A groundbreaking study published in the European Heart Journal reveals an innovative application of artificial intelligence (AI) that could revolutionize cardiovascular disease detection, particularly in women. By leveraging routine mammography screenings, typically conducted to detect breast cancer, researchers have developed an AI-based system capable of quantifying breast arterial calcifications — a critical marker of cardiovascular risk. This approach promises to harness an existing healthcare infrastructure to identify women at risk of heart attacks, strokes, and other severe cardiovascular events without additional costs or procedures.
Calcium deposits in the arteries, known as arterial calcification, signal the hardening or stiffening of blood vessels, a primary contributor to cardiovascular disease progression. Traditionally, assessment of such calcifications requires dedicated imaging and expensive diagnostic tests. However, the team’s AI model can extract quantitative measurements of these calcifications directly from standard mammographic images, interpreting subtle indicators that are frequently overlooked in routine breast cancer screenings.
The study analyzed mammograms from 123,762 women with no baseline cardiovascular disease, applying machine learning algorithms to segment and measure the extent of calcification in breast arteries. These measurements were then correlated with longitudinal health outcomes, including incidence of stroke, heart attack, heart failure, and cardiovascular mortality. Remarkably, the AI’s quantification of breast arterial calcification (BAC) emerged as a powerful independent predictor of subsequent cardiovascular events.
According to Dr. Hari Trivedi, lead investigator from Emory University, the analysis showed a dose-dependent relationship between calcium burden and cardiovascular risk. Women with mild calcifications had approximately a 30% heightened risk of severe cardiovascular events compared to those with no calcification. This risk escalated sharply with moderate calcifications showing over 70% increased risk, and severe calcifications doubling or tripling the likelihood of adverse cardiovascular outcomes.
What makes this advancement particularly transformative is its applicability across various subgroups of patients. The predictive value of AI-detected BAC held true even for younger women under 50 years old, a demographic often considered at low cardiovascular risk and frequently underrepresented in preventive cardiology efforts. This challenges existing paradigms about risk stratification and underscores the significance of early detection.
Moreover, the diversity of the study cohort—comprising multiple racial and ethnic groups, and spanning two major US healthcare systems—adds robustness and generalizability to the findings. This inclusivity addresses a critical gap in cardiovascular research, as women remain disproportionately underdiagnosed and undertreated despite heart disease being the leading cause of female mortality worldwide.
In clinical practice, integrating this AI technology into mammography could seamlessly expand the utility of breast cancer screening platforms. Women undergoing mammograms would receive concurrent cardiovascular risk assessments without any change in workflow or additional tests, enabling earlier physician-patient discussions around preventative cardiology measures such as cholesterol monitoring, lifestyle modification, or initiation of pharmacotherapy.
For healthcare providers, this tool offers a pragmatic approach to identify high-risk individuals who might otherwise remain undetected by traditional screening strategies reliant on self-reporting or sporadic clinical encounters. Policymakers could capitalize on existing mammography infrastructures, which reach tens of millions of women annually, to scale up cardiovascular disease prevention at population levels with minimal incremental resource expenditure.
Future directions include clinical trials designed to evaluate the operational aspects of implementing AI-based BAC assessment protocols within standard mammography services. Such studies will clarify optimal methods for notifying patients and healthcare providers, addressing ethical considerations, and ensuring equitable access across diverse healthcare settings.
An editorial by Professor Lori B. Daniels from the University of California, San Diego, emphasizes the untapped potential of breast arterial calcification as a cardiovascular biomarker. She highlights the disconnect between high mammography adherence rates—up to two-thirds of women aged 50–69 in Europe and nearly 70% of women over 45 in the United States—and the relatively low awareness of personal cardiovascular risk factors like cholesterol levels.
Professor Daniels advocates for a paradigm shift, urging the medical community to move breast arterial calcification from a mere incidental observation in mammograms to a standardized, actionable metric. This transition could transform the landscape of cardiovascular prevention in women, leveraging a trusted screening platform to bridge the gap between cancer detection and heart health management.
This novel AI application embodies the convergence of medical imaging, machine learning, and cardiovascular medicine, showcasing the interdisciplinary innovation redefining modern healthcare. As cardiovascular disease remains the leading cause of death for women globally, such advances are crucial for closing the gender gap in diagnosis and treatment, ultimately saving countless lives through early, personalized intervention.
In summary, AI-enabled quantification of breast arterial calcifications from routine mammography offers a promising, cost-effective strategy to predict cardiovascular risk in women. By capitalizing on an established cancer screening tool, this approach holds the potential to revolutionize preventive cardiology, ensuring timely identification and management of cardiovascular disease in a population traditionally underserved by risk assessment frameworks.
Subject of Research: People
Article Title: Artificial intelligence–based quantification of breast arterial calcifications to predict cardiovascular morbidity and mortality
News Publication Date: 9-Mar-2026
Web References:
https://dx.doi.org/10.1093/eurheartj/ehag128
References:
- Dapamede et al., European Heart Journal, 2026.
- Editorial by Lori B. Daniels, European Heart Journal, 2026.
Image Credits: European Heart Journal / Hari Trivedi
Keywords
- Mammography
- Artificial intelligence
- Cardiovascular disorders
- Heart disease
- Cardiac arrest
