Skip to main content

Thank you for visiting nature.com. 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.

  • Letter
  • Published:

Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network

A Publisher Correction to this article was published on 24 January 2019

This article has been updated

Abstract

Computerized electrocardiogram (ECG) interpretation plays a critical role in the clinical ECG workflow1. Widely available digital ECG data and the algorithmic paradigm of deep learning2 present an opportunity to substantially improve the accuracy and scalability of automated ECG analysis. However, a comprehensive evaluation of an end-to-end deep learning approach for ECG analysis across a wide variety of diagnostic classes has not been previously reported. Here, we develop a deep neural network (DNN) to classify 12 rhythm classes using 91,232 single-lead ECGs from 53,549 patients who used a single-lead ambulatory ECG monitoring device. When validated against an independent test dataset annotated by a consensus committee of board-certified practicing cardiologists, the DNN achieved an average area under the receiver operating characteristic curve (ROC) of 0.97. The average F1 score, which is the harmonic mean of the positive predictive value and sensitivity, for the DNN (0.837) exceeded that of average cardiologists (0.780). With specificity fixed at the average specificity achieved by cardiologists, the sensitivity of the DNN exceeded the average cardiologist sensitivity for all rhythm classes. These findings demonstrate that an end-to-end deep learning approach can classify a broad range of distinct arrhythmias from single-lead ECGs with high diagnostic performance similar to that of cardiologists. If confirmed in clinical settings, this approach could reduce the rate of misdiagnosed computerized ECG interpretations and improve the efficiency of expert human ECG interpretation by accurately triaging or prioritizing the most urgent conditions.

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

Access options

Rent or buy this article

Prices vary by article type

from$1.95

to$39.95

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

Fig. 1: ROC and precision-recall curves.
Fig. 2: Confusion matrices.

Similar content being viewed by others

Data availability

The test dataset used to support the findings of this study is publicly available at https://irhythm.github.io/cardiol_test_set without restriction. Restrictions apply to the availability of the training dataset, which was used under license from iRhythm Technologies, Inc. for the current study. iRhythm Technologies, Inc. will consider requests to access the training data on an individual basis. Any data use will be restricted to noncommercial research purposes, and the data will only be made available on execution of appropriate data use agreements.

Change history

  • 24 January 2019

    In the version of this article originally published, the x axis labels in Fig. 1a were incorrect. The labels originally were ‘Specificity,’ but should have been ‘1 – Specificity.’ Also, the x axis label in Fig. 2b was incorrect. It was originally ‘DNN predicted label,’ but should have been ‘Average cardiologist label.’ The errors have been corrected in the PDF and HTML versions of this article.

References

  1. Schläpfer, J. & Wellens, H. J. Computer-interpreted electrocardiograms: benefits and limitations. J. Am. Coll. Cardiol. 70, 1183–1192 (2017).

    Article  PubMed  Google Scholar 

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

    Article  CAS  PubMed  Google Scholar 

  3. Holst, H., Ohlsson, M., Peterson, C. & Edenbrandt, L. A confident decision support system for interpreting electrocardiograms. Clin. Physiol. 19, 410–418 (1999).

    Article  CAS  PubMed  Google Scholar 

  4. Schlant, R. C. et al. Guidelines for electrocardiography. A report of the American College of Cardiology/American Heart Association Task Force on assessment of diagnostic and therapeutic cardiovascular procedures (Committee on Electrocardiography). J. Am. Coll. Cardiol. 19, 473–481 (1992).

    Article  CAS  PubMed  Google Scholar 

  5. Shah, A. P. & Rubin, S. A. Errors in the computerized electrocardiogram interpretation of cardiac rhythm. J. Electrocardiol. 40, 385–390 (2007).

    Article  PubMed  Google Scholar 

  6. Guglin, M. E. & Thatai, D. Common errors in computer electrocardiogram interpretation. Int. J. Cardiol. 106, 232–237 (2006).

