Effect of aperture size on plane wave ultrasound strain estimation

Abstract

Assessment of axial strain and lateral strain is of major importance for studying the mechanical properties of tissue. In this study the effect of beamforming on the accuracy of axial and lateral strain estimation was investigated. Radio frequency (RF) data were generated for non steered plane wave insonation using an eccentric plaque vessel simulated with Field II and experiments with a homogeneous vessel phantom using a Verasonics, Inc. (Redmond, WA). Beamforming was executed by employing two techniques, delay-and-sum (DAS) and Fourier based reconstruction (FBR). In the course of DAS beamforming, RF-data were processed with different f-number settings and its impact on 2D cross-correlation based strain estimation was investigated. Subsequently, beamforming was performed with FBR. Strain estimation obtained from FBR and non apodized (NAPOD) DAS beamformed RF data were compared. In the simulation study, strain estimation was analyzed by calculating the root-mean-square error (RMSE) with respect to the theoretical strain. Simulation results showed that lateral strain estimation can be improved by decreasing the f-number, but that there was almost no influence on axial strain. The strain estimation in phantom experiments was evaluated in terms of the elastographic signal-to-noise ratio (SNR <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">e</sub> ). The experimental results also showed that lateral strain tends to improve for decreasing f-number, but hardly affects axial strain. Comparing the strain results between NAPOD DAS and FBR, FBR provides a reduction of 0.13% RMSE for axial strain and NAPOD DAS showed a reduction of 0.27% in lateral strain. In conclusion, the f-number influences lateral strain estimation, but seems to have a negligible effect on axial strain estimation. FBR is more accurate for axial strain estimation, whereas NAPOD DAS is more accurate for lateral strain estimation.