{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,3,8]],"date-time":"2026-03-08T03:25:12Z","timestamp":1772940312489,"version":"3.50.1"},"posted":{"date-parts":[[2026,3,4]]},"group-title":"Cardiovascular Medicine","reference-count":29,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,3,4]],"date-time":"2026-03-04T00:00:00Z","timestamp":1772582400000},"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":[[2026,3,4]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:sec>\n                  <jats:title>Background and Aims<\/jats:title>\n                  <jats:p>Accurate classification of mitral stenosis (MS) remains a significant clinical challenge. This study aimed to develop an artificial intelligence (AI) framework to automatically detect clinically significant MS from echocardiography.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Methods<\/jats:title>\n                  <jats:p>We developed EchoNet-MS, an open-source end-to-end integrated approach combining video based convolutional neural networks to assess MS severity and differentiate rheumatic etiology from echocardiography and validated its performance across four cohorts.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Results<\/jats:title>\n                  <jats:p>EchoNet-MS was trained and validated in total of 431,612 videos from 44,671 studies from three different healthcare system. Combining assessments from multiple echocardiographic videos, the model was trained on a Kaiser Permanente Northern California (KPNC) cohort of 8,677 studies from 7,576 patients with a range of MS severity. The model was validated on a KPNC held-out test cohort (N=1,623) and a temporally distinct cohort (N=19,206), as well as Stanford Healthcare (SHC) cohort (N=3,333) and Cedars-Sinai Medical Center (CSMC) cohort (N=72,909). EchoNet-MS achieved excellent discrimination of severe MS with AUC 0.937 [95% CI: 0.913 \u2013 0.958] in the KPNC held-out cohort, 0.994 [0.986 \u2013 0.999] in the temporally distinct cohort, 0.991 [0.986 \u2013 0.995] in SHC, and 0.973 [0.958 \u2013 0.987] in CSMC. The model achieved excellent performance in classifying both rheumatic or non-rheumatic MS with AUC ranging from 0.890 and 0.967.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Conclusions<\/jats:title>\n                  <jats:p>EchoNet-MS accurately assesses MS severity and etiology using information from multiple echocardiographic views. Its strong performance generalizes robustly to external cohorts and shows potential as an automated clinical decision support tool.<\/jats:p>\n                <\/jats:sec>","DOI":"10.64898\/2026.03.03.26347557","type":"posted-content","created":{"date-parts":[[2026,3,5]],"date-time":"2026-03-05T04:40:12Z","timestamp":1772685612000},"source":"Crossref","is-referenced-by-count":0,"title":["Automated Phenotyping of Mitral Stenosis Using Deep Learning"],"prefix":"10.64898","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-6252-6134","authenticated-orcid":false,"given":"Hirotaka","family":"Ieki","sequence":"first","affiliation":[{"name":"Department of Medicine, Division of Cardiovascular Medicine, Stanford University School of Medicine, Stanford, CA"},{"name":"Stanford Cardiovascular Institute, Stanford University, Stanford, CA"},{"name":"Division of Research, Kaiser Permanente Northern California, Pleasanton, CA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yuki","family":"Sahashi","sequence":"additional","affiliation":[{"name":"Department of Cardiology, Smidt Heart Institute, Cedars-Sinai Medical Center, Los Angeles, CA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Milo\u0161","family":"Vukadinovic","sequence":"additional","affiliation":[{"name":"Division of Research, Kaiser Permanente Northern California, Pleasanton, CA"},{"name":"Department of Bioengineering, University of California, Los Angeles, CA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Meenal","family":"Rawlani","sequence":"additional","affiliation":[{"name":"Department of Cardiology, Smidt Heart Institute, Cedars-Sinai Medical Center, Los Angeles, CA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Isabel","family":"Kim","sequence":"additional","affiliation":[{"name":"Department