Synthesis of Chest-Lead ECG Using Temporal Convolutional Networks

Cardiovascular diseases (CVDs) are a leading cause of mortality globally, and therefore timely and accurate diagnosis is crucial to patient safety. The standard 12-lead electrocardiography (ECG) is routinely used to diagnose heart disease. Most wearable monitoring devices provide insufficient ECG information because of the limitations in the number of leads and measurement positions. This study presents a patient-specific chest-lead synthesis method based on temporal convolutional network (TCN) to exploit both intra- and inter-lead correlations of ECG signals. Performance can be further enhanced by using the variational mode decomposition (VMD), which reduces the non-stationary characteristic of ECG signals and helps to improve the synthesis accuracy. Experiments on PTB diagnostic database demonstrate that the proposed method is effective and has good performance in synthesis of chest-lead ECG signals from a single limb lead.

[1]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[2]  Pineda,et al.  Generalization of back-propagation to recurrent neural networks. , 1987, Physical review letters.

[3]  Sun Kook Yoo,et al.  Synthesis of Electrocardiogram V-Lead Signals From Limb-Lead Measurement Using R-Peak Aligned Generative Adversarial Network , 2020, IEEE Journal of Biomedical and Health Informatics.

[4]  Uday Maji,et al.  Detection And Characterisation Of QRS Complex In VMD Domain , 2015 .

[5]  Yun Pan,et al.  A lightweight piecewise linear synthesis method for standard 12-lead ECG signals based on adaptive region segmentation , 2018, PloS one.

[6]  Paul Rubel,et al.  A Novel Neural-Network Model for Deriving Standard 12-Lead ECGs From Serial Three-Lead ECGs: Application to Self-Care , 2010, IEEE Transactions on Information Technology in Biomedicine.

[7]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[8]  Jimeng Sun,et al.  Opportunities and Challenges in Deep Learning Methods on Electrocardiogram Data: A Systematic Review , 2020, ArXiv.

[9]  Pachamuthu Rajalakshmi,et al.  Accurate and reliable 3-lead to 12-lead ECG reconstruction methodology for remote health monitoring applications , 2013, 2013 IEEE 15th International Conference on e-Health Networking, Applications and Services (Healthcom 2013).

[10]  Jiss J. Nallikuzhy,et al.  DCT-Based Linear Regression Approach for 12-Lead ECG Synthesis , 2015 .

[11]  Jaakko Malmivuo,et al.  12-Lead ECG System , 1995 .

[12]  Gill R. Tsouri,et al.  Patient-Specific 12-Lead ECG Reconstruction From Sparse Electrodes Using Independent Component Analysis , 2014, IEEE Journal of Biomedical and Health Informatics.

[13]  Vladlen Koltun,et al.  An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling , 2018, ArXiv.

[14]  G.B. Moody,et al.  PhysioNet: a Web-based resource for the study of physiologic signals , 2001, IEEE Engineering in Medicine and Biology Magazine.

[15]  Gretchen A. Stevens,et al.  World Health Organization cardiovascular disease risk charts: revised models to estimate risk in 21 global regions , 2019, The Lancet. Global health.

[16]  Mirza Mansoor Baig,et al.  A comprehensive survey of wearable and wireless ECG monitoring systems for older adults , 2013, Medical & Biological Engineering & Computing.

[17]  Dominique Zosso,et al.  Variational Mode Decomposition , 2014, IEEE Transactions on Signal Processing.

[18]  Kwang-Ting Cheng,et al.  Standard 12-lead ECG synthesis using a GA optimized BP neural network , 2015, 2015 Seventh International Conference on Advanced Computational Intelligence (ICACI).