🛎 Citation

If you find our work helpful for your research, please cite:

@article{zheng2024lagrange,
  title={Lagrange Duality and Compound Multi-Attention Transformer for Semi-Supervised Medical Image Segmentation},
  author={Zheng, Fuchen and Li, Quanjun and Li, Weixuan and Chen, Xuhang and Dong, Yihang and Huang, Guoheng and Pun, Chi-Man and Zhou, Shoujun},
  journal={arXiv preprint arXiv:2409.07793},
  year={2024}
}

📋Lagrange-Duality-and-CMAformer

Lagrange Duality and Compound Multi-Attention Transformer (CMAformer) for Semi-Supervised Medical Image Segmentation

Fuchen Zheng, Quanjun Li, Weixuan Li, Xuhang Chen, Yihang Dong, Guoheng Huang, Chi-Man Pun 📮and Shoujun Zhou 📮( 📮 Corresponding authors)

University of Macau, SIAT CAS, Guangdong University of Technology

🚧 Installation

Requirements: Ubuntu 20.04

1. Create a virtual environment: conda create -n your_environment python=3.8 -y and conda activate your_environment
2. Install Pytorch :pip install torch==2.3.0 torchvision==0.18.0 torchaudio==2.3.0 --index-url https://download.pytorch.org/whl/cu118 Or you can use Tsinghua Source for installation

pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple
pip install torch==2.0.0+cu118 torchvision==0.15.1+cu118 torchaudio==2.0.1+cu118 -f https://download.pytorch.org/whl/torch_stable.html

3. pip install tqdm scikit-learn albumentations==1.0.3 pandas einops axial_attention

4. python train_lits2017_png.py

1. Prepare the dataset

LiTS2017 datasets

• The LiTS2017 datasets can be downloaded here: {LiTS2017}.

• The Synapse dataset can be downloaded here: {Synapse multi-organ}.

• After downloading the datasets, you should run ./data_prepare/preprocess_lits2017_png.py to convert .nii files into .png files for training. (Save the downloaded LiTS2017 datasets in the data folder in the following format.)

• './data_prepare/'

  • preprocess_lits2017_png.py

• './data/'

  • LITS2017
    • ct
      • .nii
    • label
      • .nii
  • trainImage_lits2017_png
    • .png
  • trainImage_lits2017_png
    • .png

Other datasets

• Other datasets just similar to LiTS2017

2. Prepare the pre_trained weights

• The weights of the pre-trained CMAformer could be downloaded here. After that, the pre-trained weights should be stored in './pretrained_weights/'.
• To use pre-trained file, you should change 2 places in './train/train_lits2017_png.py'
  • 1. Change 'default=True' in 'parser.add_argument('--pretrained', default=False, type=str2bool)'
  • 2. Change 'pretrained_path= "./your_pretrained_file_path"' after 'if args.model_name == 'CMAformer':'

3. Train the CMAformer

cd ./train/
python train_lits2017_png.py 

4. Obtain the outputs

• After trianing, you could obtain the results in './trained_models/LiTS_CMAformer/'

🧧 Acknowledgement

This work was supported in part by the National Key R&D Project of China (2018YFA0704102, 2018YFA0704104), in part by Natural Science Foundation of Guangdong Province (No. 2023A1515010673), and in part by Shenzhen Technology Innovation Commission (No. JSGG20220831110400001), in part by Shenzhen Development and Reform Commission (No. XMHT20220104009), in part by the Science and Technology Development Fund, Macau SAR, under Grant 0141/2023/RIA2 and 0193/2023/RIA3.