Code for paper MICCAI2022 paper "Evidence fusion with contextual discounting for multi-modality medical image segmentation".
We propose a new deep framework allowing us to merge multi-MRI image segmentation results using the formalism of Dempster-Shafer theory while taking into account the reliability of different modalities relative to different classes.
Environment requirement:
Before using the code, please install the required packages according to the instructions( refer to https://github.com/iWeisskohl/Evidential-neural-network-for-lymphoma-segmentation )
Models:
Copy the models from net into ./monai/networks/nets
Pre-Trained weights of ES module for flair, t1, t1Gd and t2 are located in ./model_single_modality
Training: ./medical-segmentation-master_enn_fusion
python TRAINING_unet_enn.py
###########Citing this paper #############
@inproceedings{huang2022evidence,
title={Evidence fusion with contextual discounting for multi-modality medical image segmentation},
author={Huang, Ling and Denoeux, Thierry and Vera, Pierre and Ruan, Su},
booktitle={International Conference on Medical Image Computing and Computer-Assisted Intervention},
pages={401--411},
year={2022},
organization={Springer}
}