Effects of the wall filter on the estimation of high blood velocity

In pulsed wave Doppler ultrasonic measurements, a high-pass wall filter is used to remove the clutter signal prior to the blood velocity estimation. For high velocity measurements, the wall filter creates dead zones where the Doppler frequency equals multiples of the pulse repetition frequency (PRF). In this work, the effect of the wall filter has been studied for two different blood velocity estimators; the crosscorrelation method (CCM) and the extended autocorrelation method (EAM). When the pulse bandwidth is sufficiently high, the Doppler signal bandwidth will exceed the wall filter cut-off frequency due to the transit-time effect, and the dead zones are partially removed. However, the chance of velocity aliasing is increased in these zones due to the filtering, both for the CCM and EAM method. The effects of the wall filter have been studied by simulations with rectilinear velocities up to four times the Nyquist limit (ν_NY). In this simulation, the pulse bandwidth is 2.5 MHz. When the cut-off frequency of the wall filter is 0.1^*PRF, no velocity aliasing has been observed. When the wall filter is increased to 0.2*PRF, there is 15% aliasing error occurring at velocity=2^*ν_NY and no velocity aliasing at v=4*ν_NY. When the wall filter is increased to 0.25*PRF, there is about 70% velocity aliasing error at twice ν_NY, and 15% velocity aliasing error at velocity four times ν_NY. The simulation results have further been verified by experimental data from subclavian artery measurements with velocities up to twice the Nyquist limit