CoChat
Sign Up
Cited by 30ECG Monitoring and AnalysisMachine Learning in HealthcareArtificial Intelligence in Healthcare

Multi-center retrospective cohort study applying deep learning to electrocardiograms to identify left heart valvular dysfunction

Open access
Vaid, Akhil, Argulian, Edgar, Lerakis, Stamatios, Beaulieu-Jones, Brett K., Krittanawong, Chayakrit, Klang, Eyal, Lampert, Joshua, Reddy, Vivek Y. +3 more·2023·Communications Medicine
DOI 10.1038/s43856-023-00240-wPMC PMC9929085
Open this paper in CoChat — free accountOpen PDF

Abstract

Background: Aortic Stenosis and Mitral Regurgitation are common valvular conditions representing a hidden burden of disease within the population. The aim of this study was to develop and validate deep learning-based screening and diagnostic tools that can help guide clinical decision making. Methods: In this multi-center retrospective cohort study, we acquired Transthoracic Echocardiogram reports from five Mount Sinai hospitals within New York City representing a demographically diverse cohort of patients. We developed a Natural Language Processing pipeline to extract ground-truth labels about valvular status and paired these to Electrocardiograms (ECGs). We developed and externally validated deep learning models capable of detecting valvular disease, in addition to considering scenarios of clinical deployment. Results: We use 617,338 ECGs paired to transthoracic echocardiograms from 123,096 patients to develop a deep learning model for detection of Mitral Regurgitation. Area Under Receiver Operating Characteristic curve (AUROC) is 0.88 (95% CI:0.88–0.89) in internal testing, and 0.81 (95% CI:0.80–0.82) in external validation. To develop a model for detection of Aortic Stenosis, we use 617,338 Echo-ECG pairs for 128,628 patients. AUROC is 0.89 (95% CI: 0.88-0.89) in internal testing, going to 0.86 (95% CI: 0.85-0.87) in external validation. The model’s performance increases leading up to the time of the diagnostic echo, and it performs well in validation against requirement of Transcatheter Aortic Valve Replacement procedures. Conclusions: Deep learning based tools can increase the amount of information extracted from ubiquitous investigations such as the ECG. Such tools are inexpensive, can help in earlier disease detection, and potentially improve prognosis.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s43856-023-00240-w

Cites

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Valvular and composite SHD AI-ECG models are promising for referral enrichment.

    “Valvular and composite SHD models are promising for referral enrichment”

    From AI-ECG for Echocardiography Triage in Structural Heart Disease: Evidence, Implementation, and Future Directions (2026) · cites this paper as [22]

    The Discussion (page unavailable) says the valvular AI-ECG models could screen patients or direct them along appropriate care pathways, supporting their potential for referral enrichment. The paper evaluates aortic stenosis and mitral regurgitation models, but the cited passages do not establish a composite structural heart disease (SHD) model. (LLM, conf=0.87)

    Experimental — accuracy not yet measured

  • Complementary multicentre work from five Mount Sinai hospitals using 617,338 ECG-echocardiogram pairs showed that ECG-based deep learning can detect both MR and AS, with external-validation AUROCs of 0.81 for MR and 0.86 for AS.

    “Complementary multicentre work from five Mount Sinai hospitals using 617,338 ECG–echocardiogram pairs likewise showed that ECG-based deep learning can detect both MR and AS, with external-validation AUROCs of 0.81 for MR and 0.86 for AS.22”

    From AI-ECG for Echocardiography Triage in Structural Heart Disease: Evidence, Implementation, and Future Directions (2026) · cites this paper as [22]

  • AI-ECG has a broader role for left-sided valvular dysfunction.

    “Vaid et al. applied deep learning to more than 600,000 ECG-echocardiogram pairs across five hospitals and externally validated models for both mitral regurgitation and aortic stenosis, supporting the broader role of AI-ECG for left-sided valvular dysfunction [31].”

    From Deep Learning Applied to 12-Lead ECGs for Detection of Structural, Metabolic and Systemic Disease (2026) · cites this paper as [31]

  • Deep learning applied to more than 600,000 ECG-echocardiogram pairs across five hospitals externally validated models for both mitral regurgitation and aortic stenosis.

    “Vaid et al. applied deep learning to more than 600,000 ECG-echocardiogram pairs across five hospitals and externally validated models for both mitral regurgitation and aortic stenosis, supporting the broader role of AI-ECG for left-sided valvular dysfunction [31].”

