Wearable Pulse Wave Monitor Resistant to Motion Artifacts

Abstract

The aim of this study is to create a wearable device for long-term pulse wave monitoring as well as to investigate the possibility of using adaptive noise cancellation approach for reducing motion artifacts occurred during the real-life recording. In our study wearable monitoring device have acquired pulse wave by using photoplethysmography approach and human movement with triaxial accelerometer. The electrical design of wearable device was based on synchronous demodulation and using 24 bits sigma-delta analog-to-digital converter. To achieve effective and robust motion artifacts reduction we create the pulse wave signal processing method based on band-pass filtering and adaptive noise cancellation. Pulse wave signals were initially pass-band filtered at 0.5–10 Hz to remove noise, electrical and physiological interferences, using a zero-phase forward and reverse digital filter, which first filtered the raw signal in the forward direction, and subsequently filtered the reversed signal, thus the resultant signal has zero-phase distortion. Adaptive noise cancellation was implemented by using a recursive least squares algorithm based on the solution of the Wiener-Hopf equation. Our studies have shown that the best results of pulse wave signal processing are achieved for the following parameters of the algorithm: the forgetting factor of 0.99; filter order of 16. Performance of proposed processing technique was evaluated by assessing signal-to-noise ratio (SNR) of the filtered signal and compared with other approaches such as wavelet multiresolution decomposition and moving average filtering. For correct estimation of SNR we used robust approach based on the eigenvalues of signal autocorrelation matrix. This study indicates that designed wearable device based on principles of photoplethysmography for unobtrusive and noninvasive recording of pulse waves and using advanced digital processing technique for removing motion artifacts could provide an effective and performance tools for improving the long-term healthcare monitoring of human vital signs.