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

arXiv: https://arxiv.org/abs/2504.05636 · HTML: https://mlsanity.com/paper/2504.05636/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 88 references that are in the map.
- First posted on arXiv 8 Apr 2025.

## Builds on (most cited first, 17 of 17)

- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Attention-based Deep Multiple Instance Learning](https://mlsanity.com/paper/1802.04712/) (2018)
- [YOLOX: Exceeding YOLO Series in 2021](https://mlsanity.com/paper/2107.08430/) (2021)
- [A Closer Look at Memorization in Deep Networks](https://mlsanity.com/paper/1706.05394/) (2017)
- [Deep Learning is Robust to Massive Label Noise](https://mlsanity.com/paper/1705.10694/) (2017)
- [Deep Neural Networks Improve Radiologists' Performance in Breast Cancer Screening](https://mlsanity.com/paper/1903.08297/) (2019)
- [End-to-end Training for Whole Image Breast Cancer Diagnosis using An All Convolutional Design](https://mlsanity.com/paper/1711.05775/) (2017)
- [VinDr-Mammo: A large-scale benchmark dataset for computer-aided diagnosis in full-field digital mammography](https://mlsanity.com/paper/2203.11205/) (2022)
- [An interpretable classifier for high-resolution breast cancer screening images utilizing weakly supervised localization](https://mlsanity.com/paper/2002.07613/) (2020)
- [OPTIMAM Mammography Image Database: a large scale resource of mammography images and clinical data](https://mlsanity.com/paper/2004.04742/) (2020)
- [Detection of masses and architectural distortions in digital breast tomosynthesis: a publicly available dataset of 5,060 patients and a deep learning model](https://mlsanity.com/paper/2011.07995/) (2020)
- [Globally-Aware Multiple Instance Classifier for Breast Cancer Screening](https://mlsanity.com/paper/1906.02846/) (2019)
- [Adaptation of a deep learning malignancy model from full-field digital mammography to digital breast tomosynthesis](https://mlsanity.com/paper/2001.08381/) (2020)
- [MAMMO: A Deep Learning Solution for Facilitating Radiologist-Machine Collaboration in Breast Cancer Diagnosis](https://mlsanity.com/paper/1811.02661/) (2018)
- [Joint 2D-3D Breast Cancer Classification](https://mlsanity.com/paper/2002.12392/) (2020)
- [Meta-repository of screening mammography classifiers](https://mlsanity.com/paper/2108.04800/) (2021)
- [Leveraging Transformers to Improve Breast Cancer Classification and Risk Assessment with Multi-modal and Longitudinal Data](https://mlsanity.com/paper/2311.03217/) (2023)

## Built on by (most cited first, 1 of 1)

- [MammoClean: Toward Reproducible and Bias-Aware AI in Mammography through Dataset Harmonization](https://mlsanity.com/paper/2511.02400/) (2025)

## Abstract (arXiv)

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.