Graph Cuts Loss to Boost Model Accuracy and Generalizability for Medical Image Segmentation

Pytorch implementation for our ICCVW paper 'Graph Cuts Loss to Boost Model Accuracy and Generalizability for Medical Image Segmentation'.

Acknowledgement

Implementations of the losses cited in our work are public avaliable.
clDice - a Novel Topology-Preserving Loss Function for Tubular Structure Segmentation (clDice)
An Elastic Interaction-Based Loss Function for Medical Image Segmentation (EIB)
Learning Active Contour Models for Medical Image Segmentation (AC)
Learning Euler's Elastica Model for Medical Image Segmentation (ACE)
Reducing the Hausdorff Distance in Medical Image Segmentation with Convolutional Neural Networks (HD)
Boundary loss for highly unbalanced segmentation (BD)

Note

Contact: Zhou Zheng (zzheng@mori.m.is.nagoya-u.ac.jp)