    Article  PubMed  Google Scholar 

  7. Poon, K., Okin, P. M. & Kligfield, P. Diagnostic performance of a computer-based ECG rhythm algorithm. J. Electrocardiol. 38, 235–238 (2005).

    Article  PubMed  Google Scholar 

  8. Amodei, D. et al. Deep Speech 2: end-to-end Speech recognition in English and Mandarin. In Proc. 33rd International Conference on Machine Learning, 173–182 (2016).

  9. He, K., Zhang, X., Ren, S. & Sun, J. Delving deep into rectifiers: surpassing human-level performance on ImageNet classification. In Proc. International Conference on Computer Vision, 1026–1034 (IEEE, 2015).

  10. Silver, D. et al. Mastering the game of Go with deep neural networks and tree search. Nature 529, 484–489 (2016).

    Article  CAS  PubMed  Google Scholar 

  11. 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 

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

    Article  CAS  PubMed  Google Scholar 

  13. Poungponsri, S. & Yu, X. An adaptive filtering approach for electrocardiogram (ECG) signal noise reduction using neural networks. Neurocomputing 117, 206–213 (2013).

    Article  Google Scholar 

  14. Ochoa, A., Mena, L. J. & Felix, V. G. Noise-tolerant neural network approach for electrocardiogram signal classification. In Proc. 3rd International Conference on Compute and Data Analysis, 277–282 (Association for Computing Machinery, 2017).

  15. Mateo, J. & Rieta, J. J. Application of artificial neural networks for versatile preprocessing of electrocardiogram recordings. J. Med. Eng. Technol. 36, 90–101 (2012).

    Article  CAS  PubMed  Google Scholar 

  16. Pourbabaee, B., Roshtkhari, M. J. & Khorasani, K. Deep convolutional neural networks and learning ECG features for screening paroxysmal atrial fibrillation patients. IEEE Trans. Syst. Man Cybern. Syst. 99, 1–10 (2017).

    Google Scholar 

  17. Javadi, M., Arani, S. A., Sajedin, A. & Ebrahimpour, R. Classification of ECG arrhythmia by a modular neural network based on mixture of experts and negatively correlated learning. Biomed. Signal Process. Control 8, 289–296 (2013).

    Article  Google Scholar 

  18. Acharya, U. R. et al. A deep convolutional neural network model to classify heartbeats. Comput. Biol. Med. 89, 389–396 (2017).

    Article  PubMed  Google Scholar 

  19. Banupriya, C. V. & Karpagavalli, S. Electrocardiogram beat classification using probabilistic neural network. In Proc. Machine Learning: Challenges and Opportunities Ahead 31–37 (2014).

  20. Al Rahhal, M. M. et al. Deep learning approach for active classification of electrocardiogram signals. Inf. Sci. (NY) 345, 340–354 (2016).

    Article  Google Scholar 

  21. Acharya, U. R. et al. Automated detection of arrhythmias using different intervals of tachycardia ECG segments with convolutional neural network. Inf. Sci. (NY) 405, 81–90 (2017).

    Article  Google Scholar 

  22. Zihlmann, M., Perekrestenko, D. & Tschannen, M. Convolutional recurrent neural networks for electrocardiogram classification. Comput. Cardiol. https://doi.org/10.22489/CinC.2017.070-060 (2017).

  23. Xiong, Z., Zhao, J. & Stiles, M. K. Robust ECG signal classification for detection of atrial fibrillation using a novel neural network. Comput. Cardiol. https://doi.org/10.22489/CinC.2017.066-138 (2017).

  24. Clifford, G. et al. AF classification from a short single lead ECG recording: the PhysioNet/Computing in Cardiology Challenge 2017. Comput. Cardiol. https://doi.org/10.22489/CinC.2017.065-469 (2017).

  25. Teijeiro, T., Garcia, C. A., Castro, D. & Felix, P. Arrhythmia classification from the abductive interpretation of short single-lead ECG records. Comput. Cardiol. https://doi.org/10.22489/CinC.2017.166-054 (2017).

  26. Goldberger, A. L. et al. PhysioBank, PhysioToolkit, and PhysioNet: components of a new research resource for complex physiologic signals. Circulation 101, E215–E220 (2000).

