AI-enabled echocardiograms identify higher risk of stroke

Mayo Clinic researchers found ECGs powered by artificial intelligence could predict irregular heart rhythms years before they are diagnosed. It could point to earlier interventions to prevent stroke.

Artificial intelligence-enabled electrocardiograms could determine risks of atrial fibrillation and possibly identify patients at risk of stroke or cognitive decline, Mayo Clinic researchers have found.

The researchers outlined their findings in a study in Mayo Clinic Proceedings, which was published online May 2.

Atrial fibrillation is an irregular heart rhythm that can eventually lead to blood clots and stroke. It is tied to one out of three ischemic strokes, where a vessel supplying blood to the brain is blocked, the authors noted. However, many patients aren’t aware they have atrial fibrillation, as one-third of all patients with AF have no symptoms.

The research suggests the potential for identifying those with atrial fibrillation much earlier. AI-enabled electrocardiogram algorithms could predict future atrial fibrillation up to 10 years before a clinical diagnosis, the study found.

The researchers conducted a population-based study of 3,729 individuals seen from November 29, 2004 through July 13, 2020. The participants did not have a known history of atrial fibrillation. The median age of the participants was 74.

Jonathan Graff-Radford, a Mayo Clinic neurologist and one of the contributing authors to the study, described the findings and their implications to the Mayo Clinic News Network.

"Application of this AI-ECG algorithm may be another way to screen individuals not only to determine risk of atrial fibrillation, but also to identify future risk of cognitive decline and stroke," Graff-Radford said.

The authors said it’s unclear if the higher probability of atrial fibrillation as identified by an artificial intelligence-enabled ECGs should lead to immediate treatment with anticoagulants, drugs that can prevent blood clots and strokes.

“The need for inexpensive, easy-to-obtain risk markers to identify potential candidates for early intervention is increasingly urgent as new anticoagulation and antidementia therapies are developed,” the authors wrote.

Still, AI-powered electrocardiograms should be considered “an attractive screening option,” since they are inexpensive, widely available, and already part of the normal workflows of health systems, the authors wrote.

Erika Weil, the lead author of the study, said the AI-enabled ECG algorithm offers “a fast and easy way to screen individuals not only to determine risk score for A-Fib, but also determine future risk of cognitive decline and infarcts.”

Weil discussed the findings of the study in this video by the Mayo Clinic. (The story continues after the video.)

Tests indicating the higher probability of atrial fibrillation were associated with declines in global cognition and attention among those patients who had no previous AF history.

In addition, high probabilities of AF in AI-enabled echocardiograms were associated with the presence of cerebral infractions on MRI exams.

Most of the infarctions were subcortical, the authors noted.  This suggests that AI-powered echocardiograms can not only predict atrial fibrillation but other indicators of cardiac disease, cognitive decline and disease affecting the blood flow to the brain.

In conclusion, the authors point to the value of AI-enabled echocardiograms, and the potential for earlier intervention for patients.

“Prospective, controlled studies are necessary to determine whether a high AF score is a biomarker to select patients for anticoagulation or more aggressive stroke risk factor modification to prevent cognitive impairment or cerebral infarcts,” the authors wrote.

The authors note the study’s participants were older, so the results can’t be extrapolated to younger people. However, the authors pointed to the large number of participants as one of the study’s strengths.