A Physiological Information Extraction Method Based on Wearable PPG Sensors with Motion Artifact Removal

In this paper, a method to extract physiological information from Photoplethysmography (PPG) signal with motion artifact detection and removal is proposed. First, PPG data recorded from wearable sensors are pre-processed by discrete wavelet transform to remove unwanted noise and extract AC/DC component. Then, the characteristic points such as peaks and troughs are identified for feature extraction and physiological information extraction. To detect the motion artifact periods, we use support vector machine (SVM) for classification. The detection performance was verified by PPG signals from 11 different healthy subjects. The proposed method operates in 7 seconds with accuracy of 94.4%, sensitivity of 90.35%, and specificity of 99.36%. The motion artifact removal is accomplished by Kalman filter to track the SpO(2) and heart rate (HR) extracted from motion artifact-corrupted periods. The parameters of Kalman filter are determined by the detection results. In the case of waving hand left-right, the average mean absolute bias of artifact-corrupted SpO(2) and HR are 1.34% and 7.29 bpm, respectively. After applying the algorithm, the bias becomes 0.8% and 4.29 bpm. In the case of waving hand up and down, the errors of SpO(2) and HR reduce from 1.31% to 0.82% and 13.97 bpm to 6.87 bpm.