TY - JOUR
T1 - GeoECG
T2 - 7th Machine Learning for Healthcare Conference, MLHC 2022
AU - Zhu, Jiacheng
AU - Qiu, Jielin
AU - Yang, Zhuolin
AU - Weber, Douglas
AU - Rosenberg, Michael A.
AU - Liu, Emerson
AU - Li, Bo
AU - Zhao, Ding
N1 - Funding Information:
The work is partly supported by the Allegheny Health Network and Mario Lemieux Center for Innovation and Research in EP.
Publisher Copyright:
© 2022 J. Zhu∗, J. Qiu∗, Z. Yang, D. Weber, M.A. Rosenberg, E. Liu, B. Li & D. Zhao.
PY - 2022
Y1 - 2022
N2 - There has been an increased interest in applying deep neural networks to automatically interpret and analyze the 12-lead electrocardiogram (ECG). The current paradigms with machine learning methods are often limited by the amount of labeled data. This phenomenon is particularly problematic for clinically-relevant data, where labeling at scale can be time-consuming and costly in terms of the specialized expertise and human effort required. Moreover, deep learning classifiers may be vulnerable to adversarial examples and perturbations, which could have catastrophic consequences, for example, when applied in the context of medical treatment, clinical trials, or insurance claims. In this paper, we propose a physiologically-inspired data augmentation method to improve performance and increase the robustness of heart disease detection based on ECG signals. We obtain augmented samples by perturbing the data distribution towards other classes along the geodesic in Wasserstein space. To better utilize domain-specific knowledge, we design a ground metric that recognizes the difference between ECG signals based on physiologically determined features. Learning from 12-lead ECG signals, our model is able to distinguish five categories of cardiac conditions. Our results demonstrate improvements in accuracy and robustness, reflecting the effectiveness of our data augmentation method.
AB - There has been an increased interest in applying deep neural networks to automatically interpret and analyze the 12-lead electrocardiogram (ECG). The current paradigms with machine learning methods are often limited by the amount of labeled data. This phenomenon is particularly problematic for clinically-relevant data, where labeling at scale can be time-consuming and costly in terms of the specialized expertise and human effort required. Moreover, deep learning classifiers may be vulnerable to adversarial examples and perturbations, which could have catastrophic consequences, for example, when applied in the context of medical treatment, clinical trials, or insurance claims. In this paper, we propose a physiologically-inspired data augmentation method to improve performance and increase the robustness of heart disease detection based on ECG signals. We obtain augmented samples by perturbing the data distribution towards other classes along the geodesic in Wasserstein space. To better utilize domain-specific knowledge, we design a ground metric that recognizes the difference between ECG signals based on physiologically determined features. Learning from 12-lead ECG signals, our model is able to distinguish five categories of cardiac conditions. Our results demonstrate improvements in accuracy and robustness, reflecting the effectiveness of our data augmentation method.
UR - http://www.scopus.com/inward/record.url?scp=85164536685&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85164536685&partnerID=8YFLogxK
M3 - Conference article
AN - SCOPUS:85164536685
SN - 2640-3498
VL - 182
SP - 172
EP - 197
JO - Proceedings of Machine Learning Research
JF - Proceedings of Machine Learning Research
Y2 - 5 August 2022 through 6 August 2022
ER -