Statistical approach for the detection of motion/noise artifacts in Photoplethysmogram

Abstract

Motion and noise artifacts (MNA) have been a serious obstacle in realizing the potential of Photoplethysmogram (PPG) signals for real-time monitoring of vital signs. We present a statistical approach based on the computation of kurtosis and Shannon Entropy (SE) for the accurate detection of MNA in PPG data. The MNA detection algorithm was verified on multi-site PPG data collected from both laboratory and clinical settings. The accuracy of the fusion of kurtosis and SE metrics for the artifact detection was 99.0%, 94.8% and 93.3% in simultaneously recorded ear, finger and forehead PPGs obtained in a clinical setting, respectively. For laboratory PPG data recorded from a finger with contrived artifacts, the accuracy was 88.8%. It was identified that the measurements from the forehead PPG sensor contained the most artifacts followed by finger and ear. The proposed MNA algorithm can be implemented in real-time as the computation time was 0.14 seconds using Matlab®.

Characterization of the relative contributions from systemic physiological noise to whole-brain resting-state functional near-infrared spectroscopy data using single-channel independent component analysis

“…Then, the data were decomposed into independent components by robustICA. To identify channels affected by motion artifacts, we used the kurtosis-based method, 31 in which kurtosis has been defined as a measure of the "peakedness" of the probability distribution. We removed the components having probability distribution sharply peaked at the center (a kurtosis value >25) and then back-projected the multichannel fNIRS data using the remaining components.…”

A new movement artifact detector for photoplethysmographic signals

“…The level of uncertainty in a stochastic signal, such as the PPG, can be quantified using Shannon entropy [10]. This concept was used to design an entropy-based detection statistic based on the hypothesis that segments contaminated by movement artifacts would have higher entropy than artifact-free segments [9].…”

“…Two higher-order statistics, kurtosis and skewness, have been applied for detection of PPG movement artifacts [8,9]. Kurtosis is defined as the standardized fourth central moment of a stochastic signal.…”

“…To follow the implementation of the entropy-based detector developed in [9], the PPG signal was band-pass filtered prior to (1) with a finite impulse response (FIR) filter of order 64 with cut-off frequencies of 0.1 and 10 Hz, to remove low frequency trends and higher frequencies where the PPG was not expected to contain any significant power.…”

“…However, before any artifact reduction technique is applied, it is necessary first to detect corrupted segments so that artifact-free data do not undergo unnecessary processing that might alter relevant information. A variety of methods have been proposed for this, including methods based on higher-order statistics in the time and frequency domains [8,9] (e.g., kurtosis, spectral kurtosis), as well as Shannon entropy [9]. However, performance of these methods varies widely and they have not been evaluated under the high noise conditions that prevail in environments such as intensive care units or the home.…”