Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Deep-learning-assisted analysis of echocardiographic videos improves predictions of all-cause mortality


Machine learning promises to assist physicians with predictions of mortality and of other future clinical events by learning complex patterns from historical data, such as longitudinal electronic health records. Here we show that a convolutional neural network trained on raw pixel data in 812,278 echocardiographic videos from 34,362 individuals provides superior predictions of one-year all-cause mortality. The model’s predictions outperformed the widely used pooled cohort equations, the Seattle Heart Failure score (measured in an independent dataset of 2,404 patients with heart failure who underwent 3,384 echocardiograms), and a machine learning model involving 58 human-derived variables from echocardiograms and 100 clinical variables derived from electronic health records. We also show that cardiologists assisted by the model substantially improved the sensitivity of their predictions of one-year all-cause mortality by 13% while maintaining prediction specificity. Large unstructured datasets may enable deep learning to improve a wide range of clinical prediction models.

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

Access options

Rent or buy this article

Prices vary by article type



Prices may be subject to local taxes which are calculated during checkout

Fig. 1: AUC across five folds for individual and collective echocardiography views.
Fig. 2: Survey results showing the performance estimates for the four cardiologists and the DNN model.
Fig. 3: Performance comparison with the SHF score.
Fig. 4: Occlusion results for the lowest and highest prediction risk scores for 1-yr mortality.

Data availability

The main data supporting the results in this study are available within the paper and its Supplementary Information. All requests for raw and analysed data will be reviewed by the Legal Department of Geisinger Clinic to verify whether the request is subject to any intellectual property or confidentiality constraints. Requests for patient-related data not included in the paper will not be considered. Any data that can be shared will be released via a Material Transfer Agreement for non-commercial research purposes.

Code availability

All requests for code will be reviewed by the Legal Department of Geisinger Clinic to verify whether the request is subject to any intellectual property or confidentiality constraints. Any code that can be shared will be released via a Material Transfer Agreement for non-commercial research purposes under the Creative Commons Attribution NonCommercial–NoDerivatives 4.0 license. Code to reproduce Supplementary Fig. 10 is available at:


  1. Payne, J. W. Task complexity and contingent processing in decision making: an information search and protocol analysis. Organ. Behav. Hum. Perform. 16, 366–387 (1976).

    Article  Google Scholar 

  2. Quer, G., Muse, E. D., Nikzad, N., Topol, E. J. & Steinhubl, S. R. Augmenting diagnostic vision with AI. Lancet 390, 221 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

  3. Jha, S. & Topol, E. J. Adapting to artificial intelligence: radiologists and pathologists as information specialists. JAMA 316, 2353–2354 (2016).

    Article  PubMed  Google Scholar 

  4. Kyriacou, E., Constantinides, A., Pattichis, C., Pattichis, M. & Panayides, A. in Biomedical Signals, Imaging, and Informatics 4th edn (eds Bronzino, J. D. & Peterson, D.) Ch. 64 (CRC Press, 2015).

  5. Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I. & Salakhutdinov, R. Dropout: a simple way to prevent neural networks from overfitting. J. Mach. Learn. Res. 15, 1929–1958 (2014).

    Google Scholar 

  6. LeCun, Y., Bengio, Y. & Hinton, G. Deep learning. Nature 521, 436–444 (2015).

    Article  CAS  PubMed  Google Scholar 

  7. Ji, S., Xu, W., Yang, M. & Yu, K. 3D convolutional neural networks for human action recognition. IEEE Trans. Pattern Anal. Mach. Intell. 35, 221–231 (2012).

    Article  Google Scholar 

  8. Karpathy, A. et al. Large-scale video classification with convolutional neural networks. In IEEE conference on Computer Vision and Pattern Recognition 1725–1732 (2014).

