{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,13]],"date-time":"2026-04-13T12:57:03Z","timestamp":1776085023406,"version":"3.50.1"},"reference-count":26,"publisher":"Oxford University Press (OUP)","issue":"4","license":[{"start":{"date-parts":[[2024,1,24]],"date-time":"2024-01-24T00:00:00Z","timestamp":1706054400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/academic.oup.com\/pages\/standard-publication-reuse-rights"}],"funder":[{"DOI":"10.13039\/100007184","name":"Yale School of Medicine","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100007184","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000050","name":"National Heart, Lung, and Blood Institute","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100000050","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000002","name":"National Institutes of Health","doi-asserted-by":"publisher","award":["K23HL153775"],"award-info":[{"award-number":["K23HL153775"]}],"id":[{"id":"10.13039\/100000002","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000070","name":"National Institute of Biomedical Imaging and Bioengineering","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100000070","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000050","name":"National Heart, Lung, and Blood Institute","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100000050","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000038","name":"Food and Drug Administration","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100000038","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Department of Defense Advanced Research"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2024,4,3]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Objective<\/jats:title>\n                    <jats:p>Artificial intelligence (AI) detects heart disease from images of electrocardiograms (ECGs). However, traditional supervised learning is limited by the need for large amounts of labeled data. We report the development of Biometric Contrastive Learning (BCL), a self-supervised pretraining approach for label-efficient deep learning on ECG images.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Materials and Methods<\/jats:title>\n                    <jats:p>Using pairs of ECGs from 78\u200a288 individuals from Yale (2000-2015), we trained a convolutional neural network to identify temporally separated ECG pairs that varied in layouts from the same patient. We fine-tuned BCL-pretrained models to detect atrial fibrillation (AF), gender, and LVEF\u2009&amp;lt;\u200940%, using ECGs from 2015 to 2021. We externally tested the models in cohorts from Germany and the United States. We compared BCL with ImageNet initialization and general-purpose self-supervised contrastive learning for images (simCLR).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>While with 100% labeled training data, BCL performed similarly to other approaches for detecting AF\/Gender\/LVEF\u2009&amp;lt;\u200940% with an AUROC of 0.98\/0.90\/0.90 in the held-out test sets, it consistently outperformed other methods with smaller proportions of labeled data, reaching equivalent performance at 50% of data. With 0.1% data, BCL achieved AUROC of 0.88\/0.79\/0.75, compared with 0.51\/0.52\/0.60 (ImageNet) and 0.61\/0.53\/0.49 (simCLR). In external validation, BCL outperformed other methods even at 100% labeled training data, with an AUROC of 0.88\/0.88 for Gender and LVEF\u2009&amp;lt;\u200940% compared with 0.83\/0.83 (ImageNet) and 0.84\/0.83 (simCLR).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Discussion and Conclusion<\/jats:title>\n                    <jats:p>A pretraining strategy that leverages biometric signatures of different ECGs from the same patient enhances the efficiency of developing AI models for ECG images. This represents a major advance in detecting disorders from ECG images with limited labeled data.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1093\/jamia\/ocae002","type":"journal-article","created":{"date-parts":[[2024,1,2]],"date-time":"2024-01-02T16:55:06Z","timestamp":1704214506000},"page":"855-865","source":"Crossref","is-referenced-by-count":25,"title":["Biometric contrastive learning for data-efficient deep learning from electrocardiographic images"],"prefix":"10.1093","volume":"31","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-8524-1203","authenticated-orcid":false,"given":"Veer","family":"Sangha","sequence":"first","affiliation":[{"name":"Section of Cardiovascular Medicine, Department of Internal Medicine, Yale University , New Haven, CT, 06510, United States"},{"name":"Department of Engineering Science, Oxford University , Oxford, OX1 3PJ, United Kingdom"}]},{"given":"Akshay","family":"Khunte","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Yale University , New Haven, CT, 06511, United States"}]},{"given":"Gregory","family":"Holste","sequence":"additional","affiliation":[{"name":"Department