CancerCellVision is a deep learning project focused on detecting and classifying cancer cells from histopathology segmentation images. Leveraging Convolutional Neural Networks (CNNs) and a robust medical imaging dataset, this project aims to aid early cancer diagnosis and support treatment planning through automated and accurate analysis.
CancerCellVision applies computer vision and deep learning techniques to medical images to identify and distinguish between healthy and cancerous cells. By automating cell classification, it supports pathologists and reduces diagnostic errors.
- 🧬 Automated cancer cell detection from segmentation images
- 🔎 CNN-based classification architecture
- 📊 High accuracy with thorough training and validation
- 📁 Modular codebase for preprocessing, training, evaluation, and inference
- 📷 Support for histopathology datasets (e.g., PatchCamelyon, MoNuSeg, etc.)
- numpy
- pandas
- tensorflow
- scikit-learn
- pillow
- fastapi
- uvicorn
- matplotlib
CancerCellVision uses histopathology image datasets such as:
- PatchCamelyon (PCam)
- MoNuSeg
Instructions for downloading and placing datasets can be found indata/README.md.
Follow these steps to set up and run the project:
git clone https://github.com/AnojAryal/CancerCellVision && \
cd CancerCellVision && \python3 -m venv venv && \
source venv/bin/activate && \pip install --upgrade pip && \
pip install -r requirements.txt && \mkdir -p results && \
touch results/temp_results.txt && \python run/detecty.pyuvicorn services.main:app --reload --port 8001We welcome contributions to improve CancerCellVision!
Feel free to open a pull request from the detection_api branch to the develoipment branch if you're interested in contributing.
Bug fixes, feature enhancements, or suggestions are all appreciated.
Please follow the project structure, write clean, well-documented code, and ensure compatibility before submitting.