Mammograms are mostly known as a tool for finding breast cancer, but new research shows they could also help detect early signs of heart disease in women.
A study presented at the American College of Cardiology’s Annual Scientific Session (ACC.25) found that artificial intelligence (AI) can use mammogram images to measure calcium buildup in the arteries of the breast.
This buildup may signal a higher risk of heart problems.
Each year, about 40 million mammograms are performed in the United States. The U.S. Centers for Disease Control and Prevention (CDC) recommends that middle-aged and older women get mammograms every one or two years.
Sometimes, these mammograms show calcium in the breast arteries, but doctors usually don’t report this information. The new study shows that AI can help by automatically analyzing these calcium levels and giving a score that shows a woman’s risk of heart disease.
Dr. Theo Dapamede from Emory University, the lead author of the study, said this discovery could allow women to be screened for both breast cancer and heart disease at the same time.
He explained that calcium in the breast arteries is a strong sign of heart problems, especially in women under 60. If doctors can spot these signs early, they can send patients to a heart specialist for more tests and care.
Heart disease is the leading cause of death in women, but many cases go undiagnosed. There’s also a lack of awareness about heart disease in women.
The researchers believe using AI with mammograms could help doctors find more women with early signs of heart trouble—using a test they’re already getting for breast cancer.
Calcium buildup in the blood vessels shows damage linked to aging or early-stage heart disease. Past studies show that women with calcium in their arteries have a 51% higher chance of heart attacks and strokes.
To create the screening tool, researchers trained an AI model to spot the calcium in the mammogram images. These calcium spots appear as bright spots in the pictures. The AI then used patient health records to predict their chances of future heart problems.
Unlike older models, this new one uses a special method to separate the blood vessels and gives a more accurate reading. The study used a large set of data from over 56,000 women who had mammograms at Emory Healthcare from 2013 to 2020 and were followed for at least five years.
Dr. Dapamede said that new AI tools make it easier to find more health information in medical images, even when those images were made for other reasons.
The AI model was able to sort women into low, medium, or high risk for heart problems. The study found that more calcium meant a higher chance of serious events like heart attacks, strokes, or even death within two to five years.
This was true for women under 60 and those between 60 and 80, but not for those over 80. This means the tool may work best for spotting risk in younger women, who could benefit more from early treatment.
The study also found that women with the most calcium had a much lower chance of staying healthy over five years. For example, 86.4% of women with high calcium lived five more years without major heart problems, compared to 95.3% of women with low calcium.
This means those with high calcium were nearly three times more likely to die within five years.
The AI model was developed by Emory Healthcare and Mayo Clinic but is not available for public use yet. The team hopes to get approval from the U.S. Food and Drug Administration in the future so the tool can be used in clinics across the country.
They also plan to see if the same technology can help detect other diseases, like kidney disease or problems in arteries outside the heart.
If you care about heart health, please read studies about how eating eggs can help reduce heart disease risk, and Vitamin K2 could help reduce heart disease risk.
For more information about heart health, please see recent studies about how to remove plaques that cause heart attacks, and results showing a new way to prevent heart attacks, strokes.
Copyright © 2025 Knowridge Science Report. All rights reserved.
