{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,6,2]],"date-time":"2026-06-02T03:47:33Z","timestamp":1780372053872,"version":"3.54.1"},"posted":{"date-parts":[[2024,10,7]]},"group-title":"Cardiovascular Medicine","reference-count":44,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2024,10,7]],"date-time":"2024-10-07T00:00:00Z","timestamp":1728259200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2024,12,27]]},"abstract":"<jats:title>ABSTRACT<\/jats:title>\n                <jats:sec>\n                  <jats:title>Background<\/jats:title>\n                  <jats:p>Identifying structural heart diseases (SHDs) early can change the course of the disease, but their diagnosis requires cardiac imaging, which is limited in accessibility.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Objective<\/jats:title>\n                  <jats:p>To leverage images of 12-lead ECGs for automated detection and prediction of multiple SHDs using an ensemble deep learning approach.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Methods<\/jats:title>\n                  <jats:p>We developed a series of convolutional neural network models for detecting a range of individual SHDs from images of ECGs with SHDs defined by transthoracic echocardiograms (TTEs) performed within 30 days of the ECG at the Yale New Haven Hospital (YNHH). SHDs were defined as LV ejection fraction &lt;40%, moderate-to-severe left-sided valvular disease (aortic\/mitral stenosis or regurgitation), or severe left ventricular hypertrophy (IVSd &gt; 1.5cm and diastolic dysfunction). We developed an ensemble XGBoost model, PRESENT-SHD, as a composite screen across all SHDs. We validated PRESENT-SHD at 4 US hospitals and the prospective, population-based Brazilian Longitudinal Study of Adult Health (ELSA-Brasil), with concurrent protocolized ECGs and TTEs. We also used PRESENT-SHD for risk stratification of new-onset SHD or heart failure (HF) in clinical cohorts and the population-based UK Biobank (UKB).<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Results<\/jats:title>\n                  <jats:p>The models were developed using 261,228 ECGs from 93,693 YNHH patients and evaluated on a single ECG from 11,023 individuals at YNHH (19% with SHD), 44,591 across external hospitals (20-27% with SHD), and 3,014 in the ELSA-Brasil (3% with SHD). In the held-out test set, PRESENT-SHD demonstrated an AUROC of 0.886 (0.877-894), 90% sensitivity, and 66% specificity. At hospital-based sites, PRESENT-SHD had AUROCs ranging from 0.854-0.900, with sensitivities and specificities of 93-96% and 51-56%, respectively. The model generalized well to ELSA-Brasil (AUROC, 0.853 [0.811-0.897], 88% sensitivity, 62% specificity). PRESENT-SHD demonstrated consistent performance across demographic subgroups, novel ECG formats, and smartphone photographs of ECGs from monitors and printouts. A positive PRESENT-SHD screen portended a 2- to 4-fold higher risk of new-onset SHD\/HF, independent of demographics, comorbidities, and the competing risk of death across clinical sites and UKB, with high predictive discrimination.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Conclusion<\/jats:title>\n                  <jats:p>We developed and validated PRESENT-SHD, an AI-ECG tool identifying a range of SHD using images of 12-lead ECGs, representing a robust, scalable, and accessible modality for automated SHD screening and risk stratification.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>CONDENSED ABSTRACT<\/jats:title>\n                  <jats:p>Screening for structural heart disorders (SHDs) requires cardiac imaging, which has limited accessibility. To leverage 12-lead ECG images for automated detection and prediction of multiple SHDs, we developed PRESENT-SHD, an ensemble deep learning model. PRESENT-SHD demonstrated excellent performance in detecting SHDs across 5 US hospitals and a population-based cohort in Brazil. The model successfully predicted the risk of new-onset SHD or heart failure in both US clinical cohorts and the community-based UK Biobank. By using ubiquitous ECG images and smartphone photographs to predict a composite outcome of multiple SHDs, PRESENT-SHD establishes a scalable paradigm for cardiovascular screening and risk stratification.