Deep Mutual Learning among Partially Labeled Datasets for Multi-Organ Segmentation

TL;DR

This paper introduces a two-stage mutual learning approach for multi-organ segmentation that effectively utilizes partially labeled datasets to improve segmentation accuracy across various organs and datasets.

Contribution

It proposes a novel two-stage mutual learning framework that leverages partial labels and pseudo labels to enhance multi-organ segmentation performance.

Findings

Achieved state-of-the-art results on nine diverse datasets.

Demonstrated the effectiveness of mutual difference and similarity learning.

Validated the approach through comprehensive ablation studies.

Abstract

The task of labeling multiple organs for segmentation is a complex and time-consuming process, resulting in a scarcity of comprehensively labeled multi-organ datasets while the emergence of numerous partially labeled datasets. Current methods are inadequate in effectively utilizing the supervised information available from these datasets, thereby impeding the progress in improving the segmentation accuracy. This paper proposes a two-stage multi-organ segmentation method based on mutual learning, aiming to improve multi-organ segmentation performance by complementing information among partially labeled datasets. In the first stage, each partial-organ segmentation model utilizes the non-overlapping organ labels from different datasets and the distinct organ features extracted by different models, introducing additional mutual difference learning to generate higher quality pseudo labels for unlabeled organs. In the second stage, each full-organ segmentation model is supervised by fully labeled datasets with pseudo labels and leverages true labels from other datasets, while dynamically sharing accurate features across different models, introducing additional mutual similarity learning to enhance multi-organ segmentation performance. Extensive experiments were conducted on nine datasets that included the head and neck, chest, abdomen, and pelvis. The results indicate that our method has achieved SOTA performance in segmentation tasks that rely on partial labels, and the ablation studies have thoroughly confirmed the efficacy of the mutual learning mechanism.