Blood Cancer Detection Using AI

Project Description

This project aims to detect blood cancer using AI techniques. It leverages deep learning models to analyze blood sample images and predict the presence of cancerous cells.

Installation

To set up the project, follow these steps:

1. Clone the repository:

   git clone https://github.com/Enosha01/Blood-Cancer-Detection-Using-AI.git
   cd blood-cancer-detection

2. Create a virtual environment:

   python -m venv env

3. Activate the virtual environment:

   • On Windows:

     .\env\Scripts\activate

   • On macOS/Linux:

     source env/bin/activate

4. Install the required packages:

   pip install -r requirements.txt

Usage

To run the application, use the following command:

python app.py

Project Structure

BLOOD CANCER DETECTION USING AI/
│
├── app.py
├── CODE.ipynb
├── dataset.txt
├── model.ipynb
├── sql.sql
├── vs blood cancer.ipynb
├── env/
│   ├── pyvenv.cfg
│   ├── Include/
│   ├── Lib/
│   └── Scripts/
├── images/
│   ├── Snap_057.jpg
│   ├── Snap_091.jpg
│   └── ...
├── models/
│   ├── CNN.h5
│   ├── FinalModel.h5
│   └── mobilenet.h5
├── static/
│   ├── css/
│   ├── fonts/
│   ├── img/
│   └── js/
├── templates/
│   ├── about.html
│   ├── blog-post.html
│   ├── index.html
│   ├── login.html
│   ├── loginhomepage.html
│   ├── README.md
│   └── Register.html
└── README.md

Dataset

The dataset used for this project consists of blood sample images. The images are stored in the images/ directory.

Model Training

To train the model, use the model.ipynb notebook. It contains the code for data preprocessing, model architecture, training, and evaluation.

Model Evaluation

The trained models are saved in the models/ directory. You can evaluate the models using the model.ipynb notebook.

Deployment

To deploy the application, use the app.py script. It sets up a web server to serve the model predictions.

Contributing

Contributions are welcome! Please open an issue or submit a pull request.

License

This project is licensed under the MIT License..

Contact

For any questions or inquiries, please contact [email protected].