Photoplethysmographic Pulse Quality Assessment Methods Based on Similarity Analysis

In this study, we proposed photoplethysmography (PPG) waveform quality assessment indices based on waveform similarity analysis. We developed three pulse quality indices (PQIs) based on Euclidian distance and dynamic time warping. In evaluation, we generated normal PPG waveform template using 950 normal pulses of 2 subject's PPG, and then evaluated the performance of pulse quality assessment with 960 total pulses of another subject including 480 abnormal pulses. As a result, significant difference between normal and abnormal pulse was found in all of the proposed PQIs (p<0.001). In abnormal pulse beat classification, the accuracy of proposed indices was 87.27 ± 4.33%, which is 23.34% higher than the average accuracy of existing PQIs such as zero crossing rate, perfusion, skewness, kurtosis or entropy.