    From Deep Learning Applied to 12-Lead ECGs for Detection of Structural, Metabolic and Systemic Disease (2026) · cites this paper as [31]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

No citation lists for this paper

No citation lists for this paper yet — checked again daily.

References 29 · Cited by 30

Research works printed but not linked: 28

  1. [1]Iung B Vahanian A Epidemiology of valvular heart disease in the adult Nat. Rev. Cardiol. 2011 8 162 172 10.1038/nrcardio.2010.202 21263455 · DOI
  2. [2]Montant P Long-term survival in asymptomatic patients with severe degenerative mitral regurgitation: a propensity score-based comparison between an early surgical strategy and a conservative treatment approach J. Thorac. Cardiovasc. Surg. 2009 138 1339 1348 10.1016/j.jtcvs.2009.03.046 19660385 · DOI
  3. [3]Travis B Partho PS Jagat N Is TAVR ready for the global aging population? Global Heart 2017 12 291 299 10.1016/j.gheart.2017.02.002 28433492 · DOI
  4. [4]Kundi H Trends in isolated surgical aortic valve replacement according to hospital-based transcatheter aortic valve replacement volumes JACC Cardiovasc. Interv. 2018 11 2148 2156 10.1016/j.jcin.2018.07.002 30343022 · DOI
  5. [5]Coleman W Weidman-Evans E Clawson R Diagnosing and managing mitral regurgitation J. Am. Acad. of PAs 2017 30 11 14 28452853
  6. [6]Grimard BH Safford RE Burns EL Aortic stenosis: diagnosis and treatment Am. Fam. Physician 2016 93 371 378 26926974
  7. [7]Carapetis JR Evaluation of a screening protocol using auscultation and portable echocardiography to detect asymptomatic rheumatic heart disease in Tongan schoolchildren Nat. Clin. Pract. Cardiovasc. Med. 2008 5 411 417 10.1038/ncpcardio1185 18398402 · DOI
  8. [8]Kattoor AJ Shanbhag A Abraham A Vallurupalli S Clinical context and detection of the murmur of advanced aortic stenosis South Med. J. 2018 111 230 234 10.14423/SMJ.0000000000000791 29719036 · DOI
  9. [9]Kasper I Effect of teaching and type of stethoscope on cardiac auscultatory performance Am. Heart J. 2006 152 85.e81 85.e87 16824835
  10. [10]Jellis CL Griffin BP Are we doing too many inpatient echocardiograms? The answer from big data may surprise you! J. Am. Coll. Cardiol. 2016 67 512 514 10.1016/j.jacc.2015.10.091 26846949 · DOI
  11. [11]Strange G Poor long-term survival in patients with moderate aortic stenosis J. Am. Coll. Cardiol. 2019 74 1851 1863 10.1016/j.jacc.2019.08.004 31491546 · DOI · Paper page
  12. [13]Bos JM Use of artificial intelligence and deep neural networks in evaluation of patients with electrocardiographically concealed long QT syndrome from the surface 12-lead electrocardiogram JAMA Cardiol. 2021 6 532 538 10.1001/jamacardio.2020.7422 33566059 PMC7876623 · DOI
  13. [14]Christopoulos G Artificial intelligence-electrocardiography to predict incident atrial fibrillation Circ. Arrhythmia Electrophysiol. 2020 13 e009355 10.1161/CIRCEP.120.009355 PMC8127001 33185118 · DOI
  14. [15]Hannun AY Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network Nat. Med. 2019 25 65 69 10.1038/s41591-018-0268-3 30617320 PMC6784839 · DOI · Paper page
  15. [16]Kagiyama N Machine learning assessment of left ventricular diastolic function based on electrocardiographic features J. Am. Coll. Cardiol. 2020 76 930 941 10.1016/j.jacc.2020.06.061 32819467 · DOI
  16. [17]Vaid A, et al. Using Deep-Learning Algorithms to Simultaneously Identify Right and Left Ventricular Dysfunction From the Electrocardiogram. JACC: Cardiovascular Imaging. 15, 395–410 (2022). PMC8917975 34656465
  17. [18]Lai D Zhang Y Zhang X Su Y Heyat MBB An automated strategy for early risk identification of sudden cardiac death by using machine learning approach on measurable arrhythmic risk markers IEEE Access 2019 7 94701 94716 10.1109/ACCESS.2019.2925847 · DOI
  18. [19]Kwon JM Deep learning-based algorithm for detecting aortic stenosis using electrocardiography J. Am. Heart Assoc. 2020 9 e014717 10.1161/JAHA.119.014717 32200712 PMC7428650 · DOI · Paper page
  19. [20]Hata, E., et al. Classification of aortic stenosis using ECG by deep learning and its analysis using grad-CAM. In: Paper presented at: 2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), (EMBC, 2020). 33018287
  20. [21]Kwon JM Artificial intelligence for detecting mitral regurgitation using electrocardiography J. Electrocardiol. 2020 59 151 157 10.1016/j.jelectrocard.2020.02.008 32146201 · DOI
  21. [22]Cohen-Shelly M Electrocardiogram screening for aortic valve stenosis using artificial intelligence Eur. Heart J. 2021 42 2885 2896 10.1093/eurheartj/ehab153 33748852 · DOI · Paper page
  22. [23]Malmivuo, J., Plonsey, R. Bioelectromagnetism. 15. 12-Lead ECG System. In: Bioelectromagnetism - principles and applications of bioelectric and biomagnetis fields 277–289 (Oxford University Press, 1975).
  23. [24]Sörnmo, L., Laguna, P. ECG signal processing. In: Bioelectrical signal processing in cardiac and neurological applications (eds. Sörnmo, L., Laguna, P.,) 453–566 (Academic Press, Burlington, 2005).
  24. [25]Tan, M., Le, Q.V. EfficientNet: rethinking model scaling for convolutional neural networks. In: Proceedings of the 36th International Conference on Machine Learning (eds Kamalika C., & Ruslan S.,) 97, 6105–6114 (PMLR, 2019).
  25. [26]Huang S-C Pareek A Seyyedi S Banerjee I Lungren MP Fusion of medical imaging and electronic health records using deep learning: a systematic review and implementation guidelines npj Digital Med. 2020 3 136 10.1038/s41746-020-00341-z PMC7567861 33083571 · DOI · Paper page
  26. [27]Ruopp MD Perkins NJ Whitcomb BW Schisterman EF Youden Index and optimal cut-point estimated from observations affected by a lower limit of detection Biom. J. 2008 50 419 430 10.1002/bimj.200710415 18435502 PMC2515362 · DOI · Paper page
  27. [28]Chubak J Pocobelli G Weiss NS Tradeoffs between accuracy measures for electronic health care data algorithms J. Clin. Epidemiol. 2012 65 343 349.e342 10.1016/j.jclinepi.2011.09.002 22197520 PMC3264740 · DOI
  28. [29]Tajbakhsh N Convolutional neural networks for medical image analysis: full training or fine tuning? IEEE Trans. Med. Imaging 2016 35 1299 1312 10.1109/TMI.2016.2535302 26978662 · DOI · Paper page

