Source-free unsupervised domain adaptation for cross-modality abdominal multi-organ segmentation

TL;DR

This paper introduces a novel source-free unsupervised domain adaptation framework for cross-modality abdominal multi-organ segmentation, effectively transferring knowledge from labeled CT to unlabeled MR data without source dataset access.

Contribution

It proposes a two-stage adaptation method combining feature statistics, entropy minimization, style compensation, and circular learning, advancing source-free domain adaptation techniques.

Findings

Outperforms existing state-of-the-art methods in abdominal multi-organ segmentation.

Achieves supervised-level performance with only one labeled MR volume.

Proves effective for reverse domain adaptation without source data.

Abstract

Domain adaptation is crucial for transferring the knowledge from the source labeled CT dataset to the target unlabeled MR dataset in abdominal multi-organ segmentation. Meanwhile, it is highly desirable to avoid the high annotation cost related to the target dataset and protect the source dataset privacy. Therefore, we propose an effective source-free unsupervised domain adaptation method for cross-modality abdominal multi-organ segmentation without source dataset access. The proposed framework comprises two stages. In the first stage, the feature map statistics-guided model adaptation combined with entropy minimization is developed to help the top segmentation network reliably segment the target images. The pseudo-labels output from the top segmentation network are used to guide the style compensation network to generate source-like images. The pseudo-labels output from the middle segmentation network is used to supervise the learning progress of the desired model (bottom segmentation network). In the second stage, the circular learning and pixel-adaptive mask refinement are used to further improve the desired model performance. With this approach, we achieved satisfactory abdominal multi-organ segmentation performance, outperforming the existing state-of-the-art domain adaptation methods. The proposed approach can be easily extended to situations in which target annotation data exist. With only one labeled MR volume, the performance can be levelled with that of supervised learning. Furthermore, the proposed approach is proven to be effective for source-free unsupervised domain adaptation in reverse direction.