This is the official code of Spatial and Planar Consistency for Semi-Supervised Volumetric Medical Image Segmentation (BMVC 2023).
Comparison with state-of-the-art models on LA and P-CT test set. Red and bold indicate the best and second best performance.
Qualitative results on LA (left) and P-CT (right). (a) MT. (b) DTC. (c) MC-Net. (d) MC-Net+. (e) SPC. (f) Ground truth. The green arrows highlight the difference among of the results.albumentations==0.5.2
mayavi==4.8.1
MedPy==0.4.0
numpy==1.18.5
opencv_python==4.2.0.32
Pillow==9.5.0
scikit_image==0.19.1
scikit_learn==1.2.2
scipy==1.4.1
SimpleITK==2.2.1
skimage==0.0
torch==1.8.0
torchio==0.18.53
torchvision==0.9.0
visdom==0.1.8.9
Data preparation Your datasets directory tree should be look like this:
dataset
├── train_sup_20
├── image
├── 1.tif
├── 2.tif
└── ...
└── mask
├── 1.tif
├── 2.tif
└── ...
├── train_unsup_80
├── image
├── val
├── image
└── mask
Training
python -m torch.distributed.launch --nproc_per_node=4 train_semi_SPC.py
Testing
python -m torch.distributed.launch --nproc_per_node=4 test_SPC.py
If our work is useful for your research, please cite our paper:
@inproceedings{Zhou_2023_BMVC,
author = {Yanfeng Zhou and yiming huang and Ge Yang},
title = {Spatial and Planar Consistency for Semi-Supervised Volumetric Medical Image Segmentation},
booktitle = {34th British Machine Vision Conference 2023, {BMVC} 2023, Aberdeen, UK, November 20-24, 2023},
publisher = {BMVA},
year = {2023},
url = {https://papers.bmvc2023.org/0084.pdf}
}