Application of Tensor Completion for Reducing the Beamforming Time in Ultrafast Ultrasound Imaging: A Doppler Ultrasound Assessment

Doppler ultrasound is a non-invasive technique used to measure blood flow in blood vessels. However, this application requires a large number of temporal frames for precise blood flow measurements, posing challenges for beamforming time, even with a simple delay and sum (DAS) beamformer. Therefore, in this study, we aim to address this issue by proposing a post-processing solution and evaluating its effectiveness for Doppler ultrasound. In this study, We introduce sparse beamforming as a method to reduce beamforming time in preprocessing followed by applying tensor completion (TC) to reconstruct the missing information of the 3-dimensional (3D) beamforming data. We applied the proposed technique to the carotid artery data, demonstrating the feasiblity of improving the beamforming computational complexity by 4 to 5 times, while achieving comparable doppler velocity estimates to the original case (i.e., without sparse beamforming).