BTMuda: A Bi-level Multi-source unsupervised domain adaptation framework for breast cancer diagnosis

TL;DR

This paper introduces BTMuda, a novel bi-level multi-source unsupervised domain adaptation framework that improves breast cancer diagnosis by effectively handling domain shifts and leveraging multiple data sources.

Contribution

The paper proposes a new bi-level multi-source UDA method with a three-branch architecture, combining CNN and Transformer, for better domain-invariant feature learning in breast cancer diagnosis.

Findings

Outperforms state-of-the-art methods on three public datasets

Effectively reduces intra- and inter-domain shifts

Enhances feature robustness with CNN-Transformer integration

Abstract

Deep learning has revolutionized the early detection of breast cancer, resulting in a significant decrease in mortality rates. However, difficulties in obtaining annotations and huge variations in distribution between training sets and real scenes have limited their clinical applications. To address these limitations, unsupervised domain adaptation (UDA) methods have been used to transfer knowledge from one labeled source domain to the unlabeled target domain, yet these approaches suffer from severe domain shift issues and often ignore the potential benefits of leveraging multiple relevant sources in practical applications. To address these limitations, in this work, we construct a Three-Branch Mixed extractor and propose a Bi-level Multi-source unsupervised domain adaptation method called BTMuda for breast cancer diagnosis. Our method addresses the problems of domain shift by dividing domain shift issues into two levels: intra-domain and inter-domain. To reduce the intra-domain shift, we jointly train a CNN and a Transformer as two paths of a domain mixed feature extractor to obtain robust representations rich in both low-level local and high-level global information. As for the inter-domain shift, we redesign the Transformer delicately to a three-branch architecture with cross-attention and distillation, which learns domain-invariant representations from multiple domains. Besides, we introduce two alignment modules - one for feature alignment and one for classifier alignment - to improve the alignment process. Extensive experiments conducted on three public mammographic datasets demonstrate that our BTMuda outperforms state-of-the-art methods.