of Cardiology, Smidt Heart Institute, Cedars-Sinai Medical Center, Los Angeles, CA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4563-1071","authenticated-orcid":false,"given":"Andrew P.","family":"Ambrosy","sequence":"additional","affiliation":[{"name":"Division of Research, Kaiser Permanente Northern California, Pleasanton, CA"},{"name":"Department of Cardiology, Kaiser Permanente San Francisco Medical Center, San Francisco, CA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9109-0811","authenticated-orcid":false,"given":"Alan S.","family":"Go","sequence":"additional","affiliation":[{"name":"Division of Research, Kaiser Permanente Northern California, Pleasanton, CA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Bryan","family":"He","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Stanford University, Stanford, CA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3429-2702","authenticated-orcid":false,"given":"Paul","family":"Cheng","sequence":"additional","affiliation":[{"name":"Department of Medicine, Division of Cardiovascular Medicine, Stanford University School of Medicine, Stanford, CA"},{"name":"Stanford Cardiovascular Institute, Stanford University, Stanford, CA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"David","family":"Ouyang","sequence":"additional","affiliation":[{"name":"Division of Research, Kaiser Permanente Northern California, Pleasanton, CA"},{"name":"Department of Cardiology, Smidt Heart Institute, Cedars-Sinai Medical Center, Los Angeles, CA"},{"name":"Department of Cardiology, Kaiser Permanente Santa Clara Medical Center, Santa Clara, CA"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"54368","reference":[{"key":"2026030702150840000_2026.03.03.26347557v1.1","doi-asserted-by":"publisher","DOI":"10.1161\/CIR.0000000000000932"},{"key":"2026030702150840000_2026.03.03.26347557v1.2","doi-asserted-by":"crossref","first-page":"4635","DOI":"10.1093\/eurheartj\/ehaf194","article-title":"2025 ESC\/EACTS Guidelines for the management of valvular heart disease","volume":"46","year":"2025","journal-title":"Eur. Heart J"},{"key":"2026030702150840000_2026.03.03.26347557v1.3","doi-asserted-by":"publisher","DOI":"10.1161\/CIRCULATIONAHA.115.020185"},{"key":"2026030702150840000_2026.03.03.26347557v1.4","doi-asserted-by":"publisher","DOI":"10.1016\/j.echo.2008.11.029"},{"key":"2026030702150840000_2026.03.03.26347557v1.5","first-page":"e008122","article-title":"Interobserver variability in applying American Society of Echocardiography\/European Association of Cardiovascular Imaging 2016 guidelines for estimation of left ventricular filling pressure","volume":"12","year":"2019","journal-title":"Circ. Cardiovasc. Imaging"},{"key":"2026030702150840000_2026.03.03.26347557v1.6","doi-asserted-by":"publisher","DOI":"10.1016\/j.jacc.2025.10.025"},{"key":"2026030702150840000_2026.03.03.26347557v1.7","doi-asserted-by":"crossref","first-page":"2141","DOI":"10.1016\/j.jcin.2024.07.038","article-title":"Outcomes following transcatheter mitral valve replacement using dedicated devices in patients with mitral annular calcification","volume":"17","year":"2024","journal-title":"JACC Cardiovasc. Interv"},{"key":"2026030702150840000_2026.03.03.26347557v1.8","doi-asserted-by":"crossref","first-page":"923","DOI":"10.1161\/CIRCULATIONAHA.124.069047","article-title":"High-throughput deep learning detection of mitral regurgitation","volume":"150","year":"2024","journal-title":"Circulation"},{"key":"2026030702150840000_2026.03.03.26347557v1.9","doi-asserted-by":"publisher","DOI":"10.1101\/2025.03.18.25323918"},{"key":"2026030702150840000_2026.03.03.26347557v1.10","doi-asserted-by":"crossref","first-page":"595","DOI":"10.1001\/jamacardio.2025.0498","article-title":"Automated deep learning phenotyping of tricuspid regurgitation in echocardiography","volume":"10","year":"2025","journal-title":"JAMA Cardiol"},{"key":"2026030702150840000_2026.03.03.26347557v1.11","doi-asserted-by":"crossref","first-page":"2780","DOI":"10.1093\/eurheartj\/ehaf248","article-title":"Deep learning for echocardiographic assessment and risk stratification of aortic, mitral, and tricuspid regurgitation: the DELINEATE-regurgitation study","volume":"46","year":"2025","journal-title":"Eur. Heart