Online 4D Ultrasound-Guided Robotic Tracking Enables 3D Ultrasound Localisation Microscopy with Large Tissue Displacements

TL;DR

This work tackles the challenge of performing Ultrasound Localisation Microscopy (ULM) in moving tissues, where respiration-induced motion can disrupt field-of-view and degrade super-resolution reconstructions. It introduces an online framework that couples high-frame-rate 4D ultrasound with real-time robot-assisted probe tracking, operating asynchronously to maintain the imaging volume while microbubbles are localized and tracked for SR imaging. The approach is validated on a moving microvasculature phantom, demonstrating real-time target tracking up to 20 mm and an imaging volume rate of 85 Hz, with post-processing yielding 3D SR images and FSC-based resolutions around tens of micrometers. This indicates a meaningful step toward SR ULM in organs with large motion, enabling more reliable vascular imaging without breath-hold requirements and enhancing potential clinical applicability.

Abstract

Super-Resolution Ultrasound (SRUS) imaging through localising and tracking microbubbles, also known as Ultrasound Localisation Microscopy (ULM), has demonstrated significant potential for reconstructing microvasculature and flows with sub-diffraction resolution in clinical diagnostics. However, imaging organs with large tissue movements, such as those caused by respiration, presents substantial challenges. Existing methods often require breath holding to maintain accumulation accuracy, which limits data acquisition time and ULM image saturation. To improve image quality in the presence of large tissue movements, this study introduces an approach integrating high-frame-rate ultrasound with online precise robotic probe control. Tested on a microvasculature phantom with translation motions up to 20 mm, twice the aperture size of the matrix array used, our method achieved real-time tracking of the moving phantom and imaging volume rate at 85 Hz, keeping majority of the target volume in the imaging field of view. ULM images of the moving cross channels in the phantom were successfully reconstructed in post-processing, demonstrating the feasibility of super-resolution imaging under large tissue motions. This represents a significant step towards ULM imaging of organs with large motion.

Figures (11)

• Figure 1: Study diagram. (a) Experiment setup: a matrix array was fixed on the tool base of a robot arm by a 3D printed holder; the robot and the Verasonics system were connected to the same desktop (host); the robot and the host were communicated via TCP/IP; a phantom with crossed tubes was placed on motorised translation stages that were controlled by a laptop; MBs were injected into the phantom by a syringe pump. (b) Phantom was moved in six different motion profiles. (c) Phantom motions were given in three different directions along the matrix array. (d) Half-Full-Half aperture transmission and sub-aperture reception were used for the multiplexed matrix array. (e) Designed asynchronous real-time motion tracking timeline, consisting of data acquisition, image processing and robot control. (f) Motions remained in images after real-time motion tracking were analysed to test the motion tracking performance for different target motion speed levels. (g) The feasibility of ULM reconstruction was verified when target motion was double the probe aperture size.
• Figure 2: Experimental Setup, diagram of which is shown in Fig. \ref{['fig:Diagram']}(a). The left side shows the overall setup. The right side shows the aperture of ultrasound array and the home-made microvasculature phantom.
• Figure 3: Camera videos and corresponding lateral-depth (X-Z) maximum intensity projection (MIP) of reconstructed 4D ultrasound images with or without robotic tracking. The object in the dashed yellow box of the MIP ultrasound image is the landmark ball scatter in the phantom and the lobes in the solid yellow box are the structure supporting the ball. This landmark moved out of the imaging FoV when the probe is fixed but kept in the FoV when the robotic tracking is on. The imaging FoV is centred on the array central axis with a lateral length of 15 mm. Log compression and a dynamic range of 20 dB was used in the visualisation. The red line denotes a reference fixed in the camera view of each row.
• Figure 4: Temporal average of reconstructed B-mode volume Maximum Intensity Projection (MIP) of the landmark to demonstrate difference with robotic tracking or not. The images with the robot tracking are brighter than those without the robot tracking, meaning more frames were overlapped with the robot tracking.
• Figure 5: Temporal average of reconstructed B-mode volume MIP of the landmark when moving target along probe’s lateral direction at different speeds. Images were less overlapped among frames when target moved faster.