  9. Gulshan, V. et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. JAMA 316, 2402–2410 (2016).

    Article  PubMed  Google Scholar 

  10. Esteva, A. et al. Dermatologist-level classification of skin cancer with deep neural networks. Nature 542, 115–118 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Setio, A. A. A. et al. Pulmonary nodule detection in ct images: false positive reduction using multi-view convolutional networks. IEEE Trans. Med. Imaging 35, 1160–1169 (2016).

    Article  PubMed  Google Scholar 

  12. Arbabshirani, M. R. et al. Advanced machine learning in action: identification of intracranial hemorrhage on computed tomography scans of the head with clinical workflow integration. npj Digit. Med. 1, 9 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

  13. Dou, Q. et al. Automatic detection of cerebral microbleeds from MR images via 3D convolutional neural networks. IEEE Trans. Med. Imaging 35, 1182–1195 (2016).

    Article  Google Scholar 

  14. Madani, A., Ong, J. R., Tibrewal, A. & Mofrad, M. R. Deep echocardiography: data-efficient supervised and semi-supervised deep learning towards automated diagnosis of cardiac disease. npj Digit. Med. 1, 59 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

  15. Van Woudenberg, N. et al. in Simulation, Image Processing, and Ultrasound Systems for Assisted Diagnosis and Navigation (eds Stoyanov, D. et al.) 74–81 (Springer, 2018).

  16. Kusunose, K. et al. A deep learning approach for assessment of regional wall motion abnormality from echocardiographic images. JACC Cardiovasc. Imagin 13, 374–381 (2019).

    Article  Google Scholar 

  17. Rajkomar, A. et al. Scalable and accurate deep learning with electronic health records. npj Digit. Med. 1, 18 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

  18. Kwon, J.-m, Kim, K.-H., Jeon, K.-H. & Park, J. Deep learning for predicting in-hospital mortality among heart disease patients based on echocardiography. Echocardiography 36, 213–218 (2019).

    Article  PubMed  Google Scholar 

  19. Avati, A. et al. Improving palliative care with deep learning. BMC Med. Inform. Decis. Mak. 18, 122 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

  20. Motwani, M. et al. Machine learning for prediction of all-cause mortality in patients with suspected coronary artery disease: a 5-year multicentre prospective registry analysis. Eur. Heart J. 38, 500–507 (2016).

    PubMed Central  Google Scholar 

  21. Hadamitzky, M. et al. Optimized prognostic score for coronary computed tomographic angiography: results from the confirm registry (coronary CT angiography evaluation for clinical outcomes: an international multicenter registry). J. Am. Coll. Cardiol. 62, 468–476 (2013).

    Article  PubMed  Google Scholar 

  22. Samad, M. D. et al. Predicting survival from large echocardiography and electronic health record datasets: optimization with machine learning. JACC Cardiovasc. Imaging 12, 681–689 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

  23. Jing, L. et al. A machine learning approach to management of heart failure populations. JACC Heart Fail. 8, 578–587 (2020).

    Article  PubMed  Google Scholar 

  24. Murillo, S. et al. Motion and deformation analysis of ultrasound videos with applications to classification of carotid artery plaques. In SPIE Medical Imaging (SPIE, 2012).

  25. Cui, X. et al. Deformable regions of interest with multiple points for tissue tracking in echocardiography. Med. Image Anal. 35, 554–569 (2017).

    Article  PubMed  Google Scholar 

  26. Raghunath, S. et al. Prediction of mortality from 12-lead electrocardiogram voltage data using a deep neural network. Nat. Med. 26, 886–891 (2020).

    Article  CAS  PubMed  Google Scholar 

  27. Gahungu, N., Trueick, R., Bhat, S., Sengupta, P. P. & Dwivedi, G. Current challenges and recent updates in artificial intelligence and echocardiography. Curr. Cardiovasc. Imaging Rep. 13, 5 (2020).

    Article  Google Scholar 

  28. Horgan, S. J. & Uretsky, S. in Essential Echocardiography: A Companion to Braunwald’s Heart Disease (eds Solomon, S. D. et al.) 460–473 (Elsevier, 2019).

