Feasibility of Wearable Armband Bipolar ECG Lead-1 for Long-term HRV Monitoring by Combined Signal Averaging and 2-stage Wavelet Denoising

Abstract

Heart rate variability (HRV) is a clinically important and prominent cardiovascular diseases diagnostic factor. Since HRV is a highly individualised measure, long-term continuous ECG and HRV tracking using a non-invasive armband-based wearable monitoring device is an appealing option for HRV trend-based indicator of general health. Therefore, we investigated the correlation between the bipolar arm-ECG Lead-1 (electrodes axis coplanar to chest and at axilla level) HRV measurements and their corresponding standard measurements from the standard chest ECG Lead I, using a 2 stage dB4 Wavelet-based denoising process supported by an iterative signalaveraged ECG optimal-thresholding adaptation algorithm on the arm-ECG signal, followed by a Pan-Tompkins QRSdetection algorithm. The conventional Pearson correlation coefficient was used as the main performance assessment metric. Four clinically common HRV time-domain metrics were studied: SDNN, RR-rms, RR-median and the interquartile-range value of normal-to-normal heartbeat intervals (IQRNN). The results revealed that RR-rms and RR-median HRV metrics from bipolar arm-ECG (Lead-1) closely correlated to the values measured from the standard Lead-I and present potential for clinical use.

Robust Arm Impedocardiography Signal Quality Enhancement Using Recursive Signal Averaging and Multi-Stage Wavelet Denoising Methods for Long-Term Cardiac Contractility Monitoring Armbands

“…It operates with a fixed bioimpedance constant AC source at 100 kHz frequency and 1 mA (rms) regulated amplitude. operation due to its wide use in ECG and ICG signals filtering reported in the literature and the team's expertise on using this particular wavelet family [22,28,29].…”

“…With ICG recordings of 480 s in this study, of particular interest are Nb values associated with a relatively small-time proportion of the recording time, say around 10% of it, or less. Therefore, to investigate the relative signal quality enhancement of the Rav process for 3 convenient values of Nb, namely, 16, 32 and 64, the process was applied to simultaneous ECG signals from the chest (standard Lead I) and arm (bipolar Lead-1) database that was readily available from a previous investigation [22]. The outcomes of this exploratory investigation with 9 cases (sample size) are presented in Table 3.…”

“…Further robust denoising methods are required for addressing Arm-ICG signal quality enhancement issues in such a way that makes the signal quality clinically acceptable and attractive for long-term ambulatory cardiac contractility (SV) monitoring for the benefit of cardiac patients. Based on previous knowledge of effective Arm-ICG and Arm-ECG denoising [7,[22][23][24][25][26][27][28][29] methods, a satisfactory and robust solution was investigated by means of advanced processes, such as discrete wavelet transform (DWT)-based filtering techniques, supported by ensemble averaging algorithms for cardiac deterministic ICG events [30], both in the Arm-ICG and Thorax-ICG on a beat-by-beat (BbyB) basis. The widely recommended mother wavelet configuration used in DWT-based denoising approaches for ICG signals is the Daubechies wavelet, and more specifically, the fourth-order (Db4) and the eighth-order (Db8) Dabechies wavelets were reported to be suitable for the ICG signal [14,17,22,28,29].…”

“…Based on previous knowledge of effective Arm-ICG and Arm-ECG denoising [7,[22][23][24][25][26][27][28][29] methods, a satisfactory and robust solution was investigated by means of advanced processes, such as discrete wavelet transform (DWT)-based filtering techniques, supported by ensemble averaging algorithms for cardiac deterministic ICG events [30], both in the Arm-ICG and Thorax-ICG on a beat-by-beat (BbyB) basis. The widely recommended mother wavelet configuration used in DWT-based denoising approaches for ICG signals is the Daubechies wavelet, and more specifically, the fourth-order (Db4) and the eighth-order (Db8) Dabechies wavelets were reported to be suitable for the ICG signal [14,17,22,28,29]. Moreover, Chabchoub et al [17] conducted an experiment to compare different wavelet families, i.e., Haar, Daubechie (Db), Symlet (Sym) and Coiflet.…”

“…However, the refined question of which of these wavelets would be more suitable for denoising the Arm-ICG waveform characteristics on a BbyB basis needs to be addressed. Therefore, at this pilot research level, an initial step for addressing this refined question was taken by focussing on the Db wavelet family and investigating their optimisation for Arm-ICG signals denoising for the BbyB mode of operation due to its wide use in ECG and ICG signals filtering reported in the literature and the team's expertise on using this particular wavelet family [22,28,29].…”