TC-CoNet

This repository contains the supported pytorch code and configuration files to reproduce of TC-CoNet.

Parts of codes are borrowed from nn-UNet. For detailed configuration of the dataset, please refer to nn-UNet.

Environment

Please prepare an environment with Python 3.7, Pytorch 1.7.1, and Windows 10.

Dataset Preparation

Datasets can be acquired via following links:

Dataset I ACDC

Dataset II The Synapse multi-organ CT dataset

Dataset III Brain_tumor

Preprocess Data

• TCCoNet_convert_decathlon_task -i D:\Codes\Medical_image\UploadGitHub\TCCoNet\DATASET\TCCoNet_raw\TCCoNet_raw_data
• TCCoNet_plan_and_preprocess -t 2

Functions of scripts

• Network architecture:
  • TCCoNet\TCCoNet\network_architecture\TCCoNet_acdc.py
  • TCCoNet\TCCoNet\network_architecture\TCCoNet_synapse.py
  • TCCoNet\TCCoNet\network_architecture\TCCoNet_tumor.py
  • TCCoNet\TCCoNet\network_architecture\TCCoNet_heart.py
  • TCCoNet\TCCoNet\network_architecture\TCCoNet_lung.py
• Trainer for dataset:
  • TCCoNet\TCCoNet\training\network_training\TCCoNetTrainerV2_TCCoNet_acdc.py
  • TCCoNet\TCCoNet\training\network_training\TCCoNetTrainerV2_TCCoNet_synapse.py
  • TCCoNet\TCCoNet\training\network_training\TCCoNetTrainerV2_TCCoNet_tumor.py
  • TCCoNet\TCCoNet\training\network_training\TCCoNetTrainerV2_TCCoNet_heart.py
  • TCCoNet\TCCoNet\training\network_training\TCCoNetTrainerV2_TCCoNet_lung.py

Train Model

• python run_training.py 3d_fullres TCCoNetTrainerV2_TCCoNet_synapse 2 0

Test Model

• python predict.py -i D:\Codes\Medical_image\UploadGitHub\TCCoNet\DATASET\TCCoNet_raw\TCCoNet_raw_data\Task002_Synapse\imagesTs -o D:\Codes\Medical_image\UploadGitHub\TCCoNet\DATASET\TCCoNet_raw\TCCoNet_raw_data\Task002_Synapse\imagesTs_infer -m D:\Codes\Medical_image\UploadGitHub\TCCoNet\DATASET\TCCoNet_trained_models\TCCoNet\3d_fullres\Task002_Synapse\TCCoNetTrainerV2_TCCoNet_synapse__TCCoNetPlansv2.1 -f 0

• python TCCoNet/inference_synapse.py

Acknowledgements

This repository makes liberal use of code from Swin Transformer, nnUNet.

Citation

@article{chen2023collaborative,
title={Collaborative networks of transformers and convolutional neural networks are powerful and versatile learners for accurate 3D medical image segmentation},
author={Chen, Yong and Lu, Xuesong and Xie, Qinlan},
journal={Computers in Biology and Medicine},
volume={164},
pages={107228},
year={2023},
publisher={Elsevier}
}