Time-varying parametric modeling and simulation of ultrasonic Doppler signals

Abstract

A new approach for the simulation of nonstationary pulsed wave (PW) ultrasonic Doppler signals is proposed. The approach does not assume a theoretical power spectral density function for the PW Doppler signal and employs a time varying ARMA model whose parameters are approximated with a linear combination of Fourier base functions. Comparisons in both time and frequency domains show that the method presented can successfully be used to simulate the PW Doppler signals. Ultrasonic Doppler is a very useful and important technique for noninvasively measuring the velocities of moving particles within the body. The technique is especially used for monitoring a number of time dependent parameters related with blood flow, particularly the flow velocity. The technique is employed for blood velocity measurement.