仮想ソースベースの開口合成による発散波イメージング：実験的研究

Ultrafast imaging with coherent compounding has revolutionized medical diagnostics by achieving high frame rates and facilitating advanced applications such as Doppler imaging and shear wave elastography. Among ultrafast techniques, diverging wave imaging (DWI) offers distinct advantages over plane wave imaging (PWI), including wider field-of-view coverage. In ultrafast imaging, either using PWI or DWI, transmit apodization cannot be set in the conventional way by applying different voltages to each transmit element as the beam is formed synthetically during coherent compounding. For DWI, the beam is formed by coherent compounding for a set of spherical waves generated from virtual sources (VSs) placed behind the probe, where applying weights in the compound phase is possible. While several studies have addressed this challenge for PWI, the optimization of those weights for DWI remains unexplored. In our earlier work, we introduced a closed-form approach, under suitable hypotheses, that maps transmit apodization weights from synthetic aperture imaging (SAI) to weights applied during coherent compounding for DWI, and we refer to that set of weights as a compound mask. The approach works for both linear and convex geometries with different arrangements of virtual sources, f-numbers, and apodization windows. Here, we present the real-time implementation of this approach on a Verasonics scanner, validating its efficacy across three VSs configurations (linear, curvilinear, and tilted distributions). Experimental results demonstrate that the compound mask improves the quality of B-mode images with all distributions of VSs for linear and convex arrays, all without compromising real-time performance.