‘Huge for patients’: AI forecasts new atrial fibrillation, stroke risk

A recently developed algorithm uses common electrocardiograms to predict patients’ future risk of developing new atrial fibrillation and stroke.

Cardiology experts created their artificial intelligence tool using more than 1.5 million ECGs from individuals treated over a 35-year timespan. AI predicted new-onset atrial fibrillation with a specificity of 81% and sensitivity of 69%, according to data published in Circulation.

With upward of 300 million ECGs completed in the U.S. each year, the authors believe their advancement may significantly enhance cardiovascular care.

"This critical work stems from our major investments in cardiology to generate algorithms that make existing cardiology tests, such as ECGs, smarter and capable of predicting future clinical events,” Joel Dudley, chief scientific officer at precision medicine firm Tempus, which helped create the algorithm, said Monday. “Our goal is to enable clinicians to act earlier in the course of disease."

Conventional 12-lead ECGs can’t typically forecast future outcomes such as stroke, the authors noted. So, they set out to create a deep neural network to better utilize the electric signals emitted by the heart.

That included 1.6 million ECGs from 430,000 patients treated at Geisinger Health System in Danville, Pennsylvania. They specifically designed the tool to predict adverse heart events in patients with no prior history of atrial fibrillation, but who were likely to experience the potentially deadly event within 12 months. 

The neural network beat out current clinical models and identified 62% of patients as high-risk who later went on to suffer an AF-related stroke within three years.

"Not only can we now predict who is at risk of developing atrial fibrillation, but this work shows that the high-risk prediction precedes many AF-related strokes," added co-senior author Brandon Fornwalt, MD, PhD, chair of Geisinger's Department of Translational Data Science and Informatics. "With that kind of information, we can change the way these patients are screened and treated, potentially preventing such severe outcomes. This is huge for patients."

Around the web

The Department of Defense is gifting Case Western researchers a grant to study the use of AI in determining whether patients require surgery. 

This study brings light to the prospect of a "fully automated solution" for echocardiogram analysis, experts reported.

Experts noted a "significant" reduction in false positives and false negatives using their modified machine learning model.

Trimed Popup
Trimed Popup