GraphCL: Graph-based Clustering for Semi-Supervised Medical Image Segmentation

Introduction

Official code for "GraphCL: Graph-based Clustering for Semi-Supervised Medical Image Segmentation".

Requirements

This repository is based on PyTorch 1.8.0, CUDA 11.1 and Python 3.6.13. Experiments on Pancreas-NIH and ACDC datasets in our paper were conducted on NVIDIA GeForce RTX 3090 GPU with an identical experimental setting. LA Dataset experiments run on an NVIDIA A800 GPU.

News

We provide NIH-Pancreas dataset codes code/pancreas, data split (and other information) could be got at CoraNet

Usage

We provide code, data_split and models for LA and ACDC dataset.

Data could be got at LA and ACDC.

To train a model,

python ./code/LA_GraphCL_train.py  #for LA training
python ./code/ACDC_GraphCL_train.py  #for ACDC training

To test a model,

python ./code/test_LA.py  #for LA testing
python ./code/test_ACDC.py  #for ACDC testing

Acknowledgements

Our code is largely based on BCP and SS-Net. Thanks for these authors for their valuable work, hope our work can also contribute to related research.