    CAS  PubMed  Google Scholar 

  27. Turakhia, M. P. et al. Diagnostic utility of a novel leadless arrhythmia monitoring device. Am. J. Cardiol. 112, 520–524 (2013).

    Article  PubMed  Google Scholar 

  28. Hand, D. J. & Till, R. J. A simple generalisation of the area under the ROC curve for multiple class classification problems. Mach. Learn. 45, 171–186 (2001).

    Article  Google Scholar 

  29. Smith, M. D. et al.in Best Care at Lower Cost: the Path to Continuously Learning Health Care in America (National Academies Press,: Washington, 2012).

  30. Lyon, A., Mincholé, A., Martínez, J. P., Laguna, P. & Rodriguez, B. Computational techniques for ECG analysis and interpretation in light of their contribution to medical advances. J. R. Soc. Interface 15, pii: 20170821 (2018).

    Article  Google Scholar 

  31. Carrara, M. et al. Heart rate dynamics distinguish among atrial fibrillation, normal sinus rhythm and sinus rhythm with frequent ectopy. Physiol. Meas. 36, 1873–1888 (2015).

    Article  PubMed  Google Scholar 

  32. Zhou, X., Ding, H., Ung, B., Pickwell-MacPherson, E. & Zhang, Y. Automatic online detection of atrial fibrillation based on symbolic dynamics and Shannon entropy. Biomed. Eng. Online 13, 18 (2014).

    Article  PubMed  PubMed Central  Google Scholar 

  33. Hong, S. et al. ENCASE: an ENsemble ClASsifiEr for ECG Classification using expert features and deep neural networks. Comput. Cardiol. https://doi.org/10.22489/CinC.2017.178-245 (2017).

  34. Nahar, J., Imam, T., Tickle, K. S. & Chen, Y. P. Computational intelligence for heart disease diagnosis: a medical knowledge driven approach. Expert Syst. Appl. 40, 96–104 (2013).

    Article  Google Scholar 

  35. Cubanski, D., Cyganski, D., Antman, E. M. & Feldman, C. L. A neural network system for detection of atrial fibrillation in ambulatory electrocardiograms. J. Cardiovasc. Electrophysiol. 5, 602–608 (1994).

    Article  CAS  PubMed  Google Scholar 

  36. Andreotti, F., Carr, O., Pimentel, M. A. F., Mahdi, A. & De Vos, M. Comparing feature-based classifiers and convolutional neural networks to detect arrhythmia from short segments of ECG. Comput. Cardiol. https://doi.org/10.22489/CinC.2017.360-239 (2017).

  37. Xu, S. S., Mak, M. & Cheung, C. Towards end-to-end ECG classification with raw signal extraction and deep neural networks. IEEE J. Biomed. Health Informatics 14, 1 (2018).

    Article  Google Scholar 

  38. Ong, S. L., Ng, E. Y. K., Tan, R. S. & Acharya, U. R. Automated diagnosis of arrhythmia using combination of CNN and LSTM techniques with variable length heart beats. Comput. Biol. Med. 102, 278–287 (2018).

    Article  Google Scholar 

  39. Shashikumar, S. P., Shah, A. J., Clifford, G. D. & Nemati, S. Detection of paroxysmal atrial fibrillation using attention-based bidirectional recurrent neural networks. In Proc. 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 715–723 (Association for Computing Machinery, 2018).

  40. Xia, Y., Wulan, N., Wang, K. & Zhang, H. Detecting atrial fibrillation by deep convolutional neural networks. Comput. Biol. Med. 93, 84–92 (2018).

    Article  PubMed  Google Scholar 

  41. He, K., Zhang, X., Ren, S. & Sun, J. Identity mappings in deep residual networks. In Proc. European Conference on Computer Vision, 630–645 (Springer, 2016).

  42. Ioffe, S. & Szegedy, C. Batch normalization: accelerating deep network training by reducing internal covariate shift. In Proc. International Conference on Machine Learning, 448–456 (2015).

