Fast adaptive beamforming through a cascade structure for ultrasound imaging.

Abstract

A major limiting factor for applying a minimum variance (MV) beamformer in medical ultrasound imaging is its high computational complexity. This paper introduces a new fast MV beamforming method with almost the same capabilities as the standard MV. The fast beamformer is implemented using a cascade structure. At the first stage, the echo signals received from the points far from the main axis are strongly suppressed using a fixed-weight beamformer. At the second stage, after spatially decimating the output of the first stage, an MV-based adaptive beamformer is used to eliminate the echo signals from the points adjacent to the focal point. The greatest advantage of the proposed method is that the second beamformer can be a low-complexity implementation of MV such as beamspace (BS) MV to further reduce the complexity, resulting in a superfast MV. The resulting beamformers were evaluated through both simulation and experimental data, and it was verified that the method was competitive with standard MV and BS methods at a lower computational cost. The new fast and superfast MV methods are capable of obtaining the same results as the MV and BS-MV, at a significantly lower computational cost.