<\/jats:p>\n                <\/jats:sec>","DOI":"10.1101\/2024.10.06.24314939","type":"posted-content","created":{"date-parts":[[2024,10,7]],"date-time":"2024-10-07T18:50:14Z","timestamp":1728327014000},"source":"Crossref","is-referenced-by-count":3,"title":["An Ensemble Deep Learning Algorithm for Structural Heart Disease Screening Using Electrocardiographic Images: PRESENT SHD"],"prefix":"10.64898","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-5664-4126","authenticated-orcid":false,"given":"Lovedeep S","family":"Dhingra","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3197-2657","authenticated-orcid":false,"given":"Arya","family":"Aminorroaya","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8524-1203","authenticated-orcid":false,"given":"Veer","family":"Sangha","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1876-8304","authenticated-orcid":false,"given":"Aline F","family":"Pedroso","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-3034-0641","authenticated-orcid":false,"given":"Sumukh Vasisht","family":"Shankar","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5243-552X","authenticated-orcid":false,"given":"Andreas","family":"Coppi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2914-4406","authenticated-orcid":false,"given":"Murilo","family":"Foppa","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7317-1367","authenticated-orcid":false,"given":"Luisa CC","family":"Brant","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sandhi M","family":"Barreto","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Antonio Luiz P","family":"Ribeiro","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2046-127X","authenticated-orcid":false,"given":"Harlan M","family":"Krumholz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4362-0720","authenticated-orcid":false,"given":"Evangelos K","family":"Oikonomou","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9467-6199","authenticated-orcid":false,"given":"Rohan","family":"Khera","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2024122919473634000_2024.10.06.24314939v3.1","doi-asserted-by":"publisher","DOI":"10.1093\/eurheartj\/suq012"},{"key":"2024122919473634000_2024.10.06.24314939v3.2","doi-asserted-by":"crossref","unstructured":"Picano E . Economic and biological costs of cardiac imaging. Cardiovasc Ultrasound. 2005;3.","DOI":"10.1186\/1476-7120-3-13"},{"key":"2024122919473634000_2024.10.06.24314939v3.3","doi-asserted-by":"publisher","DOI":"10.1007\/s12350-009-9104-y"},{"key":"2024122919473634000_2024.10.06.24314939v3.4","doi-asserted-by":"crossref","unstructured":"Alkhouli M , Alqahtani F , Holmes DR , Berzingi C . Racial disparities in the utilization and outcomes of structural heart disease interventions in the United States. J Am Heart Assoc. 2019;8.","DOI":"10.1161\/JAHA.119.012125"},{"key":"2024122919473634000_2024.10.06.24314939v3.5","doi-asserted-by":"crossref","unstructured":"Samad Z , Sivak JA , Phelan M , Schulte PJ , Patel U , Velazquez EJ . Prevalence and outcomes of left-sided valvular heart disease associated with Chronic kidney disease. J Am Heart Assoc. 2017;6.","DOI":"10.1161\/JAHA.117.006044"},{"key":"2024122919473634000_2024.10.06.24314939v3.6","doi-asserted-by":"crossref","first-page":"980","DOI":"10.1016\/j.cjca.2021.02.003","article-title":"Sex and race differences in the pathophysiology, diagnosis, treatment, and outcomes of valvular heart diseases","volume":"37","year":"2021","journal-title":"Can J Cardiol"},{"key":"2024122919473634000_2024.10.06.24314939v3.7","doi-asserted-by":"crossref","first-page":"4349","DOI":"10.1093\/eurheartj\/ehaa485","article-title":"Timing of intervention in asymptomatic patients with valvular heart disease","volume":"41","year":"2020","journal-title":"Eur Heart