Study on conditioning and feature extraction algorithm of photoplethysmography signal for physiological parameters detection

Abstract

Photoplethysmography(PPG) signal can reflect many physiological parameters, such as heart function, blood vessel elasticity, blood viscosity and so on. It was a novel noninvasive method with the advantage of convenience and accuracy. It was important to find efficient pre-processing and feature extraction algorithms to deal with original PPG signal, which was interfered by many other factors. Many practical methods including median filtering and FIR filtering was used. A new algorithm based on wavelet transformation was proposed for eliminating the baseline drift. Feature points extraction was another key issue. An improved differential algorithm was used to solve this problem. All of these practical algorithms provided an effective platform for physiological parameters detection.