Every year, tens of millions of women attend routine mammogram appointments. They're there for one reason: to look for breast cancer. They get their results, they go home, and the image is filed away.
But there's something hiding in those images that nobody has been looking at — something that could predict whether a woman is about to have a heart attack.
A landmark study published in the **European Heart Journal** in March 2026, involving more than **120,000 women**, shows that **artificial intelligence** can read standard mammogram images and detect an overlooked early warning sign of cardiovascular disease: calcium deposits in the walls of breast arteries.
The finding offers what researchers describe as a **"two-for-one" screening**: the same image that checks for breast cancer now, with AI, can assess a woman's heart disease risk — with no extra radiation, no extra test, and no extra appointment.
**What Is Breast Arterial Calcification?**
Breast arterial calcification (BAC) is exactly what it sounds like: calcium deposits forming on the inner walls of the small arteries running through breast tissue. These deposits are a recognised marker of **arterial stiffening** — the same underlying process that leads to cardiovascular disease.
BAC has been visible on mammograms for decades. Radiologists have noticed it. But it was never routinely reported to patients or their doctors, because it has nothing to do with breast cancer — the reason for the scan. It was essentially invisible information: present, but ignored.
AI changed that.
**What the Study Found**
The AI algorithms developed for this study can classify BAC into four categories: **absent, mild, moderate, or severe**. Across 120,000 women, the relationship between BAC severity and cardiovascular risk was clear and consistent:
❤️ **Mild BAC** → approximately **30% higher risk** of major cardiovascular events (heart attack, stroke, heart failure) ❤️ **Moderate BAC** → significantly elevated risk ❤️ **Severe BAC** → **two to three times** the risk of a serious cardiovascular event
Critically, this predictive value was **independent of traditional cardiovascular risk factors** — it added information over and above what cholesterol levels, blood pressure, age, and smoking history already predicted. And the link held even for women under 50, a group traditionally considered to be at lower risk.
**Why Women in Particular**
Cardiovascular disease is the **leading cause of death in women** worldwide — killing more women than all cancers combined. Yet heart disease in women is chronically under-diagnosed and under-treated. Women are less likely than men to be assessed for cardiovascular risk, and less likely to receive preventive interventions.
Mammograms are one of the most consistently attended preventive health screenings among women — particularly in countries with national screening programmes. Turning that existing contact point into a cardiovascular risk assessment, without adding any burden to the patient, is a profound opportunity.
Some imaging centres have already begun offering AI-based BAC analysis as an optional add-on to routine mammography. The push now is to integrate it into routine clinical practice and establish clear guidelines for how to communicate these findings to patients and clinicians.
**What It Means in Practice**
Imagine going for your routine mammogram and receiving, alongside your breast cancer result, a note: *"Your scan shows no signs of breast cancer. However, the AI analysis also detected mild breast arterial calcification. We recommend discussing this with your GP, as it is associated with a modestly elevated risk of cardiovascular events. A cholesterol test and blood pressure check are advised."*
That single sentence — triggered by a scan you were already having — could prompt a blood test, a prescription for a statin, a conversation about lifestyle, and ultimately prevent a heart attack that would otherwise have arrived without warning.
**The Bigger Pattern**
This is part of a broader revolution in what medical imaging can tell us. AI doesn't see images the way humans do. It sees patterns across millions of examples — correlations invisible to the human eye, signals that exist in the data but were never recognised as meaningful until the scale and pattern-matching of machine learning made them legible.
Every scan you've ever had contains information that no one thought to look for. AI is beginning to look.
The mammogram was already saving lives. Now, with AI reading a signal hiding in plain sight, it might be saving different lives in a completely different way. 💙
*Sources: European Heart Journal (March 2026) · ESC (European Society of Cardiology) · Medical News Today · Aunt Minnie · ICT&Health · 120,000+ women study cohort*
