A Fast Fourier Beamformer for Convex-array Passive Acoustic Mapping

Passive acoustic mapping (PAM) with convex arrays is often used to monitor microbubble (MB) cavitation activities when the field-of-view (FOV) is large. However, current PAM reconstruction methods for convex arrays are based on the delay and sum (DAS) operation, which is a pixel-based beamformer. In this work, we propose a more efficient Fourier beamformer (FB) for PAM with convex arrays, which is based on extrapolating the acoustic fields using the helical wave spectrum and a transfer function. In the tube experiments, our method was compared with the time-domain (TD-) and frequency-domain (FD-) DAS methods using the metrics including spatial resolution, signal-to-background-ratio (SBR) and computational time. Compared with the FD-DAS and TD-DAS methods, the PAM images formed by FB have comparable image quality and are 47 times and 186 times faster, respectively. A wall-less vessel phantom was used to further demonstrate the cavitation source localization capability of our PAM method under the guidance of B-mode images. The cavitation source was well localized in the wall-less tunnel in the phantom. These results demonstrated that our method is highly efficient and may be useful for real-time monitoring of MB cavitation activities in a large FOV using convex arrays.