Detecting structural heart disease from electrocardiograms using AI

Timothy J. Poterucha, Linyuan Jing, Ramon Pimentel Ricart, Michael Adjei-Mosi, Joshua Finer, Dustin Hartzel, Christopher Kelsey, Aaron Long, Daniel Rocha, Jeffrey A. Ruhl, David vanMaanen, Marc A. Probst, Brock Daniels, Shalmali D. Joshi, Olivier Tastet, Denis Corbin, Robert Avram, Joshua P. Barrios, Geoffrey H. Tison, I-Min Chiu, David Ouyang, Alexander Volodarskiy, Michelle Castillo, Francisco A. Roedan Oliver, Paloma P. Malta, Siqin Ye, Gregg F. Rosner, Jose M. Dizon, Shah R. Ali, Qi Liu, Corey K. Bradley, Prashant Vaishnava, Carol A. Waksmonski, Ersilia M. DeFilippis, Vratika Agarwal, Mark Lebehn, Polydoros N. Kampaktsis, Sofia Shames, Ashley N. Beecy, Deepa Kumaraiah, Shunichi Homma, Allan Schwartz, Rebecca T. Hahn, Martin Leon, Andrew J. Einstein, Mathew S. Maurer, Heidi S. Hartman, John Weston Hughes, Christopher M. Haggerty and Pierre Elias ()
Additional contact information
Timothy J. Poterucha: Columbia University Irving Medical Center
Linyuan Jing: NewYork-Presbyterian Hospital
Ramon Pimentel Ricart: Columbia University Irving Medical Center
Michael Adjei-Mosi: Columbia University Vagelos College of Physicians and Surgeons
Joshua Finer: NewYork-Presbyterian Hospital
Dustin Hartzel: NewYork-Presbyterian Hospital
Christopher Kelsey: NewYork-Presbyterian Hospital
Aaron Long: Columbia University Irving Medical Center
Daniel Rocha: NewYork-Presbyterian Hospital
Jeffrey A. Ruhl: NewYork-Presbyterian Hospital
David vanMaanen: NewYork-Presbyterian Hospital
Marc A. Probst: Columbia University Irving Medical Center
Brock Daniels: Weill Cornell Medicine
Shalmali D. Joshi: Columbia University
Olivier Tastet: Montreal Heart Institute
Denis Corbin: Montreal Heart Institute
Robert Avram: Montreal Heart Institute
Joshua P. Barrios: San Francisco
Geoffrey H. Tison: San Francisco
I-Min Chiu: Cedars Sinai
David Ouyang: Cedars Sinai
Alexander Volodarskiy: NewYork-Presbyterian Hospital-Queens
Michelle Castillo: Columbia University Irving Medical Center
Francisco A. Roedan Oliver: Columbia University Irving Medical Center
Paloma P. Malta: Columbia University Irving Medical Center
Siqin Ye: Columbia University Irving Medical Center
Gregg F. Rosner: Columbia University Irving Medical Center
Jose M. Dizon: Columbia University Irving Medical Center
Shah R. Ali: Columbia University Irving Medical Center
Qi Liu: Columbia University Irving Medical Center
Corey K. Bradley: Columbia University Irving Medical Center
Prashant Vaishnava: Columbia University Irving Medical Center
Carol A. Waksmonski: Columbia University Irving Medical Center
Ersilia M. DeFilippis: Columbia University Irving Medical Center
Vratika Agarwal: Columbia University Irving Medical Center
Mark Lebehn: Columbia University Irving Medical Center
Polydoros N. Kampaktsis: Columbia University Irving Medical Center
Sofia Shames: Columbia University Irving Medical Center
Ashley N. Beecy: Weill Cornell Medicine
Deepa Kumaraiah: Columbia University Irving Medical Center
Shunichi Homma: Columbia University Irving Medical Center
Allan Schwartz: Columbia University Irving Medical Center
Rebecca T. Hahn: Columbia University Irving Medical Center
Martin Leon: Columbia University Irving Medical Center
Andrew J. Einstein: Columbia University Irving Medical Center
Mathew S. Maurer: Columbia University Irving Medical Center
Heidi S. Hartman: Columbia University Irving Medical Center
John Weston Hughes: Columbia University Irving Medical Center
Christopher M. Haggerty: NewYork-Presbyterian Hospital
Pierre Elias: Columbia University Irving Medical Center

Nature, 2025, vol. 644, issue 8075, 221-230

Abstract: Abstract Early detection of structural heart disease is critical to improving outcomes, but widespread screening remains limited by the cost and accessibility of imaging tools such as echocardiography1,2. Recent advances in machine learning applied to heart rhythm recordings have shown promise in identifying disease3,4, although previous work has been limited by development in narrow populations or targeting only select heart conditions5. Here we introduce a deep learning model, EchoNext, trained on more than 1 million heart rhythm and imaging records across a large and diverse health system to detect many forms of structural heart disease. The model demonstrated high diagnostic accuracy in internal and external validation, outperforming cardiologists in a controlled evaluation and showing consistent performance across different care settings and racial and/or ethnic groups. The models were prospectively evaluated in a clinical trial of patients without previous cardiac imaging, successfully identifying previously undiagnosed heart disease. These findings support the potential of artificial intelligence to expand access to heart disease screening at scale. To enable further development and transparency, we have publicly released model weights and a large, annotated dataset linking heart rhythm data to imaging-based diagnoses.

Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
https://www.nature.com/articles/s41586-025-09227-0 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:nat:nature:v:644:y:2025:i:8075:d:10.1038_s41586-025-09227-0

Ordering information: This journal article can be ordered from
https://www.nature.com/

DOI: 10.1038/s41586-025-09227-0

Access Statistics for this article

Nature is currently edited by Magdalena Skipper

More articles in Nature from Nature
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().