How AI Can Turn Your Mammogram Into a Heart Health Check
You already get mammograms to screen for breast cancer. But did you know those same images may hold clues about your heart? Thanks to advances in artificial intelligence, researchers are unlocking hidden signs in mammograms that could help assess future cardiovascular risk, without extra tests or scans.
The AI Breakthrough: Mammograms + Age = Heart Risk Insight
In a recent study published in Heart (BMJ), researchers developed a deep learning model that uses **only mammographic image features and age** to predict a woman’s risk of cardiovascular events.
This isn’t about reading breasts for heart issues, it’s about detecting **breast arterial calcification (BAC)** and other imaging patterns linked to vascular health. In the study, the AI model performed similarly to established risk calculators — using data already available in the mammogram.
Why Mammograms Are a Unique Opportunity
Cardiovascular disease is the leading cause of death in women, yet risk is often under recognised. Meanwhile, many women already undergo mammography regularly. The study authors point out that combining breast screening with cardiovascular insight offers a “two-for-one” screening opportunity.
Because this AI model requires no extra blood tests or clinical data, it could be especially useful in settings where full cardiovascular screening is harder to access.
What the Research Found
The team analyzed data from nearly **49,000 women** in Victoria, Australia (the Lifepool cohort) with about 8.8 years of follow-up. Over that time, more than 3,300 participants had a first cardiovascular event (such as heart attack, stroke, heart failure, or coronary disease).
The AI model’s predictive accuracy (c-index ~0.72) matched traditional methods like PREVENT and PREDICT — despite relying only on the image and age.
How This Adds to Other AI Efforts
Another line of research looked specifically at BAC quantification using AI across **116,135 women** from two health systems. This transformer-based model classified BAC severity (none, mild, moderate, severe) and found that **even mild BAC was associated** with increased risk of major adverse cardiovascular events (MACE). The highest levels of BAC showed hazard ratios ~2.0 for events — independent of traditional risk factors.
Together, these studies suggest mammogram-derived AI signals carry independent prognostic value and may detect risk even in younger women.
Strengths, Limitations, and What’s Next
Strengths: The model’s simplicity — no need for extra tests or patient history — makes it pragmatic and scalable. Embedding risk insight into an already common screening adds value at low cost.
Limitations: Mammogram machines differ, potentially affecting consist. Finally, AI models depend heavily on their training populations; generalization is an open question.
Future work will need prospective validation, economic evaluation, and implementation planning — especially across diverse populations. Questions remain: Which cardiovascular outcomes are best predicted? Will clinicians act on mammogram-based risk scores?
What You Can Do Today
If you get mammograms, you might ask your provider whether they assess or report **breast arterial calcification (BAC)** in the imaging. Even if AI tools aren’t in use yet, being aware of this evolving field is empowering.
Continue prioritising established heart-healthy practices: balanced nutrition, regular physical activity, not smoking, and managing blood pressure, cholesterol, and glucose with your healthcare team.
Final Thoughts
AI is opening a new frontier in women’s health — turning mammograms into a window not just for detecting breast cancer, but for highlighting hidden cardiovascular risk. While this technology is not yet standard care, its promise to enrich preventive screening is compelling and worth watching.
Disclaimer
This content is for educational purposes only. It does not diagnose or treat any condition. Always consult a qualified healthcare provider for medical guidance.
