A Multi-Modal AI System for Screening Mammography: Integrating 2D and 3D Imaging to Improve Breast Cancer Detection in a Prospective Clinical Study

2025-04-08

AI Summary

To address high false-positive recall rates and excessive radiologist workload in digital breast tomosynthesis (DBT) screening, this work proposes the first clinically optimized tri-modal AI system integrating full-field digital mammography (FFDM), synthetic mammography, and DBT. The system simultaneously performs breast-level malignancy risk prediction and bounding-box localization of suspicious lesions. Methodologically, it introduces a novel 2D/3D joint feature extraction architecture, cross-modal attention-based fusion, and an interpretable lesion localization module. In internal evaluation, the system achieves an AUROC of 0.945. During multicenter prospective deployment, it maintains 100% sensitivity while substantially reducing recall rates for low-risk cases. External validation demonstrates a 35.31%–69.14% reduction in AUROC gap relative to baseline models, significantly enhancing both screening efficacy and clinical deployability.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Although digital breast tomosynthesis (DBT) improves diagnostic performance over full-field digital mammography (FFDM), false-positive recalls remain a concern in breast cancer screening. We developed a multi-modal artificial intelligence system integrating FFDM, synthetic mammography, and DBT to provide breast-level predictions and bounding-box localizations of suspicious findings. Our AI system, trained on approximately 500,000 mammography exams, achieved 0.945 AUROC on an internal test set. It demonstrated capacity to reduce recalls by 31.7% and radiologist workload by 43.8% while maintaining 100% sensitivity, underscoring its potential to improve clinical workflows. External validation confirmed strong generalizability, reducing the gap to a perfect AUROC by 35.31%-69.14% relative to strong baselines. In prospective deployment across 18 sites, the system reduced recall rates for low-risk cases. An improved version, trained on over 750,000 exams with additional labels, further reduced the gap by 18.86%-56.62% across large external datasets. Overall, these results underscore the importance of utilizing all available imaging modalities, demonstrate the potential for clinical impact, and indicate feasibility of further reduction of the test error with increased training set when using large-capacity neural networks.

Problem

• Reducing false-positive recalls in breast cancer screening
• Integrating 2D and 3D imaging for improved detection
• Decreasing radiologist workload while maintaining high sensitivity

Innovation

• Multi-modal AI integrates 2D and 3D imaging
• Reduces recalls and radiologist workload significantly
• Scalable with large-capacity neural networks