J"},{"key":"2024122919473634000_2024.10.06.24314939v3.8","doi-asserted-by":"publisher","DOI":"10.1161\/01.CIR.0000085166.44904.79"},{"key":"2024122919473634000_2024.10.06.24314939v3.9","doi-asserted-by":"publisher","DOI":"10.1093\/eurheartj\/ehi559"},{"key":"2024122919473634000_2024.10.06.24314939v3.10","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-018-0240-2"},{"key":"2024122919473634000_2024.10.06.24314939v3.11","doi-asserted-by":"publisher","DOI":"10.1161\/CIRCULATIONAHA.122.062646"},{"key":"2024122919473634000_2024.10.06.24314939v3.12","doi-asserted-by":"crossref","first-page":"2885","DOI":"10.1093\/eurheartj\/ehab153","article-title":"Electrocardiogram screening for aortic valve stenosis using artificial intelligence","volume":"42","year":"2021","journal-title":"Eur Heart J"},{"key":"2024122919473634000_2024.10.06.24314939v3.13","first-page":"419","article-title":"Artificial intelligence enabled ECG screening for left ventricular systolic dysfunction: a systematic review","volume":"28","year":"2023","journal-title":"Heart Fail Rev"},{"key":"2024122919473634000_2024.10.06.24314939v3.14","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","year":"2023","journal-title":"NPJ Digit Med"},{"key":"2024122919473634000_2024.10.06.24314939v3.15","doi-asserted-by":"crossref","unstructured":"Oikonomou EK , Sangha V , Dhingra LS , et al. Artificial intelligence-enhanced risk stratification of cancer therapeutics-related cardiac dysfunction using electrocardiographic images. bioRxiv. 2024:2024.03.12.24304047.","DOI":"10.1101\/2024.03.12.24304047"},{"key":"2024122919473634000_2024.10.06.24314939v3.16","doi-asserted-by":"crossref","unstructured":"Sangha V , Dhingra LS , Oikonomou EK , et al. Identification of hypertrophic cardiomyopathy on electrocardiographic images with deep learning. medRxiv. 2023:2023.12.23.23300490.","DOI":"10.1101\/2023.12.23.23300490"},{"key":"2024122919473634000_2024.10.06.24314939v3.17","doi-asserted-by":"crossref","unstructured":"Dhingra LS , Sangha V , Aminorroaya A , et al. A multicenter evaluation of the impact of procedural and pharmacological interventions on deep learning-based electrocardiographic markers of hypertrophic cardiomyopathy. bioRxiv. 2024.","DOI":"10.1101\/2024.01.15.24301011"},{"key":"2024122919473634000_2024.10.06.24314939v3.18","unstructured":"Dhingra LS , Aminorroaya A , Camargos AP , et al. Using artificial intelligence to predict heart failure risk from single-lead electrocardiographic signals: A multinational assessment. bioRxiv. 2024:2024.05.27.24307952."},{"key":"2024122919473634000_2024.10.06.24314939v3.19","doi-asserted-by":"publisher","DOI":"10.1161\/CIRCULATIONAHA.121.057869"},{"key":"2024122919473634000_2024.10.06.24314939v3.20","doi-asserted-by":"publisher","DOI":"10.1016\/j.echo.2018.06.004"},{"key":"2024122919473634000_2024.10.06.24314939v3.21","doi-asserted-by":"publisher","DOI":"10.1016\/j.echo.2017.01.007"},{"key":"2024122919473634000_2024.10.06.24314939v3.22","doi-asserted-by":"crossref","first-page":"58","DOI":"10.4103\/2543-1463.282191","article-title":"Recommendations for noninvasive evaluation of native valvular regurgitation","volume":"4","year":"2020","journal-title":"J Indian Acad Echocardiogr Cardiovasc Imaging"},{"key":"2024122919473634000_2024.10.06.24314939v3.23","doi-asserted-by":"crossref","first-page":"317","DOI":"10.1177\/8756479319836781","article-title":"Diastolic dysfunction and the new recommendations for echocardiographic assessment of left ventricular diastolic function: Summary of guidelines and novelties in diagnosis and grading","volume":"35","year":"2019","journal-title":"J Diagn Med Sonogr"},{"key":"2024122919473634000_2024.10.06.24314939v3.24","unstructured":"Anon. Pillow. PyPI Accessed February 29, 2024. https:\/\/pypi.org\/project\/pillow\/."