Iteratively-Reweighted Beamforming For High-Resolution Ultrasound Imaging

Abstract

Ultrasound imaging typically employs delay-and-sum (DAS) beamformers for image reconstruction. An apodization window is typically used to suppress the beam-pattern’s sidelobes. This approach introduces a trade-off between the mainlobe width versus the sidelobe attenuation and therefore offers limited performance. We consider a statistical framework for beamforming and present two variants. In the first one, the signal of interest is modeled as a Laplacian distributed random variable and the interference is modeled as additive and Gaussian distributed. A closed-form solution is obtained to this optimization problem. In the second variant, we propose an iteratively-reweighted (IR) beamforming algorithm, which solves a constrained optimization problem to determine the optimal apodization weights. This beamformer results in a sharper mainlobe that translates to a finer lateral resolution. The proposed method is compared with the standard DAS beamformer and a recently proposed statistically modeled beamformer, namely iMAP for different number of plane-wave (PW) insonifications. This algorithm is independent of the imaging modality employed and exhibits a superior performance in terms of lateral resolution.