ECG Denoising using a Single-Node Dynamic Reservoir Computing Architecture.

Authors

Aya N. Elbedwehy, Researcher of Electronics and Communications Engineering Department, Faculty of Engineering, Mansoura University, 35516 Mansoura City, EgyptFollow
Mohy Eldin Abo-Elsoud, Professor of Electronics and Communications Engineering Department, Faculty of Engineering, Mansoura University, 35516 Mansoura City, EgyptFollow
Ahmed Elnakib, Associate Professor of Electronics and Communications Engineering Department, Faculty of Engineering, Mansoura University, 35516 Mansoura City, EgyptFollow

Corresponding Author

Elnakib, Ahmed

Subject Area

Biomedical Engineering

Article Type

Original Study

Abstract

Accurate detection of heart disease requires purely realistic electrocardiogram (ECG) signals. In the process of acquisition and transmission, various noises destroy the clean ECG signal, making diagnosis difficult. Here, we apply a single node Reservoir computing (SNRC) architecture based on a recurrent neural network (RNN) to solve this problem by minimizing typical electromyogram noise (EMG) and power line interference (PLI) that damage the ECG signal. MIT-BIH, the standard online arrhythmia database, is used to collect data and test the quality of the proposed method. To evaluate the SNRC architecture, we use two performance indicators, namely, SNR output improvement (SNRimp) and the Percentage Root mean square Difference (PRD). The proposed SNRC architecture is superior to the latest technology and can achieve higher SNRimp and lower PRD for all types of typical ECG noise under study. These results indicate that the proposed SNRC architecture is expected to efficiently restore the dynamics of ECG signals in vivo

Keywords

Electrocardiogram; Reservoir Computing; denoising; Single node reservoir computing

Recommended Citation

N. Elbedwehy, Aya; Abo-Elsoud, Mohy Eldin; and Elnakib, Ahmed (2021) "ECG Denoising using a Single-Node Dynamic Reservoir Computing Architecture.," Mansoura Engineering Journal: Vol. 46 : Iss. 4 , Article 14.
Available at: https://doi.org/10.21608/bfemu.2021.209673