},{"key":"2024122919473634000_2024.10.06.24314939v3.25","unstructured":"Anon. Mingxing Tan and Quoc V Le . EfficientNet: Rethinking model scaling for convolutional neural networks. International Conference on Machine Learning, 2019"},{"key":"2024122919473634000_2024.10.06.24314939v3.26","doi-asserted-by":"publisher","DOI":"10.1093\/jamia\/ocae002"},{"key":"2024122919473634000_2024.10.06.24314939v3.27","doi-asserted-by":"publisher","DOI":"10.1002\/sim.7501"},{"key":"2024122919473634000_2024.10.06.24314939v3.28","doi-asserted-by":"publisher","DOI":"10.1002\/sim.4154"},{"key":"2024122919473634000_2024.10.06.24314939v3.29","doi-asserted-by":"publisher","DOI":"10.1002\/sim.6370"},{"key":"2024122919473634000_2024.10.06.24314939v3.30","doi-asserted-by":"publisher","DOI":"10.1136\/bmj.q902"},{"key":"2024122919473634000_2024.10.06.24314939v3.31","doi-asserted-by":"publisher","DOI":"10.1016\/j.jacc.2019.12.030"},{"key":"2024122919473634000_2024.10.06.24314939v3.32","doi-asserted-by":"crossref","first-page":"755","DOI":"10.1161\/CIRCULATIONAHA.121.058696","article-title":"Multinational Federated Learning Approach to Train ECG and Echocardiogram Models for Hypertrophic Cardiomyopathy Detection","volume":"146","year":"2022","journal-title":"Circulation"},{"key":"2024122919473634000_2024.10.06.24314939v3.33","doi-asserted-by":"publisher","DOI":"10.1016\/j.mayocp.2021.04.023"},{"key":"2024122919473634000_2024.10.06.24314939v3.34","doi-asserted-by":"crossref","unstructured":"Taborsky M , Aiglova R , Fedorco M , et al. Detection of arrhythmias in patients with cardiac amyloidosis using implantable ECG recorders. Eur Heart J. 2022;43.","DOI":"10.1093\/eurheartj\/ehac544.1792"},{"key":"2024122919473634000_2024.10.06.24314939v3.35","doi-asserted-by":"publisher","DOI":"10.1101\/2023.04.25.23289130"},{"key":"2024122919473634000_2024.10.06.24314939v3.36","doi-asserted-by":"publisher","DOI":"10.1016\/j.jcmg.2021.08.004"},{"key":"2024122919473634000_2024.10.06.24314939v3.37","doi-asserted-by":"crossref","unstructured":"Oikonomou EK , Sangha V , Shankar SV , et al. Tracking the pre-clinical progression of transthyretin amyloid cardiomyopathy using artificial intelligence-enabled electrocardiography and echocardiography. bioRxiv. 2024:2024.08.25.24312556.","DOI":"10.1101\/2024.08.25.24312556"},{"key":"2024122919473634000_2024.10.06.24314939v3.38","doi-asserted-by":"crossref","first-page":"557","DOI":"10.1161\/CIRCULATIONAHA.123.065717","article-title":"Cardiovascular health care in low- and middle-income countries","volume":"149","year":"2024","journal-title":"Circulation"},{"key":"2024122919473634000_2024.10.06.24314939v3.39","doi-asserted-by":"publisher","DOI":"10.1161\/CIRCULATIONAHA.105.541664"},{"key":"2024122919473634000_2024.10.06.24314939v3.40","doi-asserted-by":"crossref","first-page":"897","DOI":"10.2459\/JCM.0000000000001110","article-title":"Early intervention or watchful waiting for asymptomatic severe aortic valve stenosis: a systematic review and meta-analysis","volume":"21","year":"2020","journal-title":"J Cardiovasc Med (Hagerstown)"},{"key":"2024122919473634000_2024.10.06.24314939v3.41","doi-asserted-by":"publisher","DOI":"10.1016\/S0272-6386(96)90357-1"},{"key":"2024122919473634000_2024.10.06.24314939v3.42","unstructured":"Topol E . Opportunistic A.i. for medical scans. Ground Truths. 2024. Accessed July 28, 2024. https:\/\/erictopol.substack.com\/p\/opportunistic-ai-for-medical-scans."},{"key":"2024122919473634000_2024.10.06.24314939v3.43","doi-asserted-by":"crossref","first-page":"217","DOI":"10.1097\/00005082-200605000-00010","article-title":"Opportunistic screening for cardiovascular problems in rural and remote health settings","volume":"21","year":"2006","journal-title":"J Cardiovasc Nurs"},{"key":"2024122919473634000_2024.10.06.24314939v3.44","first-page":"e13686","article-title":"Impact of incorporating echocardiographic screening into a clinical prediction model to optimise utilisation of echocardiography in primary care","volume":"75","year":"2021","journal-title":"Int J Clin Pract"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1101\/2024.10.06.24314939","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,15]],"date-time":"2026-01-15T20:14:05Z","timestamp":1768508045000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.1101\/2024.10.06.24314939"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,10,7]]},"references-count":44,"URL":"https:\/\/doi.org\/10.1101\/2024.10.06.24314939","relation":{},"subject":[],"published":{"date-parts":[[2024,10,7]]},"subtype":"preprint"}}