This repository provides the necessary tools to classify CT scans using a deep learning model designed to identify COVID-19 cases. The project is intended for researchers, developers, and anyone interested in AI applications in healthcare.
- 🔍 Project Overview
- 🚀 Getting Started
- 🌐 Web Application
- 📊 MLflow and DagsHub Integration
- 🔮 Future Work
- 🤝 Contributing
- 📜 License
- 🙏 Acknowledgments
The objective of this project is to develop a machine learning model capable of classifying CT scans into three categories: COVID-19, CAP, and non-COVID. TensorFlow and Keras are used for model development, with MLflow integrated for experiment tracking and performance analysis.
Before starting, ensure that Python 3.8 or higher is installed. Required Python packages are listed in the requirements.txt file.
-
Clone the repository:
git clone https://github.com/Jesteban247/COVID-19-_CT_Scan_Classification.git cd COVID-19-_CT_Scan_Classification -
Install the required packages:
pip install -r requirements.txt
To train and evaluate the model, refer to the Jupyter notebooks (.ipynb files) provided in the repository. These notebooks include detailed instructions for data preprocessing, model training, and performance evaluation. Open and execute the notebooks in environments like Jupyter Lab or Google Colab.
-
Install all dependencies by following the instructions in the
requirements.txtfile. -
Run the application using the command:
python main.py
-
Access the application through a web browser to start making predictions.
The web application offers an easy-to-use interface for uploading CT scans and obtaining predictions. It includes a scrollable gallery to display images and prediction results.
Screen.Recording.2024-08-23.at.11.03.12.PM.mov
This project integrates with MLflow and DagsHub to track and log experiment data. Metrics such as accuracy, loss, precision, recall, and F1-score are recorded for comprehensive performance analysis.
Screen.Recording.2024-08-23.at.11.05.49.PM.mov
- 📈 Expand the dataset to improve model generalization.
- 🛠️ Further fine-tune the model for enhanced accuracy.
- 🔄 Implement live data prediction capabilities.
Contributions are welcome. Interested individuals can fork the repository and submit a pull request for review.
This project is licensed under the MIT License. For more information, refer to the LICENSE file.
Acknowledgment is given to the contributors and the open-source community for their support and resources, which have been invaluable to the project's development.