AI accurately detects subtle signs of atrial fibrillation even in normal tests

pharmafile | August 2, 2019 | News story | Medical Communications AF, AI, ECG, atrial fibrillation, electrcardiogram, pharma 

A new study from the Mayo Clinic shows that artificial intelligence (AI) can detect signs of atrial fibrillation (AF) in an electrocardiogram (EKG), even if the heart is in a normal rhythm at the time of the test.

Atrial fibrillation, in which a person’s heart beats at an irregular rhythm, may not occur during a standard 10-12 second EKG, even if the patient has experienced an irregular heart beat before. Thus, due to its fleeting nature, the common condition is often hard to diagnose.  As such, the AI tool, an account of which is published in The Lancet, may help improve the efficacy of the EKG and improve current methods of screening for atrial fibrillation.

“When people come in with a stroke, we really want to know if they had AF in the days before the stroke, because it guides the treatment,” says Dr Paul Friedman, chair of the Department of Cardiovascular Medicine at Mayo Clinic. “Blood thinners are very effective for preventing another stroke in people with AF. But for those without AF, using blood thinners increases the risk of bleeding without substantial benefit. That’s important knowledge. We want to know if a patient has AF.”

“An EKG will always show the heart’s electrical activity at the time of the test, but this is like looking at the ocean now and being able to tell that there were big waves yesterday,” says Dr Friedman. “AI can provide powerful information about the invisible electrical signals that our bodies give off with each heartbeat — signals that have been hidden in plain sight.”

The new tool was developed using data from 450,000 EKGs taken from the Mayo Clinic’s collection of over 7 million EKGs. The researchers trained the AI to notice subtle differences between patients with atrial fibrillation and normal EKGs.

The researchers then tested the AI on EKGs from a group of 36,280 patients, of whom 3,051 were known to have atrial fibrillation. The AI correctly identified atrial fibrillation with 90% accuracy.

Dr Jeroen Hendriks, of the University of Adelaide commented: “Rather than finding the needle in the haystack by prolonged monitoring, authors basically suggest that AI will be able to judge by looking at the haystack if it has a needle hidden in it.”

Louis Goss

Related Content

Bayer and Aignostics to collaborate for AI oncology research

Bayer and Aignostics have announced that they have entered into a strategic collaboration for several …

Isomorphic Labs to collaborate with Eli Lilly

Isomorphic Labs has announced that it has entered into a strategic research collaboration with Eli …

Merck collaborates with Acceleration Consortium for AI experimentation

Merck and the Acceleration Consortium have announced their AI-driven experimentation planner Bayesian Back End (BayBE) …

Latest content