of Electrical and Computer Engineering, The University of Texas at Austin , Austin, TX, 78712, United States"}]},{"given":"Bobak J","family":"Mortazavi","sequence":"additional","affiliation":[{"name":"Department of Computer Science & Engineering, Texas A&M University , College Station, TX, 77843, United States"},{"name":"Center for Outcomes Research and Evaluation (CORE), Yale New Haven Hospital , New Haven, CT, 06510, United States"}]},{"given":"Zhangyang","family":"Wang","sequence":"additional","affiliation":[{"name":"Department of Electrical and Computer Engineering, The University of Texas at Austin , Austin, TX, 78712, United States"}]},{"given":"Evangelos K","family":"Oikonomou","sequence":"additional","affiliation":[{"name":"Section of Cardiovascular Medicine, Department of Internal Medicine, Yale University , New Haven, CT, 06510, United States"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9467-6199","authenticated-orcid":false,"given":"Rohan","family":"Khera","sequence":"additional","affiliation":[{"name":"Section of Cardiovascular Medicine, Department of Internal Medicine, Yale University , New Haven, CT, 06510, United States"},{"name":"Center for Outcomes Research and Evaluation (CORE), Yale New Haven Hospital , New Haven, CT, 06510, United States"},{"name":"Section of Health Informatics, Department of Biostatistics, Yale School of Public Health , New Haven, CT, 06510, United States"}]}],"member":"286","published-online":{"date-parts":[[2024,1,24]]},"reference":[{"issue":"1","key":"2024040320063057200_ocae002-B1","doi-asserted-by":"crossref","first-page":"1760","DOI":"10.1038\/s41467-020-15432-4","article-title":"Automatic diagnosis of the 12-lead ECG using a deep neural network","volume":"11","author":"Ribeiro","year":"2020","journal-title":"Nat Commun"},{"issue":"1","key":"2024040320063057200_ocae002-B2","doi-asserted-by":"crossref","first-page":"65","DOI":"10.1038\/s41591-018-0268-3","article-title":"Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network","volume":"25","author":"Hannun","year":"2019","journal-title":"Nat Med"},{"issue":"1","key":"2024040320063057200_ocae002-B3","doi-asserted-by":"crossref","first-page":"70","DOI":"10.1038\/s41591-018-0240-2","article-title":"Screening for cardiac contractile dysfunction using an artificial intelligence-enabled electrocardiogram","volume":"25","author":"Attia","year":"2019","journal-title":"Nat Med"},{"issue":"1","key":"2024040320063057200_ocae002-B4","doi-asserted-by":"crossref","first-page":"36","DOI":"10.1161\/CIRCULATIONAHA.121.057869","article-title":"rECHOmmend: an ECG-based machine learning approach for identifying patients at increased risk of undiagnosed structural heart disease detectable by echocardiography","volume":"146","author":"Ulloa-Cerna","year":"2022","journal-title":"Circulation"},{"issue":"7","key":"2024040320063057200_ocae002-B5","doi-asserted-by":"crossref","first-page":"e014717","DOI":"10.1161\/JAHA.119.014717","article-title":"Deep learning-based algorithm for detecting aortic stenosis using electrocardiography","volume":"9","author":"Kwon","year":"2020","journal-title":"J Am Heart Assoc"},{"issue":"7","key":"2024040320063057200_ocae002-B6","doi-asserted-by":"crossref","first-page":"465","DOI":"10.1038\/s41569-020-00503-2","article-title":"Artificial intelligence-enhanced electrocardiography in cardiovascular disease management","volume":"18","author":"Siontis","year":"2021","journal-title":"Nat Rev Cardiol"},{"issue":"19","key":"2024040320063057200_ocae002-B7","doi-asserted-by":"crossref","first-page":"11941","DOI":"10.3390\/ijerph191911941","article-title":"ECG standards and formats for interoperability between mHealth and healthcare information systems: a scoping review","volume":"19","author":"Cuevas-Gonz\u00e1lez","year":"2022","journal-title":"Int J Environ Res Public Health"},{"issue":"10","key":"2024040320063057200_ocae002-B8","doi-asserted-by":"crossref","first-page":"681","DOI":"10.1016\/j.ijmedinf.2011.06.008","article-title":"A review of ECG storage formats","volume":"80","author":"Bond","year":"2011","journal-title":"Int J Med Inform"},{"issue":"138","key":"2024040320063057200_ocae002-B9","doi-asserted-by":"crossref","DOI":"10.1098\/rsif.2017.0821","article-title":"Computational techniques for ECG analysis and interpretation in light of their contribution to medical advances","volume":"15","author":"Lyon","year":"2018","journal-title":"J R Soc Interface"},{"issue":"1","key":"2024040320063057200_ocae002-B10","doi-asserted-by":"crossref","first-page":"24","DOI":"10.1186\/s42444-022-00075-x","article-title":"Clinical significance, challenges and limitations in using artificial intelligence for electrocardiography-based diagnosis","volume":"23","author":"Chung","year":"2022","journal-title":"Int J Arrhythmia"},{"issue":"1","key":"2024040320063057200_ocae002-B11","doi-asserted-by":"crossref","first-page":"1583","DOI":"10.1038\/s41467-022-29153-3","article-title":"Automated