[Skip to Content]
[Skip to Content Landing]
May 2018

Moving From Big Data to Deep Learning—The Case of Atrial Fibrillation

Author Affiliations
  • 1Department of Medicine, Stanford University School of Medicine, Stanford, California
  • 2Center for Digital Health, Stanford University School of Medicine, Stanford, California
  • 3Veterans Affairs Palo Alto Health Care System, Palo Alto, California
  • 4Associate Editor, JAMA Cardiology
JAMA Cardiol. 2018;3(5):371-372. doi:10.1001/jamacardio.2018.0207

While our understanding of the physiology of atrial fibrillation (AF) has been advanced by many developments in electrocardiography (ECG), electrophysiology, and imaging, the hallmark of the irregular pulse remains its cornerstone and fundamental to its identity. The earliest description of AF as an irregular pulse is believed to have been made more than 4000 years ago, when the Yellow Emperor of China described in detail the proper examination of the pulse and associated conditions of disease.1 Since then, Maimonides, Stokes, Wenckebach, and MacKenzie have all published detailed investigations of the irregular pulse.2