Spaces:
No application file
title: Demo Cell Detection Gradio
emoji: 📈
colorFrom: indigo
colorTo: indigo
sdk: docker
pinned: false
license: gpl-3.0
metrics:
- accuracy
- map
tags:
- yolo
- cell detection
Model Card for Automatic Identification and Counting of Blood Cells
This model is designed for the automatic identification and counting of blood cells from smear images, using a machine learning approach.
Model Details
Model Description
- Developed by: Mahmudul Alam
- Shared by: Pablo Rocamora-Garcia
- Model type: Object Detection and Classification (YOLO)
- License: GNU General Public License v3.0
Model Sources
- Repository: https://github.com/MahmudulAlam/Automatic-Identification-and-Counting-of-Blood-Cells
- Paper: Machine learning approach of automatic identification and counting of blood cells
- Demo: (Demo link or more information needed)
Uses
Direct Use
This model is intended for use in medical diagnosis to evaluate overall health condition through the automatic identification and counting of red blood cells, white blood cells, and platelets from blood smear images.
Out-of-Scope Use
The model is not intended for use outside of the scope of medical imaging and blood cell analysis. Misuse or application in other domains may result in inaccurate or irrelevant results.
How to Get Started with the Model
To get started with the model, download the trained weights, set up the environment with TensorFlow and TF-slim, and run detect.py. Detailed instructions can be found in the repository.
Training Details
Training Data
The model was trained using the Complete Blood Count (CBC) Dataset, which contains images of blood smears for red blood cells, white blood cells, and platelets.
Results
Performance on Blood Cell Detection
The performance of different CNN architectures with the YOLO algorithm for detecting different types of blood cells is summarized in the following table:
| Model | RBC Accuracy (%) | WBC Accuracy (%) | Platelet Accuracy (%) | mAP | Execution Time (ms) |
|---|---|---|---|---|---|
| Tiny YOLO | 96.09 | 86.89 | 96.36 | 0.623 | 660 |
| VGG-16 | 72.98 | 100.00 | 90.91 | 0.713 | 3106 |
| ResNet50 | 79.80 | 95.08 | 87.27 | 0.743 | 3711 |
| InceptionV3 | 87.75 | 100.00 | 96.36 | 0.682 | 2630 |
| MobileNet | 74.24 | 93.44 | 83.64 | 0.520 | 784 |
Accuracy of Counting Blood Cells
The accuracy of counting Red Blood Cells (RBCs), White Blood Cells (WBCs), and platelets using the proposed method is detailed in the table below:
| Cell Type | Ground Truths | Estimated Count | Accuracy (%) |
|---|---|---|---|
| RBCs | 792 | 823 | 96.09 |
| WBCs | 65 | 53 | 86.89 |
| Platelets | 155 | 353 | 96.36 |
Citation
BibTeX:
@article{https://doi.org/10.1049/htl.2018.5098,
title={Machine learning approach of automatic identification and counting of blood cells},
author={Alam, Mohammad Mahmudul and Islam, Mohammad Tariqul},
journal={Healthcare Technology Letters},
volume={6},
number={4},
pages={103-108},
year={2019},
doi={https://doi.org/10.1049/htl.2018.5098},
url={https://ietresearch.onlinelibrary.wiley.com/doi/abs/10.1049/htl.2018.5098}
}