A study by digital stethoscope manufacturer Eko found that its artificial intelligence algorithm was able to detect heart murmurs in pediatric patients as accurately as a cardiologist, potentially allowing a broader number of physicians to screen patients without needing a trained specialist.
The machine-learning AI was trained on thousands of heartbeat recordings, the company said, before being tested against a data set of pediatric heart sounds and echocardiograms. Five pediatric cardiologists also listened to the recordings and determined whether they contained heart murmurs, with the AI outperforming four of them.
Eko is currently pursuing FDA clearance for the algorithm and plans to roll out the new feature to its monitoring devices already on the market, after receiving the regulatory greenlight. The agency previously cleared Eko’s handheld device combining a digital stethoscope and electrocardiogram, the Duo, in June 2017.
“Eko imagines a future where every physician is able to augment their clinical judgment and experience, leveraging thousands and thousands of previous medical cases in seconds,” Eko CEO Connor Landgraf said in a statement.
The study’s results, presented at the American Heart Association’s Scientific Sessions meeting in Chicago, could help narrow the skill gap between the expertly trained ears of the 27,000 cardiologists currently in the U.S. and millions of clinicians worldwide, Landgraf said.
Previous studies have shown that internal medicine and family practice physician residents can misdiagnose a large percentage of common cardiac events, while trained cardiologists can effectively diagnose 90% of cardiac events using a stethoscope, the company said.