This repo is the PyTorch implementation of our paper:
The code was developed using Python 3.8 with PyTorch 1.11.0, and CUDA 11.3. All experiments in our paper were conducted on a single NVIDIA A40 GPU with 48 GB of memory.
Install the main packages:
pytorch == 1.11.0
torchvision == 0.12.0
cudatoolkit == 11.3.1
The original datasets can be downloaded from the following links:
PS: Please cite the original dataset papers when using these datasets in your publications.
Follow the list files provided in the data folders to split the datasets.
python train_rematch.py
python eval.py
If you find this project useful, please consider citing our paper:
@article{shen2026segmentation,
title = {Segmentation-synthesis co-training for semi-supervised domain generalizable medical image segmentation},
author = {Zhiqiang Shen and Qingshan Hou and Peng Cao and Jinzhu Yang and Huazhu Fu and Osmar R. Zaiane and Zhaolin Chen},
journal = {Artificial Intelligence in Medicine},
volume = {181},
pages = {103512},
year = {2026},
issn = {0933-3657},
doi = {https://doi.org/10.1016/j.artmed.2026.103512}
If you have any questions or suggestions, please feel free to contact me (xxszqyy@gmail.com).