Cardiac Segmentation ACDC

Deep learning-based cardiac segmentation using PyTorch, MONAI, and U-Net models.

About

This project leverages deep learning models for cardiac segmentation, particularly for the ACDC dataset. The goal is to segment the heart's anatomical structures to aid in medical analysis.

Features:

• 2D and 3D U-Net models for segmentation.
• Attention mechanisms for improved segmentation accuracy.
• Support for both 2D and 3D datasets (using MONAI and PyTorch).

Installation

Prerequisites:

• Python 3.x
• PyTorch
• MONAI
• other Python dependencies from requirements.txt

Setup:

1. Clone the repository:

   git clone https://github.com/arkanandi/Cardiac_Segmentation_ACDC.git
   cd Cardiac_Segmentation_ACDC

2. Install the required Python libraries:

   pip install -r requirements.txt

3. Download the ACDC dataset and place it in the appropriate folder (see the dataset instructions in the repository for details).

4. To train the model, run:

   python train_2d.py
   # or
   python train_3d.py

5. To make predictions:

   python predict_2d.py
   # or
   python predict_3d.py

Usage

Once you have trained the model, you can use it to predict cardiac structures on new datasets.

Contributing

Feel free to fork the repository and submit pull requests. Issues and suggestions are always welcome.

1. Fork this repository.
2. Clone your fork:

   git clone https://github.com/your-username/Cardiac_Segmentation_ACDC.git

3. Create a new branch:

   git checkout -b feature-name

4. Make changes and commit:

   git commit -am 'Add new feature'

5. Push to your fork:

   git push origin feature-name

License

This project is licensed under the MIT License - see the LICENSE file for details.