Get started

Please fill in the form below and an Ezra representative will contact you within 1 business day.

Notify me

Please fill in the form below and we will be in touch when we launch in your city.

Blog / Artificial Intelligence, Healthcare Trends, Heart Health

Artificial Intelligence Better Able to Predict Heart Attacks Than Other Risk Models

Jul. 02 2019 by Sheherzad Raza Preisler Blog Editor, PR, & Social Media Coordinator
Artificial Intelligence Better Able to Predict Heart Attacks Than Other Risk Models

An encouraging study on the ability of Artificial Intelligence (AI) to predict heart attacks, among other cardiac diseases, was published in Radiology on June 25. In fact, machine learning (ML) was found to be more effective than traditionally employed risk models. 

The research team, led by Yale’s Kevin M. Johnson, designed an ML system that was able to analyze coronary computed tomography angiography (CCTA) images and distill details from them in order to construct a “more comprehensive prognostic picture.”

In a press release, Johnson explained his method: “Starting from the ground up, I took imaging features from the coronary CT. Each patient had 64 of these features and I fed them into an ML algorithm. The algorithm is able to pull out the patterns in the data and predict that patients with certain patterns are more likely to have an adverse event like a heart attack than patients with other patterns.”

For the study, Johnson and his team compared their novel ML system to the coronary artery disease reporting and data system (CAD-RADS), which is currently the industry standard in encapsulating a patient’s CCTA outcomes. They used data from over 6,000 patients, and followed them for an average of 9 years post-CCTA. During the study, 380 of the patients passed away: 43 from heart attacks, 70 from coronary artery disease.

What Johnson et al. found was encouraging: their ML model was more accurate than the CAD-RADS-generated risk estimate. Because of the exciting results, Johnson is already continuing to add to the ML system by incorporating risk factors unrelated to results from imaging tests, which he believes will strengthen the method.