Code for this paper: Semi-supervised Medical Image Segmentation through Dual-task Consistency (DTC)
- More details and comparison methods will be released if the paper is accepted.
- The multi-classes DTC is under doing, and also will be released as we finished it.
Some important required packages include:
- Pytorch version >=0.4.1.
- TensorBoardX
- Python == 3.6
- Some basic python packages such as Numpy, Scikit-image, SimpleITK, Scipy ......
Follow official guidance to install Pytorch.
- Clone the repo:
git clone https://github.com/HiLab-git/DTC.git
cd DTC
-
Put the data in data/2018LA_Seg_Training Set.
-
Train the model
cd code
python train_la_dtc.py
- Test the model
python test_LA.py
Our best model is saved in the model dir DTC_model, and the pretrained SASSNet and UAMT model can be download from SASSNet_model and UA-MT_model. Our implemented 3D version of CCT (with main decoder and three auxiliary decoders) will be updated as soon as possible, and the other comparison method can be found in SSL4MIS
- The training set consists of 16 labeled scans and 64 unlabeled scans and the testing set includes 20 scans.
| Methods | DICE (%) | Jaccard (%) | ASD (voxel) | 95HD (voxel) | Reference | Released Date |
|---|---|---|---|---|---|---|
| UAMT | 88.88 | 80.21 | 2.26 | 7.32 | MICCAI2019 | 2019-10 |
| SASSNet | 89.54 | 81.24 | 2.20 | 8.24 | MICCAI2020 | 2020-07 |
| Original DTC | 89.42 | 80.98 | 2.10 | 7.32 | Arxiv | 2020-09 |
| LG-ER-MT | 89.62 | 81.31 | 2.06 | 7.16 | MICCAI2020 | 2020-10 |
| DUWM | 89.65 | 81.35 | 2.03 | 7.04 | MICCAI2020 | 2020-10 |
| Updated DTC | 89.85 | 81.72 | 1.81 | 7.03 | This repo | 2020-10 |
If you find this repository is useful in your research, please consider to cite:
@article{luo2020semi,
title={Semi-supervised Medical Image Segmentation through Dual-task Consistency},
author={Luo, Xiangde and Chen, Jieneng and Song, Tao and Chen, Yinan and Wang, Guotai and Zhang, Shaoting},
journal={arXiv preprint arXiv:2009.04448},
year={2020}
}
- This code is adapted from UA-MT, SASSNet, SegWithDistMap.
- We thank Dr. Lequan Yu, M.S. Shuailin Li and Dr. Jun Ma for their elegant and efficient code base.
- More semi-supervised learning approaches for medical image segmentation have summarized in this repository SSL4MIS.