Authors
Vaid, A., Argulian, E., Lerakis, S., Beaulieu-Jones, B. K., Krittanawong, C., Klang, E., Lampert, J., Reddy, V. Y., Narula, J., Nadkarni, G. N., & Glicksberg, B. S.
Affiliations
  • Icahn School of Medicine at Mount Sinai (US)
  • NYU Langone Health (US)
  • Tel Aviv University (IL)
Corresponding author
Glicksberg, Benjamin S.
ORCID
Venue
Communications Medicine
Publisher
Springer Science and Business Media LLC
Published
2023-02-14
Volume
3
Issue
1
Pages
24-24
ISSN
2730-664X (ISSN-L)
DOI
10.1038/s43856-023-00240-w
PMC
PMC9929085
OpenAlex
W4320725812
Cited by
30
References
29
Citations by year
2026: 10 · 2025: 10 · 2024: 4 · 2023: 6
FWCI
5.89
Type
journal article
Language
en
Open access
Licence
Locations
  • Communications Medicine (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
U.S. Department of Health & Human Services | NIH | National Center for Advancing Translational Sciences (U54 TR001433-05)
Topics
  • Health Sciences › Medicine › Cardiology and Cardiovascular Medicine › ECG Monitoring and Analysis
  • Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning in Healthcare
  • Health Sciences › Health Professions › Health Information Management › Artificial Intelligence in Healthcare
Keywords
aortic stenosis, mitral regurgitation, electrocardiogram, deep learning, transthoracic echocardiography, natural language processing, transcatheter aortic valve implantation, external validation, disease screening

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review