AI detection of pediatric heart murmurs comparable to cardiologists
Weeks after Eko announced plans to develop machine-learning technology to help physicians detect heart diseases, the AI-based network's results were comparable to cardiologists in detection accuracy.
“This is the first head-to-head study demonstrating that a deep neural network can detect pediatric heart murmurs with comparable accuracy to a cardiologist,” authors John Maidens, with Eko Devices, and Nicholas B. Slamon, MD, a pediatric critical care physician at Nemours Children’s Health System, wrote in a clinical study abstract recently published in the Journal of the American Heart Association.
The company’s deep neural network was trained on thousands of heart sound recordings before being tested on an independent dataset of audio recordings from a total of 54 patients. The study only included 42 patients who had echo-confirmed pathologic heart murmurs and normal heart sounds. The deep neural network performance was tested against five cardiologists and was able to outperform a majority of them.
“The ROC curve lies entirely below the 95 percent CI for one-fifth clinicians, entirely above the 95 percent CI for one-fifth clinicians and within the 95 percent CI for three-fifth (of the) clinicians,” the authors wrote.
Further study of the network is justified due to the small sample size.
“When it comes to healthcare, data almost always leads to better results because practitioners are able to make more informed decisions,” Slamon said in a statement. “Eko’s technology is leveraging the largest available dataset of previously captured heart sounds to elevate the skills of clinicians and in turn provide guidance on how to diagnose, and subsequently treat, serious, often fatal cardiac conditions. It’s a powerful advancement for the world of medicine.”
In late October, the Mayo Clinic in Rochester, Minnesota, and Eko announced plans to develop technology that uses machine learning to help physicians detect easily-missed heart diseases in patients. Eko is currently seeking FDA clearance for the murmur screening algorithm and hopes it will soon be used with its devices.
Danielle covers Clinical Innovation & Technology as a senior news writer for TriMed Media. Previously, she worked as a news reporter in northeast Missouri and earned a journalism degree from the University of Illinois at Urbana-Champaign. She's also a huge fan of the Chicago Cubs, Bears and Bulls.