Fast Computation of Wideband Beam Pattern for Designing Large-Scale 2-D Arrays

For real-time and high-resolution 3-D ultrasound imaging, the design of sparse distribution and weights of elements of a large-scale wideband 2-D array is needed to reduce hardware cost and achieve better directivity. However, due to the high time consumption of computing the wideband beam pattern, the design methods that need massive iterations have rarely been applied to design large-scale wideband 2-D arrays by directly computing the wideband beam pattern. In this paper, a fast method is proposed to realize the computation of a wideband beam pattern of arbitrary 2-D arrays in the far field in order to design large-scale wideband 2-D arrays. The proposed fast method exploits two important techniques: 1) nonuniform fast Fourier transform (FFT) and 2) short inverse FFT. Compared with the commonly used ultrasound simulator Field II, two orders of magnitude improvement in computation speed is achieved with comparable accuracy. The proposed fast method enables massive iterations of direct wideband beam pattern computation of arbitrary large-scale 2-D arrays. A design example in this paper demonstrates that the proposed fast method can help achieve better performance in designing large-scale wideband 2-D arrays.