Towards Understanding ECG Rhythm Classification Using Convolutional Neural Networks and Attention Mappings

Sebastian D. Goodfellow, Andrew Goodwin, Robert Greer, Peter C. Laussen, Mjaye Mazwi, Danny Eytan
Proceedings of the 3rd Machine Learning for Healthcare Conference, PMLR 85:83-101, 2018.

Abstract

Access to electronic health record (EHR) data has motivated computational advances in medical research. However, various concerns, particularly over privacy, can limit access to and collaborative use of EHR data. Sharing synthetic EHR data could mitigate risk. In this paper, we propose a new approach, medical Generative Adversarial Network (medGAN), to generate realistic synthetic patient records. Based on input real patient records, medGAN can generate high-dimensional discrete variables (e.g., binary and count features) via a combination of an autoencoder and generative adversarial networks. We also propose minibatch averaging to efficiently avoid mode collapse, and increase the learning efficiency with batch normalization and shortcut connections. To demonstrate feasibility, we showed that medGAN generates synthetic patient records that achieve comparable performance to real data on many experiments including distribution statistics, predictive modeling tasks and a medical expert review. We also empirically observe a limited privacy risk in both identity and attribute disclosure using medGAN.

Cite this Paper

BibTeX
@InProceedings{pmlr-v85-goodfellow18a, title = {Towards Understanding ECG Rhythm Classification Using Convolutional Neural Networks and Attention Mappings}, author = {Goodfellow, Sebastian D. and Goodwin, Andrew and Greer, Robert and Laussen, Peter C. and Mazwi, Mjaye and Eytan, Danny}, booktitle = {Proceedings of the 3rd Machine Learning for Healthcare Conference}, pages = {83--101}, year = {2018}, editor = {Doshi-Velez, Finale and Fackler, Jim and Jung, Ken and Kale, David and Ranganath, Rajesh and Wallace, Byron and Wiens, Jenna}, volume = {85}, series = {Proceedings of Machine Learning Research}, month = {17--18 Aug}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v85/goodfellow18a/goodfellow18a.pdf}, url = {https://proceedings.mlr.press/v85/goodfellow18a.html}, abstract = {Access to electronic health record (EHR) data has motivated computational advances in medical research. However, various concerns, particularly over privacy, can limit access to and collaborative use of EHR data. Sharing synthetic EHR data could mitigate risk. In this paper, we propose a new approach, medical Generative Adversarial Network (medGAN), to generate realistic synthetic patient records. Based on input real patient records, medGAN can generate high-dimensional discrete variables (e.g., binary and count features) via a combination of an autoencoder and generative adversarial networks. We also propose minibatch averaging to efficiently avoid mode collapse, and increase the learning efficiency with batch normalization and shortcut connections. To demonstrate feasibility, we showed that medGAN generates synthetic patient records that achieve comparable performance to real data on many experiments including distribution statistics, predictive modeling tasks and a medical expert review. We also empirically observe a limited privacy risk in both identity and attribute disclosure using medGAN.} }
Endnote
%0 Conference Paper %T Towards Understanding ECG Rhythm Classification Using Convolutional Neural Networks and Attention Mappings %A Sebastian D. Goodfellow %A Andrew Goodwin %A Robert Greer %A Peter C. Laussen %A Mjaye Mazwi %A Danny Eytan %B Proceedings of the 3rd Machine Learning for Healthcare Conference %C Proceedings of Machine Learning Research %D 2018 %E Finale Doshi-Velez %E Jim Fackler %E Ken Jung %E David Kale %E Rajesh Ranganath %E Byron Wallace %E Jenna Wiens %F pmlr-v85-goodfellow18a %I PMLR %P 83--101 %U https://proceedings.mlr.press/v85/goodfellow18a.html %V 85 %X Access to electronic health record (EHR) data has motivated computational advances in medical research. However, various concerns, particularly over privacy, can limit access to and collaborative use of EHR data. Sharing synthetic EHR data could mitigate risk. In this paper, we propose a new approach, medical Generative Adversarial Network (medGAN), to generate realistic synthetic patient records. Based on input real patient records, medGAN can generate high-dimensional discrete variables (e.g., binary and count features) via a combination of an autoencoder and generative adversarial networks. We also propose minibatch averaging to efficiently avoid mode collapse, and increase the learning efficiency with batch normalization and shortcut connections. To demonstrate feasibility, we showed that medGAN generates synthetic patient records that achieve comparable performance to real data on many experiments including distribution statistics, predictive modeling tasks and a medical expert review. We also empirically observe a limited privacy risk in both identity and attribute disclosure using medGAN.
APA
Goodfellow, S.D., Goodwin, A., Greer, R., Laussen, P.C., Mazwi, M. & Eytan, D.. (2018). Towards Understanding ECG Rhythm Classification Using Convolutional Neural Networks and Attention Mappings. Proceedings of the 3rd Machine Learning for Healthcare Conference, in Proceedings of Machine Learning Research 85:83-101 Available from https://proceedings.mlr.press/v85/goodfellow18a.html.

Related Material