MedSeg-DFormer

A PyTorch implementation of D-Former: A U-shaped Dilated Transformer for 3D Medical Image Segmentation.

This repository hosts the codebase designed to enable high-quality segmentation of 3D medical images using a novel dilated transformer approach.

Features

• U-shaped architecture with dilated transformer blocks
• Optimized for medical image segmentation
• Comprehensive documentation and modular code structure

Usage

Clone the repository and follow the instructions in the implementation tutorials to get started:

# Clone the repository
git clone https://github.com/transmutat/MedSeg-DFormer.git
cd MedSeg-DFormer

# Install dependencies
pip install -r requirements.txt

# Run the training
python train.py

For detailed guidance, please review /docs.

Citation

If you use this work in your research, kindly cite it:

@article{dformer,
  title={D-former: A u-shaped dilated transformer for 3d medical image segmentation},
  author={Wu, Yixuan and Liao, Kuanlun and Chen, Jintai and Wang, Jinhong and Chen, Danny Z and Gao, Honghao and Wu, Jian},
  journal={Neural Computing and Applications},
  pages={1--14},
  year={2022},
  publisher={Springer}
}