Wavelet Decomposition and Feature Extraction from Pulse Signals of the Radial Artery

Abstract

Pressure signals from the radial artery are captured using a novel pulse recording instrument and processed using wavelets. The motivation behind this study is the validation of the noninvasive method of pulse diagnosis. The signals from the radial artery were captured using a robust electro-mechanical system, 'Nadi Yantra' which uses piezoelectric based pressure sensors to capture the palpations. Percussion peaks were identified and the interval between these was used to calculate Heart Rate Varibility (HRV), a useful tool for assessing the status of the autonomic nervous system noninvasively. Time domain indices were calculated from direct measurement of peak-peak (PP) intervals and from differences between the PP intervals. Frequency domain indices such as very low frequency (VLF) power, low frequency (LF) power, high frequency (HF) power, LF/HF ratio were also calculated. A nonlinear Poincare analysis was carried out. Results from 7 datasets are depicted in this paper.