The effect of tissue anisotropy on ultrasound strain imaging (USI): a preliminary study

Abstract

The anisotropic mechanical properties and view-dependent ultrasonic backscattering have been extensively demonstrated in soft tissues, but their potential effects on ultrasound elastographic techniques remain unclear. The aim of this study was to investigate the performance of a speckle-tracking-based 2D ultrasound strain imaging (USI) method in silico and in vitro when both the tissue mechanical and acoustic anisotropies were taken into account. First, synthetic 2D radio-frequency images of a phantom under quasi-static compression were simulated. The phantom was defined to be either transversely isotropic or isotropic to separately examine the effect of mechanical and acoustic anisotropies on strain estimation. The two anisotropic effects were further jointly incorporated in the simulated phantom with attenuation to resemble realistic ultrasonic images of soft tissues. Underperformance of the USI method was found in the fiber planes no matter whether the fiber was perpendicular or parallel to the ultrasound beam. The same configurations were further investigated in an in vitro porcine skeletal muscle phantom. Low-quality strain fields were found in the fiber planes with the myofibers either perpendicular or parallel to the beam. This confirmed the findings in the simulation study. Finally, the fundamental behind tissue anisotropy in ultrasound motion estimation was explored. The effect of mechanical anisotropy could be explained by strain filter. Sonographic signal-to-noise ratio (SNRs) was found to be related to the performance of the USI method under acoustic anisotropy. Speckle size seemed to have a limited impact on strain estimation, but further investigations would be required.