Enhanced Ultrasound Image Formation With Computationally Efficient Cross-Angular Delay Multiply and Sum Beamforming

Cameron A B Smith; Matthieu Toulemonde; Marcelo Lerendegui; Kai Riemer; Dina Malounda; Peter D Weinberg; Mikhail G Shapiro; Meng-Xing Tang

doi:10.1016/j.ultrasmedbio.2025.05.023

Enhanced Ultrasound Image Formation With Computationally Efficient Cross-Angular Delay Multiply and Sum Beamforming

Ultrasound Med Biol. 2025 Jun 28:S0301-5629(25)00177-2. doi: 10.1016/j.ultrasmedbio.2025.05.023. Online ahead of print.

Authors

Cameron A B Smith¹, Matthieu Toulemonde², Marcelo Lerendegui³, Kai Riemer³, Dina Malounda⁴, Peter D Weinberg³, Mikhail G Shapiro⁴, Meng-Xing Tang⁵

Affiliations

¹ Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, CA, USA; Department of Bioengineering, Imperial College London, London, UK.
² Department of Bioengineering, Imperial College London, London, UK; Department of Earth Science and Engineering, Imperial College London, London, UK.
³ Department of Bioengineering, Imperial College London, London, UK.
⁴ Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, CA, USA.
⁵ Department of Bioengineering, Imperial College London, London, UK. Electronic address: mengxing.tang@imperial.ac.uk.

PMID: 40582893
DOI: 10.1016/j.ultrasmedbio.2025.05.023

Abstract

Objective: Ultrasound imaging as a clinical tool is commonly achieved using the delay and sum (DAS) beamforming algorithm, which has limited resolution and suffers from high side lobes. Recently nonlinear processing has proven to be an effective way to enhance image quality. In this work, we describe a new beamforming algorithm called Cross-Angular Delay Multiply and Sum (CADMAS).

Methods: CADMAS takes advantage of nonlinear compounding across planewave steering angles to enhance image contrast. This is then implemented with a mathematical reformulation to produce images at a low computational cost. We tested CADMAS in both conventional B-Mode and amplitude modulation imaging across in vitro and in vivo datasets, and for microbubbles and gas vesicles. We compared the results to DAS and two other nonlinear beamformers, frame multiply and sum (FMAS) and delay multiply and sum (DMAS), as well as the combination of the two.

Results: Our results show on average across our datasets a 7.6 to 40 dB improvement in image contrast over DAS and a 4.8 to 20 dB improvement over DMAS. The computation time of CADMAS is between 1 and 2 times that of DAS; in our implementation we experienced a less than 6% increase in computation time.

Conclusion: Our results show a robust improvement in contrast across multiple datasets both in vitro and in vivo, demonstrating its potential for biological and clinical applications.

Keywords: AM; Acoustics; B-Mode; CADMAS; Computational efficiency; Contrast; DMAS; FMAS; Gas vesicles; Ultrafast imaging; Ultrasound.