An AI-powered test that can rapidly identify life-threatening heart condition atrial fibrillation has been developed by researchers from Mayo Clinic
An AI-powered electrocardiograph (EKG) — a device that measures electrical activity of the heart — can quickly detect atrial fibrillation (AF), which causes an irregular and abnormally fast heartbeat.
The technology could help identify existing or emerging AF within seconds, compared with current tests that take anything from weeks to years, according to new Mayo Clinic research.
Rapid diagnosis can then help guide both preventative therapy and emergency treatment.
Up to 6.1 million people in the US suffer from AF, and with the ageing of the country’s population, this figure is expected to increase along with the risk of stroke and long-term heart problems.
Dr Paul Friedman, chair of the Department of Cardiovascular Medicine at Mayo Clinic, said: “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.
“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.”
How the electrocardiograph detects signs of atrial fibrillation
According to Centres for Disease Control and Prevention, when a person has AF, the beating in the upper chambers of the heart becomes irregular, meaning blood doesn’t flow as well as it should from the atria to the lower chambers of the organ.
This can lead to life-threatening complications such as blood clots, stroke, heart failure and other cardio-related problems.
At present, clinicians rely on standard cardiac telemetry to detect AF seen commonly after a stroke.
The study, involving almost 181,000 patients between 1993 and 2017, and published in medical journal The Lancet, is the first to use deep learning to identify patients with undetected AF and had an overall accuracy of 83%.
Identification of risk factors for stroke has always been a challenge for doctors, especially if the heart is in normal rhythm during a test.
Researchers from the Mayo Clinic describe how they trained a neural network to recognise subtle differences in nearly 650,000 EKGs from the 181,000 patients.
Dr Friedman, who is a cardiac electrophysiologist, says that he is “surprised” by the findings of this research. He believes AI-powered EKGs could direct the right treatment for disease caused by AF, even without symptoms.
Additionally, the technology could eventually be made available on a large scale through the use of a smartphone or watch app.
Due to the growing incidences of chronic heart diseases, along with technological advancements, the global AF market is estimated to reach US $14.68bn by 2026, according to a new report by a market research and consulting firm Reports and Data.
But the team said the modelling now needed to be tested further to see if it could be deployed on the front-line.
“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,” said 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.”