Cross-Channel Dynamic Weighting RPCA: A De-Noising Algorithm for Multi-Channel Arterial Pulse Signal

Abstract

Pulse wave analysis (PWA) has been widely used in the medical field. A novel multi-channel sensor is employed in arterial pulse acquisition and brings richer physiological information to PWA. However, the noise of this sensor is distributed in the main frequency band of the pulse signal, which seriously interferes with subsequent analyses and is difficult to eliminate by existing methods. This study proposes a cross-channel dynamic weighting robust principal component analysis algorithm. A channel-scaled factor technique is used to manipulate the weighting factors in the nuclear norm. This factor can adaptively adjust the weights among the channels according to the signal pattern of each channel, optimizing the feature extraction in multi-channel signals. A series of performance evaluations were conducted, and four well-known de-noising algorithms were used for comparison. The results reveal that the proposed algorithm achieved one of the best de-noising performances in the time and frequency domains. The mean of in the amplitude relative error (ARE) was 23.4% smaller than for the WRPCA algorithm. Moreover, our algorithm could accelerate convergence and reduce the computational time complexity by approximately 34.6%. These results demonstrate the performance and efficiency of the algorithm. Meanwhile, the idea can be extended to other multi-channel physiological signal de-noising and feature extraction fields.

A reliable evaluation approach for multichannel signal denoising algorithms based on a novel arterial pulse acquisition system

“…

Fig. 5 Illustration of the waveforms and spectrums of the noise of some participants [ 13 ]. (a) the spectrum (upper) and the waveform (lower) of the noise in a 3 × 4 PPS tactile sensor; (b) the spectrum (upper) and the waveform (lower) of the noise in a 4 × 6 PPS tactile sensor; (c) the spectrum (upper) and the waveform (lower) of the noise in a 5 × 5−1 PPS tactile sensor.

…”

“…RPCA-based methods. RPCA [ 27 ], Weighted RPCA [ 28 ] and cc-DWRPCA [ 13 ] are employed in this study. These algorithms denoise the pulse signals by extracting the common features of all cycles for every channel.…”

“…The formula of ARE [ 13 ] is: Where N P and N C are the total number of cardiac cycles and channels, n p and n c are the number of the current cardiac cycle and channel, a t is the amplitude of the single-reference pulse signals at each sampling point, and a t is the amplitude of the filtered pulse signals at each sampling point.…”

“…Most 3DPI studies make use of the tactile sensors developed by PPS (Pressure Profile Systems Inc., Los Angeles, CA, USA), which are widely accepted as the preferred sensors for acquiring physiological signals in a wide range of clinical settings [ 12 ]. Noises of 3DPIs include background noise, contact noise, and motion artefacts [ 13 ]. Intrinsic characteristics of multichannel sensor components cause the background noise of tactile sensors.…”

“…However, these traditional algorithms cannot accommodate the denoising requirements for 3DPIs because they perform the same denoising process for all channels. In recent years, in order to accommodate the denoising requirements for 3DPIs, algorithms have been developed using Robust Principal Component Analysis based algorithms, including Robust Principal Component Analysis (RPCA), Weighted Principal Component Analysis (WRPCA), and Cross-Channel Dynamic Weighting Principal Component Analysis (cc-DWRPCA) [ 13 , 27 , 28 ].…”