A Study of Double-Stage DMAS and p-DMAS for Their Relation in Baseband Ultrasound Beamforming. (July 2020)

• Record Type: Journal Article
• Authors: Shen, Che-Chou
• Is Part Of: Biomedical signal processing and control. Volume 60(2020)
• Publication Date: 2020-07
• Subjects: Delay-Multiply-and-Sum (DMAS) -- Double-Stage DMAS (DS-DMAS) -- p-DMAS -- Baseband beamforming -- Spatial coherence; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2020.101964
• Languages: English
• ISSNs: 1746-8094

Abstract

Abstract: Delay-multiply-and-sum (DMAS) imaging is capable of suppressing clutter interference by introducing spatial coherence of received channel data into the beamforming process. Recently, double-stage DMAS beamforming (DS-DMAS) has been proposed by firstly calculating the column sum of auto-covariance matrix and then applying DMAS in the second stage to further highlight the channel coherence. In contrast, p -DMAS beamforming emphasizes the spatial coherence by adopting a single-stage p -th-root magnitude scaling of the channel signal and p -th-power restoration after channel sum. Though their implementation differs, DS-DMAS and p -DMAS produce visually similar image quality. In this study, the relation between DS-DMAS and p -DMAS is analytically demonstrated and verified using simulations and experiments. Results indicate that DS-DMAS beamforming is virtually equivalent to p -DMAS with p = 3 particularly when high coherent target such as wire reflectors are being imaged. Reconstructed B-mode image and corresponding quantitative metrics also suggest high similarity between DS-DMAS and p -DMAS for speckle-generating object. For numerical B-mode image, their absolute difference is below −43.5 dB for wire reflectors and −42.6 dB for pure speckle phantom. Due to their virtually equivalent performance, p -DMAS beamforming is a better alternative than DS-DMAS in terms of computational complexity.