multilabel diagnosis on electrocardiographic images and signals","volume":"13","author":"Sangha","year":"2022","journal-title":"Nat Commun"},{"issue":"9","key":"2024040320063057200_ocae002-B12","doi-asserted-by":"crossref","first-page":"765","DOI":"10.1161\/CIRCULATIONAHA.122.062646","article-title":"Detection of left ventricular systolic dysfunction from electrocardiographic images","volume":"148","author":"Sangha","year":"2023","journal-title":"Circulation"},{"issue":"1","key":"2024040320063057200_ocae002-B13","doi-asserted-by":"crossref","first-page":"11","DOI":"10.1186\/s13244-020-00931-1","article-title":"Bringing AI to the clinic: blueprint for a vendor-neutral AI deployment infrastructure","volume":"12","author":"Leiner","year":"2021","journal-title":"Insights Imaging"},{"key":"2024040320063057200_ocae002-B14","author":"Chen","year":"2020"},{"key":"2024040320063057200_ocae002-B15","author":"Azizi","year":"2021"},{"key":"2024040320063057200_ocae002-B16","author":"Ciga","year":"2020"},{"key":"2024040320063057200_ocae002-B17","author":"Kiyasseh","year":"2020"},{"key":"2024040320063057200_ocae002-B18","author":"Gopal","year":"2021"},{"issue":"2","key":"2024040320063057200_ocae002-B19","doi-asserted-by":"crossref","first-page":"e1009862","DOI":"10.1371\/journal.pcbi.1009862","article-title":"Patient contrastive learning: a performant, expressive, and practical approach to electrocardiogram modeling","volume":"18","author":"Diamant","year":"2022","journal-title":"PLoS Comput Biol"},{"issue":"12","key":"2024040320063057200_ocae002-B20","doi-asserted-by":"crossref","first-page":"2011","DOI":"10.1093\/jamia\/ocaa088","article-title":"MINIMAR (MINimum Information for Medical AI Reporting): developing reporting standards for artificial intelligence in health care","volume":"27","author":"Hernandez-Boussard","year":"2020","journal-title":"J Am Med Inform Assoc"},{"issue":"6","key":"2024040320063057200_ocae002-B21","doi-asserted-by":"crossref","DOI":"10.3390\/s20061796","article-title":"ECG monitoring systems: review, architecture, processes, and key challenges","volume":"20","author":"Serhani","year":"2020","journal-title":"Sensors (Basel)"},{"issue":"8","key":"2024040320063057200_ocae002-B22","doi-asserted-by":"crossref","first-page":"977","DOI":"10.1161\/01.CIR.0000085166.44904.79","article-title":"Natural history of asymptomatic left ventricular systolic dysfunction in the community","volume":"108","author":"Wang","year":"2003","journal-title":"Circulation"},{"issue":"1","key":"2024040320063057200_ocae002-B23","doi-asserted-by":"crossref","first-page":"124","DOI":"10.1038\/s41746-023-00869-w","article-title":"Detection of left ventricular systolic dysfunction from single-lead electrocardiography adapted for portable and wearable devices","volume":"6","author":"Khunte","year":"2023","journal-title":"NPJ Digit Med"},{"issue":"1","key":"2024040320063057200_ocae002-B24","doi-asserted-by":"crossref","first-page":"154","DOI":"10.1038\/s41597-020-0495-6","article-title":"PTB-XL, a large publicly available electrocardiography dataset","volume":"7","author":"Wagner","year":"2020","journal-title":"Sci Data"},{"issue":"3","key":"2024040320063057200_ocae002-B25","doi-asserted-by":"crossref","first-page":"837","DOI":"10.2307\/2531595","article-title":"Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach","volume":"44","author":"DeLong","year":"1988","journal-title":"Biometrics"},{"issue":"6","key":"2024040320063057200_ocae002-B26","doi-asserted-by":"crossref","first-page":"756","DOI":"10.1038\/s41551-023-01049-7","article-title":"Robust and data-efficient generalization of self-supervised machine learning for diagnostic imaging","volume":"7","author":"Azizi","year":"2023","journal-title":"Nat Biomed Eng"}],"container-title":["Journal of the American Medical Informatics Association"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/jamia\/article-pdf\/31\/4\/855\/57148524\/ocae002.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/jamia\/article-pdf\/31\/4\/855\/57148524\/ocae002.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,4,3]],"date-time":"2024-04-03T16:07:04Z","timestamp":1712160424000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/jamia\/article\/31\/4\/855\/7588722"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,1,24]]},"references-count":26,"journal-issue":{"issue":"4","published-online":{"date-parts":[[2024,1,24]]},"published-print":{"date-parts":[[2024,4,3]]}},"URL":"https:\/\/doi.org\/10.1093\/jamia\/ocae002","relation":{"has-preprint":[{"id-type":"doi","id":"10.1101\/2023.09.13.23295494","asserted-by":"object"}]},"ISSN":["1067-5027","1527-974X"],"issn-type":[{"value":"1067-5027","type":"print"},{"value":"1527-974X","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2024,4,1]]},"published":{"date-parts":[[2024,1,24]]}}}