Velocity Profile Detection Through Adaptive Spectral Estimators

The echoes backscattered from red blood cells moving in a vessel at different depths, can be elaborated to obtain the so-called spectral profile, reporting the distribution of Doppler frequencies along the investigating M-line. The typical processing is based on power spectral density estimation applied to the slow-time samples gathered at each depth. The Welch estimator is typically employed, but it features a good spectral resolution only if the observation window (OW) (i.e. data processed for a single estimate) is composed of at least 64-128 samples. Adaptive spectral estimators capable of producing good resolution with reduced OWs, have recently been proposed for spectrogram calculation. In this work we apply Capon and APES adaptive estimator to spectral profile assessment, obtaining good profiles with short OWs. This result can be exploited for improving the temporal resolution and/or introducing multiple Doppler lines. Two examples are reported: in the first, the blood velocity distribution during the fast systolic acceleration in carotid artery is detailed with high temporal resolution; in the second, 4 spectral profiles are simultaneously detected at different sites of the carotid bifurcation.