Confidence-enhanced flow trajectory imaging using speckle tracking with multi-angle plane waves

Abstract

Flow trajectory imaging, which tracks flow particles over time to map path-dependent trajectories, is a valuable tool for analyzing flow dynamics in pathological vascular conditions. Accurate trajectory estimation depends on robust vector flow imaging (VFI) methods to minimize cumulative errors. However, Doppler-based VFI methods are limited by aliasing and ensemble-sensitive estimation errors, while speckle tracking (ST) methods suffer from poor lateral velocity estimation due to beam diffraction. Additionally, existing trajectory imaging techniques lack a confidence metric to address unreliable estimates, leading to error propagation. To overcome these limitations, we propose a novel ST-based VFI technique leveraging multi-angle ultrafast plane wave imaging. Projected axial velocity components are estimated from baseband beamformed data for each steered plane wave and combined using a normalized cross-correlation (NCC)-weighted least-squares approach to generate velocity vectors. To ensure robust trajectory estimation, peak NCC values are incorporated to discard unreliable estimates and enhance confidence in the results. Simulation and experimental results demonstrate that the proposed method outperforms conventional ST and autocorrelation-based VFI by at least threefold and twofold in bias and standard deviation, respectively. It also shows better performance at low signal-to-noise ratios. In vivo measurements of the right carotid artery over one cardiac cycle highlight the ability of our method to visualize dynamic flow patterns and provide more accurate flow trajectory estimation compared to existing Doppler- and ST-based methods.