A solution to diffraction biases in sonoelasticity: The acoustic impulse technique

TLDR

Several methods estimate soft tissue viscoelasticity using low‑frequency vibrations (10–500 Hz) by measuring shear‑wave phase velocity (~5 m/s). The article demonstrates that velocity and attenuation measurements are biased, but that low‑frequency transient excitation can eliminate these biases. Biases arise from boundary reflections, piston‑source diffraction, and longitudinal components of piston‑generated shear waves that interfere with low‑frequency compressional waves. Using the corrected method, accurate elasticity and viscosity values were obtained in phantom and beef muscle experiments, and a relative hardness image of a two‑medium phantom was produced.

Abstract

Several methods have been proposed to estimate the viscoelastic properties of soft biological tissues using forced low-frequency vibrations (10-500 Hz). Those methods are based on the measurement of phase velocity of the shear waves (approximately 5 m/s). It is shown in this article that the measurements of velocity as well as attenuation are subjected to biases. These biases are related to reflected waves created at boundaries, to the nonnegligible size of the piston source which causes diffraction effects and to the influence of a low-frequency compressional wave. Indeed, a theoretical analysis of the field radiated by a point source explains how mechanical vibrations of a piston generate a shear wave with a longitudinal component and how this component can interfere with a low-frequency compressional wave. However, by using a low-frequency transient excitation, these biases can be avoided. Then the precise numerical values of elasticity and viscosity can be deduced. Experiments in phantoms and beef muscles are shown. Moreover, a relative hardness imaging of a phantom composed of two media with different elasticities is presented.