  29. Zhang, J. et al. Fully automated echocardiogram interpretation in clinical practice: feasibility and diagnostic accuracy. Circulation 138, 1623–1635 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

  30. Li, M. et al. Unified model for interpreting multi-view echocardiographic sequences without temporal information. Appl. Soft Comput. 88, 106049 (2020).

    Article  Google Scholar 

  31. Ge, R. et al. K-net: Integrate left ventricle segmentation and direct quantification of paired echo sequence. IEEE Trans. Med. imaging 39, 1690–1702 (2019).

    Article  PubMed  Google Scholar 

  32. Ge, R. et al. Echoquan-net: direct quantification of echo sequence for left ventricle multidimensional indices via global-local learning, geometric adjustment and multi-target relation learning. In International Conference on Artificial Neural Networks (eds Tetko, I. et al.) 219–230 (Springer, 2019).

  33. Jafari, M. H. et al. Automatic biplane left ventricular ejection fraction estimation with mobile point-of-care ultrasound using multi-task learning and adversarial training. Int. J. Comput. Assist. Radiol. Surg. 14, 1027–1037 (2019).

    Article  PubMed  Google Scholar 

  34. Ouyang, D. et al. Video-based AI for beat-to-beat assessment of cardiac function. Nature 580, 252–256 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Ghorbani, A. et al. Deep learning interpretation of echocardiograms. npj Digit. Med. 3, 10 (2020).

    Article  PubMed  PubMed Central  Google Scholar 

  36. Behnami, D. et al. Automatic cine-based detection of patients at high risk of heart failure with reduced ejection fraction in echocardiograms. Comput. Methods Biomech. Biomed. Eng. Imaging Vis. (2019).

  37. Yadlowsky, S. et al. Clinical implications of revised pooled cohort equations for estimating atherosclerotic cardiovascular disease risk. Ann. Intern. Med. 169, 20–29 (2018).

    Article  PubMed  Google Scholar 

  38. Levy, W. C. et al. The Seattle Heart Failure model. Circulation 113, 1424–1433 (2006).

    Article  PubMed  Google Scholar 

  39. McCarty, C. A. et al. The eMERGE network: a consortium of biorepositories linked to electronic medical records data for conducting genomic studies. BMC Med. Genomics 4, 13 (2011).

    Article  PubMed  PubMed Central  Google Scholar 

  40. Wehner, G. J. et al. Routinely reported ejection fraction and mortality in clinical practice: where does the nadir of risk lie? Eur. Heart J. 41, 1249–1257 (2020).

    Article  PubMed  Google Scholar 

  41. Liao, Z. et al. On modelling label uncertainty in deep neural networks: automatic estimation of intra-observer variability in 2D echocardiography quality assessment. IEEE Trans. Med. Imaging 39, 1868–1883 (2019).

    Article  PubMed  Google Scholar 

  42. Behnami, D. et al. Dual-view joint estimation of left ventricular ejection fraction with uncertainty modelling in echocardiograms. In International Conference on Medical Image Computing and Computer-Assisted Intervention (eds Shen, D. et al.) 696–704 (Springer, 2019).

  43. Yancy, C. W. et al. 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. J. Am. Coll. Cardiol. 62, e147–e239 (2013).

    Article  PubMed  Google Scholar 

  44. Lund, L. H., Aaronson, K. D. & Mancini, D. M. Predicting survival in ambulatory patients with severe heart failure on beta-blocker therapy. Am. J. Cardiol. 92, 1350–1354 (2003).

    Article  PubMed  Google Scholar 

  45. Kavalieratos, D. et al. Palliative care in heart failure: rationale, evidence, and future priorities. J. Am. Coll. Cardiol. 70, 1919–1930 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

  46. Rudin, C. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nat. Mach. Intell. 1, 206–215 (2019).

    Article  Google Scholar 

  47. Arcadu, F. et al. Deep learning algorithm predicts diabetic retinopathy progression in individual patients. npj Digit. Med. 2, 92 (2019).

