Optimal location of pulse compression in coded excitation for medical ultrasound imaging

Abstract

Coded excitation methods can enhance signal-to-noise ratio (SNR) in medical ultrasound imaging. Specifically, the location of pulse compression determines the computational complexity and image quality. Thus, the optimal location of pulse compression is important for coded excitation-based medical ultrasound imaging systems. In this paper, to determine the optimal location of pulse compression, the axial resolution with three different types of codes (i.e. Barker, Golay, and weighted Chirp) was examined. The pulse compression filter can be applied to various locations, i.e., behind analog-to-digital converter (ADC), beamformer (BF), quadrature demodulator (QD). In the Field II simulation, the codes (i.e., Barker, Golay and weighted Chirp with 13, 16 and 16 cycles, respectively) are used. On the other hand, for phantom experiments, pre-beamformed RF data from Barker and Golay codes were captured by a 4-MHz convex probe connected to a commercial ultrasound machine with a research package (SonixTouch, Ultrasonix, Vancouver, BC, Canada.) For all cases, the best image quality was given when pulse compression was placed behind ADC, but it is difficult to be implemented due to high hardware complexity. Behind BF and QD, for Barker and Golay, the degradation in axial resolution was observed due to nonlinear sampling during dynamic receive focusing. This degradation was minimal for Golay. On the other hand, Chirp is robust to the nonlinear sampling artifact regardless the location of pulse compression. Thus, the optimal location of pulse compression for Chirp is the behind beamformer that yields the smaller hardware complexity.