bmvc 2023 paper: Multi-Scale Cross Contrastive Learning for Semi-Supervised Medical Image Segmentation

Requirements

Some important required packages include:

• Pytorch version >=1.7.1
• TensorBoardX
• Python == 3.8
• Some basic python packages such as Numpy, Scikit-image, SimpleITK, Scipy ......

Follow official guidance to install Pytorch.

Prepare data

The datasets we used are provided by TransUnet's authors. [Get processed data in this link] (Synapse/BTCV: https://drive.google.com/drive/folders/1ACJEoTp-uqfFJ73qS3eUObQh52nGuzCd and ACDC: https://drive.google.com/drive/folders/1KQcrci7aKsYZi1hQoZ3T3QUtcy7b--n4).

Usage

1. Clone the repo:

git clone https://github.com/kathyliu579/MCSC.git
cd MCSC

2. Download the processed data and put the data in ../data/ACDC, please read and follow the README.

3. Train the model

cd code
python python train_MCSC_2D_256_28.py or bash/sh train_acdc_unet_semi_seg.sh

4. Test the model

python test_2D_fully.py or bash/sh test_acdc_unet_semi_seg.sh

References

• SwinUnet
• SSL4MIS

Citation

[@inproceedings{Liu_2023_BMVC, author = {Qianying Liu and Xiao Gu and Paul Henderson and Fani Deligianni}, title = {Multi-Scale Cross Contrastive Learning for Semi-Supervised Medical Image Segmentation}, booktitle = {34th British Machine Vision Conference 2023, {BMVC} 2023, Aberdeen, UK, November 20-24, 2023}, publisher = {BMVA}, year = {2023}, url = {https://bmvc2022.mpi-inf.mpg.de/BMVC2023/0868.pdf} }]