EFFECT OF PULSE NUMBER ON THE PULSE-BASED SYNCHRONIZED-AVERAGING ANALYSIS IN LASER DOPPLER SIGNAL

Abstract

Pulse parameters calculated from the LDF waveform based on a time-domain synchronized averaging analysis were shown to be able to discriminate the difference in microvascular resistance. However, its applicability may depend on the validation of signal stationarity. In this study, our aim is to investigate the effect of pulse number, which may destroy the signal stationarity, on various pulse LDF parameters. Analysis was performed in data obtained on healthy volunteers. When one pulse parameter is deviated from the standard value for more than 10%, it was regarded as an error; EP (error probability) was then defined as the occurring probability of error. It was revealed in this study that average parameter deviations for FDT and PDT were smaller than 5% for all tested pulse numbers. If we set 10% as the parameter-deviation criterion as well as the acceptable EP range, there should be at least 120 pulses for FDT and PDT, and 210 pulses for FNA and PW. The study presented here has established the criteria for appropriate pulse number to achieve the signal stationarity; we can thus get accurate pulse parameters so that the microcirculatory discriminability of the pulse-based time-averaging analysis on LDF signal can be improved. The proposed quantitative method to verify the validation of signal stationarity when utilizing time-averaging can also be applied to analysis of other bio-signals.