MammoDINO: Anatomically Aware Self-Supervision for Mammographic Images

2025-10-13

AI Summary

To address the limitations of self-supervised learning (SSL) in mammography—namely, scarce annotated data and domain shift—this paper introduces MammoDINO, the first SSL framework specifically designed for mammographic imaging. Methodologically, it features: (1) a breast-tissue-aware data augmentation sampler that preserves anatomical fidelity during pretext task construction; and (2) a cross-slice contrastive learning mechanism jointly leveraging 2D full-field digital mammograms (FFDM) and 3D digital breast tomosynthesis (DBT) slices to model hierarchical breast anatomy and spatial lesion correlations. Trained entirely without labels, MammoDINO achieves state-of-the-art performance across five benchmark datasets for breast cancer screening, demonstrating strong generalization. As a foundation model, it supports diverse downstream computer-aided diagnosis (CAD) tasks, significantly reducing clinical annotation burden and radiologist workload.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Self-supervised learning (SSL) has transformed vision encoder training in general domains but remains underutilized in medical imaging due to limited data and domain specific biases. We present MammoDINO, a novel SSL framework for mammography, pretrained on 1.4 million mammographic images. To capture clinically meaningful features, we introduce a breast tissue aware data augmentation sampler for both image-level and patch-level supervision and a cross-slice contrastive learning objective that leverages 3D digital breast tomosynthesis (DBT) structure into 2D pretraining. MammoDINO achieves state-of-the-art performance on multiple breast cancer screening tasks and generalizes well across five benchmark datasets. It offers a scalable, annotation-free foundation for multipurpose computer-aided diagnosis (CAD) tools for mammogram, helping reduce radiologists' workload and improve diagnostic efficiency in breast cancer screening.

Problem

• Develops self-supervised learning for mammography with limited annotated data
• Captures clinically meaningful features using anatomical-aware augmentation strategies
• Creates annotation-free foundation for multipurpose breast cancer screening tools

Innovation

• Self-supervised learning framework for mammography
• Breast tissue aware data augmentation sampler
• Cross-slice contrastive learning from 3D structure