Domain generalization in deep learning-based mass detection in mammography: A large-scale multi-center study

TL;DR

This large-scale multi-center study investigates domain generalization in deep learning-based mammography mass detection, comparing multiple models and analyzing sources of domain shift to improve robustness across diverse clinical settings.

Contribution

It introduces a single-source training pipeline that enhances domain generalization and provides an in-depth analysis of factors affecting detection performance across different domains.

Findings

The proposed workflow outperforms transfer learning in four of five unseen domains.

Differences in patient age, breast density, and mass characteristics significantly impact detection accuracy.

The study offers best practices for improving domain robustness in deep learning breast cancer detection.

Abstract

Computer-aided detection systems based on deep learning have shown great potential in breast cancer detection. However, the lack of domain generalization of artificial neural networks is an important obstacle to their deployment in changing clinical environments. In this work, we explore the domain generalization of deep learning methods for mass detection in digital mammography and analyze in-depth the sources of domain shift in a large-scale multi-center setting. To this end, we compare the performance of eight state-of-the-art detection methods, including Transformer-based models, trained in a single domain and tested in five unseen domains. Moreover, a single-source mass detection training pipeline is designed to improve the domain generalization without requiring images from the new domain. The results show that our workflow generalizes better than state-of-the-art transfer learning-based approaches in four out of five domains while reducing the domain shift caused by the different acquisition protocols and scanner manufacturers. Subsequently, an extensive analysis is performed to identify the covariate shifts with bigger effects on the detection performance, such as due to differences in patient age, breast density, mass size, and mass malignancy. Ultimately, this comprehensive study provides key insights and best practices for future research on domain generalization in deep learning-based breast cancer detection.