A.I. Predicts Heart Attack Risk Using X-ray


View of an x-ray film. | Image by OZMedia, Shutterstock

A new computer system can accurately predict a patient’s 10-year risk of heart attack or stroke based on a single chest X-ray, according to researchers from the Radiological Society of America.

The researchers leveraged artificial intelligence (AI) to identify patterns on a standard X-ray that correlated with hardening arteries. Hopefully, the discovery could one day pave the way for doctors to treat vulnerable individuals before conditions deteriorate beyond recovery.

In a media release, lead author Jakob Weiss, M.D., a radiologist affiliated with the Cardiovascular Imaging Research Center at Massachusetts General Hospital and the AI in Medicine program at the Brigham and Women’s Hospital in Boston, said the team’s “deep learning model offers a potential solution for population-based opportunistic screening of cardiovascular disease risk using existing chest X-ray images.”

According to Weiss, the technology could screen patients who might “benefit from statin medication but are currently untreated.”

The researchers used what is known as deep learning, a complex series of algorithms that allow computers to identify patterns in data and then produce forecasts. As more data is ingested, predictions, in theory, become more accurate. The advancement offers a potential revolution in heart therapy.

Typically, medication for heart disease can be prescribed based on elevated 10-year risk levels. The current method uses the ASCVD (atherosclerotic cardiovascular disease) grade to assess a patient’s risk based on numerous factors, like blood pressure, smoking history, age, and Type 2 diabetes status. Individuals scoring 7.5% or more are considered candidates for statins.

“The variables necessary to calculate ASCVD risk are often not available, which makes approaches for population-based screening desirable,” Dr. Weiss says. Since X-rays are common, Weiss believes the team’s approach may be valuable for identifying “individuals at high risk.”

The model used in the research is known as CXR-CVD risk. As part of its training, the model analyzed nearly 150,000 chest X-rays.

While X-rays have long been known to possess information “beyond traditional diagnostic findings,” Weiss says the data was not leveraged because adequate methods did not exist to analyze it. AI, however, is now “making it possible,” according to Weiss.

According to the researchers, the AI accurately predicted heart attacks and strokes in a group of roughly 11,000 people, finding a “significant association” between the risk level produced by the model and the actual cardiac event.

“Based on a single existing chest X-ray image, our deep learning model predicts future major adverse cardiovascular events with similar performance and incremental value to the established clinical standard,” said Weiss.

The promising findings will require further research — including controlled, randomized trials — before the technology can be validated and widely adopted. It is hoped that, eventually, it may help reduce the nearly 18 million lives lost each year to cardiovascular disease, the world’s top killer.

While not involved in the latest research, Dr. Donald Lloyd-Jones, chair of preventive medicine at Northwestern University’s Feinberg School of Medicine and former president of the American Heart Association, is encouraged by the results. “This is exactly the kind of application that artificial intelligence is best for… we need to continue to do things like this to really understand if we can find, particularly, patients who would otherwise slip through the cracks. I think that’s where it may be most useful,” he said.

If you enjoyed this article, please support us today!

Formed in 2021, we provide fact-based, non-partisan news. The Dallas Express is a non-profit organization funded by charitable support and advertising.

Please join us on the important journey to make Dallas a better place!

We welcome and appreciate comments on The Dallas Express as part of a healthy dialogue. We do ask that you be kind. Kind to each other and to everyone else in your comments. For more information, please refer to our Complete Comment Moderation Policy.

Subscribe to Comments
Notify of

Inline Feedbacks
View all comments