J"},{"key":"2026030702150840000_2026.03.03.26347557v1.12","doi-asserted-by":"crossref","first-page":"116","DOI":"10.1016\/j.echo.2022.10.014","article-title":"Identifying aortic stenosis with a single parasternal long-axis video using deep learning","volume":"36","year":"2023","journal-title":"J. Am. Soc. Echocardiogr"},{"key":"2026030702150840000_2026.03.03.26347557v1.13","doi-asserted-by":"crossref","first-page":"4592","DOI":"10.1093\/eurheartj\/ehad456","article-title":"Severe aortic stenosis detection by deep learning applied to echocardiography","volume":"44","year":"2023","journal-title":"Eur. Heart J"},{"key":"2026030702150840000_2026.03.03.26347557v1.14","unstructured":"Vukadinovic, M. et al. Comprehensive echocardiogram evaluation with view primed vision language AI. Nature 1\u20133 (2025)."},{"key":"2026030702150840000_2026.03.03.26347557v1.15","doi-asserted-by":"publisher","DOI":"10.1001\/jama.2025.8731"},{"key":"2026030702150840000_2026.03.03.26347557v1.16","doi-asserted-by":"publisher","DOI":"10.1016\/j.echo.2017.02.009"},{"key":"2026030702150840000_2026.03.03.26347557v1.17","doi-asserted-by":"publisher","DOI":"10.1016\/j.echo.2018.06.004"},{"key":"2026030702150840000_2026.03.03.26347557v1.18","doi-asserted-by":"crossref","unstructured":"Tran, D. et al. A Closer Look at Spatiotemporal Convolutions for Action Recognition. in 2018 IEEE\/CVF Conference on Computer Vision and Pattern Recognition 6450\u20136459 (IEEE, 2018).","DOI":"10.1109\/CVPR.2018.00675"},{"key":"2026030702150840000_2026.03.03.26347557v1.19","first-page":"770","article-title":"Deep Residual Learning for Image Recognition","volume":"0","author":"Ren & Sun","year":"2016","journal-title":"2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)"},{"key":"2026030702150840000_2026.03.03.26347557v1.20","unstructured":"Pytorch-Lightning: Pretrain, Finetune ANY AI Model of ANY Size on Multiple GPUs, TPUs with Zero Code Changes. (Github)."},{"key":"2026030702150840000_2026.03.03.26347557v1.21","doi-asserted-by":"crossref","unstructured":"Defazio, A. et al. The Road Less Scheduled. in Advances in Neural Information Processing Systems (eds. Globerson, A. et al.) vol. 37 9974\u201310007 (Curran Associates, Inc., 2024).","DOI":"10.52202\/079017-0320"},{"key":"2026030702150840000_2026.03.03.26347557v1.22","unstructured":"Micikevicius, P. , et al. Mixed Precision Training. arXiv [cs.AI] (2017)."},{"key":"2026030702150840000_2026.03.03.26347557v1.23","doi-asserted-by":"publisher","DOI":"10.1016\/j.cpc.2010.04.018"},{"key":"2026030702150840000_2026.03.03.26347557v1.24","unstructured":"Smilkov, D. , Thorat, N. , Kim, B. , Vi\u00e9gas, F. & Wattenberg, M . SmoothGrad: removing noise by adding noise. arXiv [cs.LG] (2017)."},{"key":"2026030702150840000_2026.03.03.26347557v1.25","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-020-2145-8"},{"key":"2026030702150840000_2026.03.03.26347557v1.26","doi-asserted-by":"crossref","first-page":"520","DOI":"10.1038\/s41586-023-05947-3","article-title":"Blinded, randomized trial of sonographer versus AI cardiac function assessment","volume":"616","year":"2023","journal-title":"Nature"},{"key":"2026030702150840000_2026.03.03.26347557v1.27","doi-asserted-by":"crossref","first-page":"386","DOI":"10.1001\/jamacardio.2021.6059","article-title":"High-throughput precision phenotyping of left ventricular hypertrophy with cardiovascular deep learning","volume":"7","year":"2022","journal-title":"JAMA Cardiol"},{"key":"2026030702150840000_2026.03.03.26347557v1.28","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-021-01335-4"},{"key":"2026030702150840000_2026.03.03.26347557v1.29","doi-asserted-by":"publisher","DOI":"10.1101\/2025.08.20.25334115"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.03.03.26347557","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,7]],"date-time":"2026-03-07T10:15:21Z","timestamp":1772878521000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.64898\/2026.03.03.26347557"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,4]]},"references-count":29,"URL":"https:\/\/doi.org\/10.64898\/2026.03.03.26347557","relation":{},"subject":[],"published":{"date-parts":[[2026,3,4]]},"subtype":"preprint"}}