Optimizing the lateral beamforming step for filtered-delay multiply and sum beamforming to improve active contour segmentation using ultrafast ultrasound imaging

As an alternative to delay-and-sum beamforming, a novel beamforming technique called filtered-delay multiply and sum (FDMAS) was introduced recently to improve ultrasound B-mode image quality. Although a considerable amount of work has been performed to evaluate FDMAS performance, no study has yet focused on the beamforming step size, Delta x, in the lateral direction. Accordingly, the performance of FDMAS was evaluated in this study by fine-tuning Delta x to find its optimal value and improve boundary definition when balloon snake active contour (BSAC) segmentation was applied to a B-mode image in ultrafast imaging. To demonstrate the effect of altering Delta x in the lateral direction on FDMAS, measurements were performed on point targets, a tissuemimicking phantom and in vivo carotid artery, by using the ultrasound array research platform II equipped with one 128-element linear array transducer, which was excited by 2-cycle sinusoidal signals. With 9-angle compounding, results showed that the lateral resolution (LR) of the point target was improved by 67.9% and 81.2%, when measured at -6 dB and -20 dB respectively, when Delta x was reduced from lambda to lambda/5. Meanwhile the image contrast ratio (CR) measured on the CIRS phantom was improved by 10.38 dB at the same Delta x reduction and the same number of compounding angles. The enhanced FDMAS results with lower side lobes and less clutter noise in the anechoic regions provides a means to improve boundary definition on a B-mode image when BSAC segmentation is applied.

Automated summary

The study evaluated the performance of FDMAS by fine-tuning Delta x to find its optimal value and improve boundary definition when BSAC segmentation was applied in ultrafast imaging. Results showed improvements in lateral resolution and image contrast ratio with reduced Delta x, offering a means to enhance boundary definition on B-mode images.