Artificial intelligence for risk-stratified breast cancer screening: a systematic review of evidence, clinical integration, and ethical implications in risk assessment tools.
Authors
Affiliations (9)
Affiliations (9)
- Breast Imaging Division, IEO European Institute of Oncology IRCCS, Milan, Italy. [email protected].
- Breast Unit, Pederzoli Hospital, Peschiera Del Garda, Italy.
- Screening Unit, Oncological Network, Prevention and Research Institute - ISPRO, Florence, Italy.
- Breast Radiology Unit, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy.
- Breast Imaging Division, IEO European Institute of Oncology IRCCS, Milan, Italy.
- Division of Breast Surgery, IEO European Institute of Oncology IRCCS, Milan, Italy.
- Department of Oncology and Hemato-Oncology, University of Milan, Milan, Italy.
- Division of Cancer Prevention and Genetics, IEO European Institute of Oncology IRCCS, Milan, Italy.
- Radiology Department, Foundation IRCCS Ca' Granda, Ospedale Maggiore Policlinico, Milan, Italy.
Abstract
Conventional age-based breast cancer screening ignores substantial inter-individual risk variation, contributing to overdiagnosis, false positives, and missed opportunities for earlier detection in high-risk women. Mammography-based artificial intelligence (AI) may enable risk-stratified screening and more efficient workflows. To systematically review evidence on mammography-based AI for personalized breast cancer screening, covering risk prediction, detection/triage, decision support, and associated ethical, economic, and equity implications. We searched MEDLINE/PubMed, Embase, Scopus, Web of Science, and the Cochrane Library (January 2015-November 2025) for studies evaluating AI-enabled personalization in breast cancer screening. Two reviewers independently screened 612 records, assessed 77 full texts, and included 30 studies; data were synthesized narratively. Image-based deep-learning risk models consistently outperformed traditional clinical risk tools and enriched future cancers within small high-risk strata, including cancers presenting as interval cancers in recent validations. Prospective trials and real-world implementations indicate that AI-supported reading can maintain or modestly improve cancer detection while reducing radiologist workload by roughly 40-50%. Decision-analytic models suggest that AI-enabled risk-stratified policies may be cost-effective but rely on assumptions and lack prospective confirmation of long-term endpoints. Key evidence gaps include prospective demonstration that acting on AI-derived risk reduces interval/advanced cancers, subgroup equity performance, explainability, and governance. AI is a credible enabler of personalized mammography screening, with the most mature near-term use cases being identification of women at highest short-term risk for intensified surveillance and workflow optimization via AI-supported reading. Interval extension for low-risk groups should be implemented only within carefully monitored pilots and prospective outcome studies, with predefined safety, equity audits, and governance safeguards.