-like fungi
6. No bacteria
7. Multiple bacteria

BiTTE also supports more detailed subcategories beyond these primary output groups.

## Intended Uses and Limitations

BiTTE is strictly intended for **Research Use Only (RUO)**.

It is **not** intended for:

- clinical diagnostics
- medical procedures
- patient management
- therapeutic selection
- any regulated clinical use

For further details, please refer to the BiTTE Learn More page on CarbConnect.

## How to Use

A video tutorial demonstrating how to use the app is available on YouTube.

## Training Data

The model was trained on a dataset of Gram-stained images of urine and blood culture specimens generously provided by:

- the School of Medicine, Kobe University
- the National Center for Global Health and Medicine (NCGM)

Specimens were Gram-stained using either the Favor or Barmy method.

Image acquisition was performed by photographing specimens through the eyepiece of an optical microscope at **1000x magnification** using a smartphone camera.

The dataset captures frequently encountered clinical bacterial species and includes:

- 15 species in urine specimens
- 19 species in aerobic blood culture specimens
- 13 species in anaerobic blood culture specimens

## Performance and Evidence

Related publication:

Kei Yamamoto, Goh Ohji, et al. *Accuracy of classification of urinary Gram-stain findings by a computer-aided diagnosis app compared with microbiology specialists*. J Med Microbiol. 2025 Apr;74(4):002008. doi: 10.1099/jmm.0.002008.

Paper link:

https://www.microbiologyresearch.org/content/journal/jmm/10.1099/jmm.0.002008

## Citation

If you use BiTTE in research, please cite:

```bibtex
@article{yamamoto2025bitte,
  author = {Yamamoto, Kei and Ohji, Goh and others},
  title = {Accuracy of classification of urinary Gram-stain findings by a computer-aided diagnosis app compared with microbiology specialists},
  journal = {Journal of Medical Microbiology},
  year = {2025},
  month = {Apr},
  volume = {74},
  number = {4},
  pages = {002008},
  doi = {10.1099/jmm.0.002008}
}
```

## Other Remarks

For guidance on achieving high-quality Gram staining, please refer to the automated gram stainer **Point of Care Gram Stainer (PoCGS)**.