← Latest briefing

Health

AI model detects heart disease risk from routine mammograms, study finds

Researchers presented findings showing machine-learning tools can analyze breast screening scans to identify high blood pressure, coronary heart disease, and stroke history in women.

The short version

  • Researchers from Tel Aviv University used an AI model to analyze over 97,000 routine mammograms, successfully identifying women with high blood pressure, coronary heart disease, and previous strokes.
  • The finding could allow existing breast cancer screening programs to serve a dual purpose by flagging cardiovascular risks without requiring additional imaging.
  • Researchers are working to improve the model's accuracy, lower false-positive rates, and expand the range of detectable conditions before moving to clinical implementation.

Key facts

  • Researchers analyzed 97,364 mammograms from 29,921 women with an average age of 54 to train a machine-learning model.[The Guardian]
  • The AI model identified women who had suffered a stroke with 86% reliability based on their mammograms, while identifying high blood pressure and coronary heart disease at 79% and 78% reliability, respectively.[The Guardian]
  • The study's results were presented at the European Society of Cardiology’s annual congress in Munich by Dr. Viana Copeland of Tel Aviv University.[The Guardian]
  • Cross-referenced medical records showed that 16% of the study's participants had high blood pressure, 2.5% had coronary heart disease, and 2.5% had previously experienced a stroke.[The Guardian]

What remains uncertain

  • The AI model requires further refinement to increase overall accuracy, reduce false results, and establish clinical reliability before real-world adoption.[The Guardian]

Sources