LT-Net: Label Transfer by Learning Reversible Voxel-wise Correspondence for One-shot Medical Image Segmentation

TL;DR

This paper presents LT-Net, a one-shot medical image segmentation approach that learns reversible voxel-wise correspondences using cycle consistency, reducing manual annotation efforts and improving segmentation accuracy.

Contribution

The paper introduces a novel cycle-consistency-based learning framework for one-shot segmentation that effectively learns reversible voxel correspondences, enhancing stability and performance.

Findings

Outperforms existing deep learning one-shot segmentation methods.

Achieves higher accuracy than classical multi-atlas segmentation.

Demonstrates robustness through extensive experiments.

Abstract

We introduce a one-shot segmentation method to alleviate the burden of manual annotation for medical images. The main idea is to treat one-shot segmentation as a classical atlas-based segmentation problem, where voxel-wise correspondence from the atlas to the unlabelled data is learned. Subsequently, segmentation label of the atlas can be transferred to the unlabelled data with the learned correspondence. However, since ground truth correspondence between images is usually unavailable, the learning system must be well-supervised to avoid mode collapse and convergence failure. To overcome this difficulty, we resort to the forward-backward consistency, which is widely used in correspondence problems, and additionally learn the backward correspondences from the warped atlases back to the original atlas. This cycle-correspondence learning design enables a variety of extra, cycle-consistency-based supervision signals to make the training process stable, while also boost the performance. We demonstrate the superiority of our method over both deep learning-based one-shot segmentation methods and a classical multi-atlas segmentation method via thorough experiments.