AI flags hard-to-detect heart disease in seconds
Researchers have developed an advanced artificial intelligence (AI) model capable of detecting signs of coronary microvascular dysfunction (CMVD) on electrocardiograms (ECGs) in a matter of seconds. The group shared its findings in NEJM AI, an algorithm-focused publication from The New England Journal of Medicine.[1]
“Our model creates a way for clinicians to accurately identify a condition that is notoriously hard to diagnose—and often missed in emergency department visits—using a 10-second ECG strip,” senior-author Venkatesh L. Murthy, MD, PhD, associate chief of cardiology for translational research and innovation at the U-M Health Frankel Cardiovascular Center and the Melvyn Rubenfire Professor of Preventive Cardiology at U-M Medical School, said in a statement.
Murthy et al. trained a deep learning model using data from more than 800,000 unlabeled ECGs. It was then fine-tuned using a mix of PET imaging results and clinical reports. The advanced AI model’s area under the ROC curve ranged from 0.763 for detecting impaired myocardial flow reserve to 0.955 for detecting impaired left ventricular ejection fraction.
“Essentially, we taught the model to ‘understand’ the electrical language of the heart without human supervision,” Murthy explained.
Early data suggest the new AI model is a smashing success. It was linked to improved diagnostic accuracy for 11 of the 12 traditional prediction tasks the researchers were evaluating.
“People who come to the ER for chest pain might have CMVD, but their angiogram will show up as ‘clear,’” co-author Sascha N. Goonewardena, MD, associate professor of cardiology at U-M Medical School, said in the same statement. “In hospitals with limited resources or non-specialty centers, using our ECG-AI model to predict myocardial flow reserve and CMVD will be an easy, cost-effective and noninvasive way to identify when a patient would benefit from advanced testing for a serious condition.”
The Department of Veterans Affairs, National Heart, Lung and Blood Institute, National Institute of Diabetes and Digestive and Kidney Diseases and National Institute on Aging all provided funding for this research.
Click here to read the full study.
Michael has more than 19 years of experience as a professional writer and editor. He has written at length about cardiology, radiology, artificial intelligence and other key healthcare topics.