Effects of aberration on super-resolution ultrasound imaging using microbubbles

Abstract

Visualizating vasculature beyond the diffraction limit can be achieved using ultrasound super-resolution (USR). Typically, ultrasound (US) scanners model the target medium as homogeneous, assuming a constant speed-of-sound for time-of-flight based calculations. However, variations in US propagation velocity caused by varying tissue layers affect beamforming operations. This aberration is likely to have considerable effect on USR accuracy when imaging at depth. Here we investigate the effect of aberration on USR localization accuracy. Wave propagation through different tissue media was modelled from a linear transducer using k-Wave, and the resulting pulse-echo data from a point scatterer located at 45mm depth was beamformed. Media included a control homogeneous propagation medium, and aberrating media typical of liver imaging. Aberration effects were estimated using RMS arrival-time fluctuations (ATF) and energy-level fluctuations (ELF) at the scatterer location. Signals were extracted and localized with existing localization techniques. Results indicated USR localisation accuracy decreased with increasing aberration. Axial localisation errors reached 836.6 mm for an ATF 30.9 ns. Furthermore, errors increased with decreasing frequency from 4 to 3 MHz. The scale of these errors relative to micro-vascular structures of interest suggests that aberration will have considerable impact on USR performance and requires attention to ensure its success.