Evaluation of the performance of strategies for the estimation of low velocity blood flow

Current color flow mapping systems are limited in their ability to detect slow blood flow, and to accurately estimate flow velocities in a region that contains vessels with rapid flow as well as slow blood flow. We present results of simulation and in vitro experiments which compare the performance of the following estimation strategies: the autocorrelator, a second order Burg autoregressive estimator, a cross correlation estimator, and the wideband maximum likelihood estimator. Simulation results are presented for a signal in the absence of noise, and in the presence of white noise and clutter