Revisiting XDoppler estimator for high spatiotemporal resolution volumetric axial velocity measurement using row-column arrays

TL;DR

This paper extends the XDoppler estimator for ultrasound velocity measurement using row-column arrays, achieving higher accuracy, better sensitivity to slow flows, and reduced aliasing in 3D imaging.

Contribution

It introduces a novel velocity estimator based on phase cross-correlation of orthogonal RCA signals, improving upon traditional methods in accuracy and sensitivity.

Findings

Outperforms traditional phase-shift autocorrelator in accuracy

Provides higher Nyquist velocity, reducing aliasing

Demonstrates improved sensitivity to slow flows and dynamic blood flow tracking

Abstract

Accurate volumetric velocity estimation is crucial in ultrasound imaging for both diagnostic and therapeutic applications. Traditional ultrasound systems, though effective for two-dimensional imaging, face major limitations in 3D imaging due to hardware and computational demands. Row-column addressed (RCA) ultrasound probes offer a promising alternative by reducing hardware complexity, thereby reducing the gap between research prototypes and clinical systems. However, this typically comes at the expense of stronger sidelobes compared with fully populated matrix arrays, leading to reduced image contrast. Several approaches have been proposed to improve the contrast of power Doppler imaging, yet the accuracy and performance of velocity Doppler estimation have received comparatively little attention. In this study, we present a method that exploits the phase information from RCA row and column signals to derive a novel velocity estimator based on cross-correlation of orthogonal apertures. This extends the XDoppler scheme, initially developed for power Doppler imaging, to velocity estimation. The XDoppler estimator is shown to provide accurate measurements of axial velocities and to outperform the traditional phase-shift autocorrelator, while offering a theoretical Nyquist velocity twice as high. In vitro experiments further demonstrate enhanced sensitivity to slow flows and reduced bias in flow rate estimation. In vivo data from a carotid artery confirm the reduced sensitivity to aliasing and reveal the ability to track dynamic blood flow velocity changes associated with arterial pulsatility. These findings suggest that the XDoppler velocity estimator could improve volumetric velocity imaging in clinical contexts.