SSL4MIS

Semi-supervised-learning-for-medical-image-segmentation.

• [New] We have transferred to a new topic about active learning and source-free domain adaptation for medical image analysis, which may be closer to the real clinical requirement. The new benchmark is here.

• We are reformatting the codebase to support the 5-fold cross-validation and randomly select labeled cases, the reformatted methods in this Branch.

• Recently, semi-supervised image segmentation has become a hot topic in medical image computing, unfortunately, there are only a few open-source codes and datasets, since the privacy policy and others. For easy evaluation and fair comparison, we are trying to build a semi-supervised medical image segmentation benchmark to boost the semi-supervised learning research in the medical image computing community. If you are interested, you can push your implementations or ideas to this repo or contact me at any time.

• This repo has re-implemented these semi-supervised methods (with some modifications for semi-supervised medical image segmentation, more details please refer to these original works): (1) Mean Teacher; (2) Entropy Minimization; (3) Deep Adversarial Networks; (4) Uncertainty Aware Mean Teacher; (5) Interpolation Consistency Training; (6) Uncertainty Rectified Pyramid Consistency; (7) Cross Pseudo Supervision; (8) Cross Consistency Training; (9) Deep Co-Training; (10) Cross Teaching between CNN and Transformer; (11) FixMatch; (12) Regularized Dropout. In addition, several backbones networks (both 2D and 3D) are also supported in this repo, such as UNet, nnUNet, VNet, AttentionUNet, ENet, Swin-UNet, etc.

• This project was originally developed for our previous works. Now and future, we are still working on extending it to be more user-friendly and support more approaches to further boost and ease this topic research. If you use this codebase in your research, please cite the following works:

  复制

    @article{media2022urpc,
    title={Semi-Supervised Medical Image Segmentation via Uncertainty Rectified Pyramid Consistency},
    author={Luo, Xiangde and Wang, Guotai and Liao, Wenjun and Chen, Jieneng and Song, Tao and Chen, Yinan and Zhang, Shichuan, Dimitris N. Metaxas, and Zhang, Shaoting},
    journal={Medical Image Analysis},
    volume={80},
    pages={102517},
    year={2022},
    publisher={Elsevier}}
    
    @inproceedings{luo2021ctbct,
    title={Semi-supervised medical image segmentation via cross teaching between cnn and transformer},
    author={Luo, Xiangde and Hu, Minhao and Song, Tao and Wang, Guotai and Zhang, Shaoting},
    booktitle={International Conference on Medical Imaging with Deep Learning},
    pages={820--833},
    year={2022},
    organization={PMLR}}
  
    @InProceedings{luo2021urpc,
    author={Luo, Xiangde and Liao, Wenjun and Chen, Jieneng and Song, Tao and Chen, Yinan and Zhang, Shichuan and Chen, Nianyong and Wang, Guotai and Zhang, Shaoting},
    title={Efficient Semi-supervised Gross Target Volume of Nasopharyngeal Carcinoma Segmentation via Uncertainty Rectified Pyramid Consistency},
    booktitle={Medical Image Computing and Computer Assisted Intervention -- MICCAI 2021},
    year={2021},
    pages={318--329}}
     
    @InProceedings{luo2021dtc,
    title={Semi-supervised Medical Image Segmentation through Dual-task Consistency},
    author={Luo, Xiangde and Chen, Jieneng and Song, Tao and  Wang, Guotai},
    journal={AAAI Conference on Artificial Intelligence},
    year={2021},
    pages={8801-8809}}
    
    @misc{ssl4mis2020,
    title={{SSL4MIS}},
    author={Luo, Xiangde},
    howpublished={\\url{https://github.com/HiLab-git/SSL4MIS}},
    year={2020}}

Literature reviews of semi-supervised learning approach for medical image segmentation (SSL4MIS).