Clutter filtering issues in speckle tracking for two-dimensional blood velocity estimation

Clutter filtering issues remains a major limitation in multi-dimensional blood velocity estimation. In this work we investigate how compounded plane wave imaging can be used to minimize clutter filtering issues when using speckle tracking for two-dimensional velocity estimation. By using a dual scan angle approach one can ensure a sufficient beam-to-flow angle for most imaging scenarios. Segmentation algorithms based on the estimated power and mean axial velocity were used to determine whether estimates were to be discarded or combined. The proposed setup was evaluated in a simulation study (Field II) for straight tube flow with parabolic flow profiles, and for a realistic CFD flow model of a carotid bifurcation. It was shown that the clutter filtering issues were significantly reduced when using compounded imaging and speckle tracking. Regions corrupted by the clutter filter in one scan was not corrupted in the other scan. Thus, consistent estimates could be obtained from the whole region of interest. In addition, the estimates were averaged and the variance reduced in regions where neither scan was corrupted. Regression analysis showed a 10-16 % decrease in standard deviation for the lateral velocity component and a 17 - 19 % decrease for the axial component.