Adaptive reduction of motion artifacts from PPG signals using a synthetic noise reference signal

Abstract

Pulse oximeters estimate both the heart rate and oxygen saturation accurately and are widely used in clinical applications for monitoring the patients at risk of hypoxia. The raw pulse oximeter signal namely Photoplethysmogram (PPG) usually suffers from motion artifacts (MA) corruption, due to the voluntary or involuntary movements of patient while recording the data from PPG sensor. The identification and elimination of these erroneous signal features has received much attention in the scientific literature over recent years. In this paper, we present a simple and efficient adaptive filtering technique for MA reduction using a synthetic noise reference signal without any extra hardware for noise reference signal generation. A thorough experimental analysis is carried out on real MA corrupted PPG data (for horizontal, vertical and bending motions of finger) to demonstrate the efficacy of the proposed method. Simulation results and statistical analysis reveal that the proposed method has shown better performance in MA reduction, making it suitable for pulse oximetry applications.