Influence of Beamforming Methods on Velocity Estimation: In Vitro Experiments

Many beamforming methods have been developed for ultrasound B-mode imaging, as an alternative to conventional Delay-And-Sum (DAS) algorithm. Fourier-based methods decrease the computational load, while adaptive ones enhance B-mode image quality. These beamformers were mainly evaluated using image quality metrics on static media, such as resolution and contrast, but never on motion estimation. Herein, the influence of six beamformers on velocity estimation is quantitatively evaluated experimentally with a rotating disk. The conventional DAS was compared with several beamformers introduced in the literature: two Fourier-based techniques (Lu's and f-k), and three adaptive ones (Minimum Variance (MV), Phase Coherence Factor (PCF) and DAS with p-th root compression of signals (p-Das)).The experimental results demonstrate that advanced beamformers are compatible with velocity estimation. Fourier-based methods slightly outperformed DAS, and adaptive beamformers can offer comparable velocity maps as DAS with similar statistics regarding bias and standard deviation. However, their performances decreased when increasing the adaptive effect.