Ultrasound-based vessel wall tracking: an auto-correlation technique with RF center frequency estimation

Abstract

Vessel diameter is related to the distending blood pressure, and is used in estimations of vessel stiffness parameters. The vessel walls can be tracked by integrating wall velocities estimated by ultrasound (US) Doppler techniques. The purpose of this work was to evaluate the performance of the modified autocorrelation estimator when applied on vessel wall motion. As opposed to the conventional autocorrelation method that only estimates the mean Doppler frequency, the modified autocorrelation method estimates both the mean Doppler frequency and the radiofrequency (RF) center frequency. To make a systematic evaluation of the estimator, we performed computer simulations of vessel wall motion, where pulse bandwidth, signal-to-noise ratio (SNR), signal-to-reverberation ratio, packet size and sample volume were varied. As reference, we also analyzed the conventional autocorrelation method and the cross-correlation method with parabolic interpolation. Under the simulation conditions considered here, the modified autocorrelation method had the lowest bias and variance of the estimators. When integrating velocity estimates over several cardiac cycles, the resulting tissue displacement curves might drift. This drift is directly related to the magnitude of the estimator bias and variance. Hence, the modified autocorrelation method should be the preferred choice of method. (E-mail: s.i.rabben@klinmed.uio.no)