Title

Rapid Exclusion of COVID Infection With the Artificial Intelligence Electrocardiogram

Authors

Zachi I. Attia, Department of Cardiovascular Medicine, Mayo Clinic College of Medicine, Rochester, MN.
Suraj Kapa, Department of Cardiovascular Medicine, Mayo Clinic College of Medicine, Rochester, MN.
Jennifer Dugan, Department of Cardiovascular Medicine, Mayo Clinic College of Medicine, Rochester, MN.
Naveen Pereira, Department of Cardiovascular Medicine, Mayo Clinic College of Medicine, Rochester, MN.
Peter A. Noseworthy, Department of Cardiovascular Medicine, Mayo Clinic College of Medicine, Rochester, MN.
Francisco Lopez Jimenez, Department of Cardiovascular Medicine, Mayo Clinic College of Medicine, Rochester, MN.
Jessica Cruz, Department of Cardiovascular Medicine, Mayo Clinic College of Medicine, Rochester, MN.
Rickey E. Carter, Department of Health Sciences Research, Mayo Clinic College of Medicine, Jacksonville, FL.
Daniel C. DeSimone, Department of Cardiovascular Medicine, Mayo Clinic College of Medicine, Rochester, MN; Division of Infectious Diseases, Mayo Clinic College of Medicine, Rochester, MN.
John Signorino, Department of Compliance, Mayo Clinic College of Medicine, Rochester, MN.
John Halamka, Mayo Clinic Platform, Mayo Clinic College of Medicine, Rochester, MN.
Nikhita R. Chennaiah Gari, Department of Hepatology and Transplant, Mayo Clinic College of Medicine, Rochester, MN.
Raja Sekhar Madathala, Department of Internal Medicine, Mayo Clinic College of Medicine, Austin, MN.
Pyotr G. Platonov, Department of Cardiology, Clinical Sciences, Lund University, Lund, Sweden.
Fahad Gul, Division of Cardiology, Heart and Vascular Institute, Einstein Healthcare Network, Philadelphia, PA.
Stefan P. Janssens, Department of Cardiovascular Diseases, University Hospitals Leuven, KU Leuven, Leuven, Belgium.
Sanjiv Narayan, Cardiovascular Institute and Department of Cardiovascular Medicine, Stanford University Medical Center, Stanford, CA.
Gaurav A. Upadhyay, Section of Cardiology, Department of Medicine, University of Chicago, Chicago, IL.
Francis J. Alenghat, Section of Cardiology, Department of Medicine, University of Chicago, Chicago, IL.
Marc K. Lahiri, Henry Ford Hospital, Heart and Vascular Institute, Detroit, MI.
Karl Dujardin, Department of Cardiology, AZ Delta Hospital, AZ Delta Campus Rumbeke, Deltalaan, Belgium.
Melody Hermel, Scripps Health and the Scripps Clinic Division of Cardiology, La Jolla, CA.
Paari Dominic, Louisiana State University Health Sciences Center, Shreveport, LA.
Karam Turk-Adawi, Qatar University, QU-Health, Doha, Qatar.
Nidal Asaad, Hamad Medical Corporation, Doha, Qatar.
Anneli Svensson, Department of Cardiology and Department of Medical and Health Sciences, Linköping University Hospital, Linköping, Sweden.
Francisco Fernandez-Aviles, Hospital General Universitario Gregorio Maranon, Instituto de Investigacion Sanitaria Gregorio Maranon, Universidad Complutense, Madrid, Spain.
Darryl D. Esakof, Department of Cardiology, Lahey Hospital & Medical Center, Burlington, MA.
Jozef Bartunek, Cardiovascular Center, Aalst, OLV Hospital, Belgium.
Amit Noheria, Department of Cardiovascular Medicine, The University of Kansas Health System, Kansas City, KS.
Arun R. Sridhar, Section of Cardiac Electrophysiology, University of Washington Medical Center, Seattle, WA.
Gaetano A. Lanza, Fondazione Policlinico Universitario A. Gemelli IRCCS, Universita Cattolica del Sacro Cuore, Cardiology Institute, Rome, Italy.
Kevin Cohoon, Division of Cardiovascular Medicine Froedtert & the Medical College of Wisconsin, Milwaukee, WI.

Document Type

Article

Publication Date

8-1-2021

Abstract

OBJECTIVE: To rapidly exclude severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection using artificial intelligence applied to the electrocardiogram (ECG). METHODS: A global, volunteer consortium from 4 continents identified patients with ECGs obtained around the time of polymerase chain reaction-confirmed COVID-19 diagnosis and age- and sex-matched controls from the same sites. Clinical characteristics, polymerase chain reaction results, and raw electrocardiographic data were collected. A convolutional neural network was trained using 26,153 ECGs (33.2% COVID positive), validated with 3826 ECGs (33.3% positive), and tested on 7870 ECGs not included in other sets (32.7% positive). Performance under different prevalence values was tested by adding control ECGs from a single high-volume site. RESULTS: The area under the curve for detection of acute COVID-19 infection in the test group was 0.767 (95% CI, 0.756 to 0.778; sensitivity, 98%; specificity, 10%; positive predictive value, 37%; negative predictive value, 91%). To more accurately reflect a real-world population, 50,905 normal controls were added to adjust the COVID prevalence to approximately 5% (2657/58,555), resulting in an area under the curve of 0.780 (95% CI, 0.771 to 0.790) with a specificity of 12.1% and a negative predictive value of 99.2%. CONCLUSION: Infection with SARS-CoV-2 results in electrocardiographic changes that permit the artificial intelligence-enhanced ECG to be used as a rapid screening test with a high negative predictive value (99.2%). This may permit the development of electrocardiography-based tools to rapidly screen individuals for pandemic control.

Publication Source (Journal or Book title)

Mayo Clinic proceedings

First Page

2081

Last Page

2094

This document is currently not available here.

COinS