Performance Evaluation of Displacement Estimators for Real-Time Ultrasonic Strain and Blood Flow Imaging With Improved Spatial Resolution

Abstract

Ultrasonic strain and blood flow imaging has great potential to improve early tumor detection. Both methods have the same task of estimating echo signal displacements. To be able to resolve tissue stiffness and vascularity of tumors while they are still small, spatial resolution must be improved. Moreover, ultrasonic real-time imaging is vital in clinical practice. In this paper, we evaluate the estimation performances of one novel and several established displacement estimators with real-time capability and compare these with results from cross-correlation-based standard estimators when echo signals that are typically available with a commercial ultrasound scanner set to yield the highest spatial resolution are used. Estimation errors are investigated as a function of sampling rate, signal-to-noise ratio, displacement and strain magnitudes, and frequency-dependent tissue attenuation. Results are from simulations using the FIELD II program to generate ultrasonic echo signals and tissue-mimicking phantom experiments using a Voluson 730 medical ultrasound scanner (GE medical systems Kretztechnik, Austria) that is equipped with a digital research interface that enables the export of beam-formed radio frequency echo data.