What Can Be Transferred: Unsupervised Domain Adaptation for Endoscopic Lesions Segmentation

TL;DR

This paper introduces a novel unsupervised domain adaptation model for endoscopic lesion segmentation, effectively transferring knowledge across datasets with diverse lesion appearances by focusing on transferable features.

Contribution

The paper proposes a new model with two modules, T_D and T_F, that selectively transfer domain-invariant features, addressing negative transfer issues in medical image segmentation.

Findings

Outperforms existing methods on medical endoscopic datasets

Effectively handles diverse lesion appearances across datasets

Improves segmentation accuracy through selective feature transfer

Abstract

Unsupervised domain adaptation has attracted growing research attention on semantic segmentation. However, 1) most existing models cannot be directly applied into lesions transfer of medical images, due to the diverse appearances of same lesion among different datasets; 2) equal attention has been paid into all semantic representations instead of neglecting irrelevant knowledge, which leads to negative transfer of untransferable knowledge. To address these challenges, we develop a new unsupervised semantic transfer model including two complementary modules (i.e., T_D and T_F ) for endoscopic lesions segmentation, which can alternatively determine where and how to explore transferable domain-invariant knowledge between labeled source lesions dataset (e.g., gastroscope) and unlabeled target diseases dataset (e.g., enteroscopy). Specifically, T_D focuses on where to translate transferable visual information of medical lesions via residual transferability-aware bottleneck, while neglecting untransferable visual characterizations. Furthermore, T_F highlights how to augment transferable semantic features of various lesions and automatically ignore untransferable representations, which explores domain-invariant knowledge and in return improves the performance of T_D. To the end, theoretical analysis and extensive experiments on medical endoscopic dataset and several non-medical public datasets well demonstrate the superiority of our proposed model.