Learnable Wiener Postfilter Based Beamformer for Improved Ultrafast Power Doppler Imaging

Ultrafast power Doppler imaging (uPDI) has been advanced to visualize superb small vessels. Yet, its effectiveness is hindered by low signal-to-noise ratio, a consequence of the unfocused transmission of plane waves. A beamformer, leveraging a united spatial-angular adaptive scaling Wiener postfilter (uSA-ASW), has been introduced to enhance uPDI quality, albeit with significant computational demands. In this study, we introduce a learnable Wiener postfilter (LWP)-based beamformer. This approach employs a deep learning model to discern the coherence weights across various channels and transmit angles, thereby accelerating the uSA-ASW beamformer for uPDI applications. Experimental results indicate that the LWP-based beamformer maintains superior quality and speed simultaneously in comparison to other beamformers.