Semi-supervised Cardiac Image Segmentation via Label Propagation and Style Transfer

TL;DR

This paper introduces a semi-supervised cardiac MRI segmentation method that leverages label propagation and style transfer to improve robustness across different data sources, achieving high performance in a competitive challenge.

Contribution

The work presents a novel combination of semi-supervised learning with label propagation and style transfer for cardiac MRI segmentation, addressing data variance and annotation scarcity.

Findings

Ranked 2nd in the M&Ms challenge 7

Effectively leverages unlabelled data for segmentation

Reduces data variance across centers and vendors

Abstract

Accurate segmentation of cardiac structures can assist doctors to diagnose diseases, and to improve treatment planning, which is highly demanded in the clinical practice. However, the shortage of annotation and the variance of the data among different vendors and medical centers restrict the performance of advanced deep learning methods. In this work, we present a fully automatic method to segment cardiac structures including the left (LV) and right ventricle (RV) blood pools, as well as for the left ventricular myocardium (MYO) in MRI volumes. Specifically, we design a semi-supervised learning method to leverage unlabelled MRI sequence timeframes by label propagation. Then we exploit style transfer to reduce the variance among different centers and vendors for more robust cardiac image segmentation. We evaluate our method in the M&Ms challenge 7 , ranking 2nd place among 14 competitive teams.