Proto_Contra_SFDA

Source-Free Domain Adaptation for Medical Image Segmentation via Prototype-Anchored Feature Alignment and Contrastive Learning

This is an official implementation of MICCAI 2023 paper [Source-Free Domain Adaptation for Medical Image Segmentation via Prototype-Anchored Feature Alignment and Contrastive Learning.] (https://arxiv.org/abs/2307.09769)

Requirements

• Linux with Python ≥ 3.7
• PyTorch ≥ 1.7.1 and torchvision matches the PyTorch insallation. Install them following the official instructions from pytorch.org to make sure of this.

Quick start

1. Clone repository.

git clone https://github.com/CSCYQJ/MICCAI23-ProtoContra-SFDA
cd MICCAI23-ProtoContra-SFDA

2. Download Data.

Both datasets are public available. CT modality data is from MICCAI 2015 Multi-Atlas Abdomen Labeling Challenge. MRI modality data is from 2019 CHAOS Challenge.

3. Source Model Training.

python main_trainer_source.py --config_file configs/train_source_seg.yaml --gpu_id 0

4. Target Domain Adaptaion PFA Stage.

python main_trainer_sfda.py --config_file configs/train_target_adapt_PFA.yaml --gpu_id 0

5. Target Domain Adaptaion CL Stage.

python main_trainer_sfda.py --config_file configs/train_target_adapt_CL.yaml --gpu_id 0

Citation

If you find this work or code is helpful in your research, please cite:

@article{yu2023source,
  title={Source-Free Domain Adaptation for Medical Image Segmentation via Prototype-Anchored Feature Alignment and Contrastive Learning},
  author={Yu, Qinji and Xi, Nan and Yuan, Junsong and Zhou, Ziyu and Dang, Kang and Ding, Xiaowei},
  journal={arXiv preprint arXiv:2307.09769},
  year={2023}
}

Acknowledgement

Many thanks to these excellent opensource projects

• PCT
• U2PL