  43. He, K., Zhang, X., Ren, S. & Sun, J. Deep residual learning for image recognition. In Proc. IEEE Conference on Computer Vision and Pattern Recognition, 770–778 (IEEE, 2016).

  44. 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 

  45. Kingma, D. P. & Ba, J. L. Adam: a method for stochastic optimization. In Proc. International Conference on Learning Representations 1–15 (2015).

  46. Hochreiter, S. & Schmidhuber, J. Long short-term memory. Neural Comput. 9, 1735–1780 (1997).

    Article  CAS  PubMed  Google Scholar 

  47. Fawcett, T. An introduction to ROC analysis. Pattern Recognit. Lett. 27, 861–874 (2006).

    Article  Google Scholar 

  48. Hanley, J. A. & McNeil, B. J. A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology 148, 839–843 (1983).

    Article  CAS  PubMed  Google Scholar 

  49. Saito, T. & Rehmsmeier, M. The precision-recall plot is more informative than the ROC plot when evaluating binary classifiers on imbalanced datasets. PLoS ONE 10, e0118432 (2015).

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

iRhythm Technologies, Inc. provided financial support for the data annotation in this work. M.H. and C.B. are employees of iRhythm Technologies, Inc. A.Y.H. was funded by an NVIDIA fellowship. G.H.T. received support from the National Institutes of Health (K23 HL135274). The only financial support provided by iRhythm Technologies, Inc. for this study was for the data annotation. Data analysis and interpretation was performed independently from the sponsor. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.

Author information

Authors and Affiliations

Authors

Contributions

M.H., A.Y.N., A.Y.H., and G.H.T. contributed to the study design. M.H. and C.B. were responsible for data collection. P.R. and A.Y.H. ran the experiments and created the figures. G.H.T., P.R., and A.Y.H. contributed to the analysis. G.H.T., A.Y.H., and M.P.T. contributed to the data interpretation and to the writing. G.H.T., M.P.T., and A.Y.N. advised and A.Y.N. was the senior supervisor of the project. All authors read and approved the submitted manuscript.

Corresponding author

Correspondence to Awni Y. Hannun.

Ethics declarations

Competing interests

M.H. and C.B. are employees of iRhythm Technologies, Inc. G.H.T. is an advisor to Cardiogram, Inc. M.P.T. is a consultant to iRhythm Technologies, Inc. None of the other authors have potential conflicts of interest.

Additional information

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

Extended data

Extended Data Fig. 1 Deep Neural Network architecture.

Our deep neural network consisted of 33 convolutional layers followed by a linear output layer into a softmax. The network accepts raw ECG data as input (sampled at 200 Hz, or 200 samples per second), and outputs a prediction of one out of 12 possible rhythm classes every 256 input samples.

Extended Data Fig. 2 Receiver operating characteristic curves for deep neural network predictions on 12 rhythm classes.

Individual cardiologist performance is indicated by the red crosses and averaged cardiologist performance is indicated by the green dot. The line represents the ROC curve of model performance. AF-atrial fibrillation/atrial flutter; AVB- atrioventricular block; EAR-ectopic atrial rhythm; IVR-idioventricular rhythm; SVT-supraventricular tachycardia; VT-ventricular tachycardia. n = 7,544 where each of the 328 30-second ECGs received 23 sequence-level predictions.

Source data

Supplementary information

Source data

Source data Fig. 1

Sensitivity, specificity and PPV values at different operating points as well as the individual and average cardiologist metrics for the arrhythmias in the figure.

Source data Fig. 2

Absolute confusions counts between arrhythmias for both the model and the cardiologists.

Extended data Fig. 2

Sensitivity and specificity values at different operating points as well as the individual and average cardiologist metrics for all of the arrhythmias.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Hannun, A.Y., Rajpurkar, P., Haghpanahi, M. et al. Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network. Nat Med 25, 65–69 (2019). https://doi.org/10.1038/s41591-018-0268-3

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41591-018-0268-3

This article is cited by

Search

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