[Skip to Navigation]
February 10, 2021

Finding New Meaning in Everyday Electrocardiograms—Leveraging Deep Learning to Expand Our Diagnostic Toolkit

Author Affiliations
  • 1Division of Cardiology, Department of Medicine, University of California, San Francisco
  • 2Cardiovascular Research Institute, San Franciso, California
  • 3Bakar Computational Health Sciences Institute, University of California, San Francisco
JAMA Cardiol. 2021;6(5):493-494. doi:10.1001/jamacardio.2020.7460

Few diagnostic tests are as synonymous with cardiac activity as the humble electrocardiogram (ECG). As the most commonly obtained cardiovascular diagnostic test, the ECG has a ubiquity that belies the underlying physiologic complexity that it captures. Similarly, although congenital long QT syndrome (LQTS) is usually recognized by its namesake prolongation of the QT interval on the 12-lead ECG, the underlying pathophysiologic mechanisms capable of causing this syndrome are heterogeneous, complex, and as yet incompletely understood. Despite the numerous advances made in elucidating the various genetic causes of LQTS, the everyday tool used by physicians to screen for LQTS remains the standard ECG.