Coherent Plane Wave Compounding combined with Euclidean distance transform for high frame rate and high contrast ultrafast imaging

Ultrasound Coherent Plane Wave Compounding (CPWC) is a beamforming technique that generally provides higher image contrast compared to single-angle plane wave imaging (PWI). However, when a reduced number of compounding angles is used to achieve higher frame rates, the contrast may degrade due to artifacts such as grating lobes, sidelobes, and ghost artifacts. In this study, our objective is to improve the image contrast in CPWC imaging while using a lower number of transmissions, reaching a higher frame rate and high contrast CPWC image. To this end, we propose a spatially weighted CPWC imaging based on the distance transform. More specifically, we propose utilizing the angle-wise Euclidean distance transform (EDT)- based filter as a joint post-beamformer to reduce the negative impact of grating lobes and ghosting artifacts in image contrast before coherently compounding the radio frequency (RF) images of different angles. By employing the proposed EDT-based technique, we do emphasize 'on-axis' information while suppressing 'off-axis' information, thereby reducing sidelobes, grating lobes, and ghosting artifacts. We applied the suggested technique to the plane wave imaging challenge in medical ultrasound (PICMUS) database, and the results revealed a significant improvement in image contrast while reaching a 5x improvement in the acquisition frame rate. Quantitative analysis of the proposed method using the generalized contrast-to-noise ratio (gCNR) across varying noise levels demonstrates its robustness in preserving cyst detectability. More specifically, the gCNR of the proposed technique remains consistently high, above 0.88, as SNR decreases from ＋9 dB to -21 dB, while CPWC shows a significant performance drop, with gCNR values dropping from 0.89 to 0.37. These results confirm that the proposed approach effectively maintains cyst detectability even under severe noise conditions.