Evaluation of Performance Tradeoffs when Using Mechanically Swept 1-D Linear Arrays for 3-D DCE-US

Elahe Moghimirad,Zhiting Xu,Hong Ding,Jeffrey Bamber,Emma Harris
DOI: https://doi.org/10.1109/tuffc.2023.3268009
IF: 3.267
2023-01-01
IEEE Transactions on Ultrasonics Ferroelectrics and Frequency Control
Abstract:Dynamic contrast-enhanced ultrasound imaging (DCE-US) may be used to characterize tumor vascular perfusion using metrics derived from time-amplitude curves (TACs). The 3-D DCE-US enables generation of 3-D parametric maps of TAC metrics that may inform on how perfusion varies across the entire tumor. The aim of this work was to understand the effect of low temporal sampling (i.e., < 1 Hz) typical of 3-D imaging using a swept 1-D array transducer on the evaluation of TAC metrics and the effect of transducer motion in combination with flow on 3-D parametric maps generated using both plane wave imaging (PWI) (seven angles) and focused imaging (FI). Correlation maps were introduced to evaluate the spatial blurring of TAC metrics. A research ultrasound scanner and a pulse-inversion algorithm were used to obtain DCE-US. The 2-D (frame rate 10 Hz) and 3-D (volume rate 0.4 Hz) images were acquired of a simple wall-less vessel phantom (flow phantom) and a cartridge phantom. Volumetric imaging provided similar TACs to that of the higher 2-D sampling rate. Varying sweep speed and acceleration/deceleration had little influence on the 3-D TAC compared to 2-D for both FI and PWI. Sweeping motion and limited temporal sampling (0.4 Hz) did not change the spatial correlation of TAC metrics measured using FI, whereas a small increase in correlation across the cartridge phantom was observed for PWI. This was attributed to grating lobe artifacts, broad beam spatial blurring, and incoherent compounding caused by motion. Increased correlation will reduce the spatial resolution with which inhomogeneity of vascular perfusion can be mapped supporting the choice of FI for DCE-US.
What problem does this paper attempt to address?