Local-Global Pseudo-label Correction for Source-free Domain Adaptive Medical Image Segmentation

TL;DR

This paper introduces LGDA, a novel source-free domain adaptation method for medical image segmentation that corrects false pseudo-labels using local context and global class prototypes, improving accuracy without source data.

Contribution

The paper proposes a new local-global pseudo-label correction approach for source-free domain adaptation in medical imaging, addressing false labels and enhancing segmentation performance.

Findings

Outperforms state-of-the-art methods on fundus image datasets

Effective correction of pseudo-label errors improves segmentation accuracy

Operates without using source data, preserving privacy

Abstract

Domain shift is a commonly encountered issue in medical imaging solutions, primarily caused by variations in imaging devices and data sources. To mitigate this problem, unsupervised domain adaptation techniques have been employed. However, concerns regarding patient privacy and potential degradation of image quality have led to an increased focus on source-free domain adaptation. In this study, we address the issue of false labels in self-training based source-free domain adaptive medical image segmentation methods. To correct erroneous pseudo-labels, we propose a novel approach called the local-global pseudo-label correction (LGDA) method for source-free domain adaptive medical image segmentation. Our method consists of two components: An offline local context-based pseudo-label correction method that utilizes local context similarity in image space. And an online global pseudo-label correction method based on class prototypes, which corrects erroneously predicted pseudo-labels by considering the relative distance between pixel-wise feature vectors and prototype vectors. We evaluate the performance of our method on three benchmark fundus image datasets for optic disc and cup segmentation. Our method achieves superior performance compared to the state-of-the-art approaches, even without using of any source data.