    Article  PubMed  PubMed Central  Google Scholar 

  48. Lee, H. et al. An explainable deep-learning algorithm for the detection of acute intracranial haemorrhage from small datasets. Nat. Biomed. Eng. 3, 173 (2019).

    Article  PubMed  Google Scholar 

  49. Venugopalan, S. et al. Translating videos to natural language using deep recurrent neural networks. In Proceedings of the 2015 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Association for Computational Linguistics, 2015).

  50. Madani, A., Arnaout, R., Mofrad, M. & Arnaout, R. Fast and accurate view classification of echocardiograms using deep learning. npj Digit. Med. 1, 6 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

  51. Tran, D., Bourdev, L., Fergus, R., Torresani, L. & Paluri, M. Learning spatiotemporal features with 3D convolutional networks. In Proc. IEEE International Conference on Computer Vision 4489–4497 (2015).

  52. Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J. & Wojna, Z. Rethinking the inception architecture for computer vision. In Proc. IEEE Conference on Computer Vision and Pattern Recognition 2818–2826 (2016).

  53. Huang, G., Liu, Z., Van Der Maaten, L. & Weinberger, K. Q. Densely connected convolutional networks. In Proc. IEEE Conference on Computer Vision and Pattern Recognition 4700–4708 (2017).

  54. Prechelt, L. in Neural Networks: Tricks of the Trade (eds Montavon, G. et al.) 55–69 (Springer, 1998).

  55. Buuren, S. & Groothuis-Oudshoorn, K. MICE: multivariate imputation by chained equations in R. J. Stat. Softw. 45, jss.v045.i03 (2011).

    Article  Google Scholar 

  56. Chen, T. & Guestrin, C. Xgboost: a scalable tree boosting system. In Proc. 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining 785–794 (ACM, 2016).

  57. Williams, B. A. & Agarwal, S. Applying the Seattle Heart Failure model in the office setting in the era of electronic medical records. Circ. J. 82, 724–731 (2018).

    Article  CAS  PubMed  Google Scholar 

Download references


This work was supported in part by funding from the Pennsylvania Dept of Health (SAP 4100070267 and 4100079720) and the Geisinger Health Plan and Clinic. The content of this article does not reflect the view of the funding sources.

Author information

Authors and Affiliations



A.E.U.-C., C.M.H. and B.K.F. conceived the study and designed the experiments. A.E.U.-C. conducted all experiments. A.E.U.-C. and S.R. wrote the software for applying deep learning to echocardiography videos. A.E.U.-C., L.J., D.P.v., D.N.H., J.D.S. and J.B.L. assembled the input data. H.L.K., G.J.W., M.S.P. and A.A.P. gave advice on experiment design. L.J., C.D.N., C.M.H. and B.K.F. manually audited the data for the cardiologist survey. C.W.G., A.A., J.M.P. and B.J.C. completed the surveys and provided insights on interpretability experiments. A.E.U.-C., C.M.H., M.S.P. and B.K.F. wrote the manuscript. All authors critically revised the manuscript.

Corresponding author

Correspondence to Brandon K. Fornwalt.

Ethics declarations

Competing interests

Geisinger receives funding from Tempus for ongoing development of predictive modelling technology and commercialization. Tempus and Geisinger have jointly applied for a patent related to this work. None of the authors has ownership interest in any of the intellectual property resulting from the partnership.

Additional information

Peer review information Nature Biomedical Engineering thanks Partho Sengupta, Purang Abolmaesumi and the other, anonymous, reviewers for their contribution to the peer review of this work. Peer reviewer reports are available.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Supplementary methods, discussion, figures and tables.

Reporting Summary

Peer Review File

Rights and permissions

Reprints and Permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Ulloa Cerna, A.E., Jing, L., Good, C.W. et al. Deep-learning-assisted analysis of echocardiographic videos improves predictions of all-cause mortality. Nat Biomed Eng 5, 546–554 (2021).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:

This article is cited by


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing