Hepatic Portal Venous Perfusion Imaging Using Vessel-Labeling Super-Resolution Ultrasound

Objective

Blood flow imaging and perfusion assessment of the hepatic portal vein are critical for the diagnosis of several liver diseases, including cirrhosis, primary and metastatic liver tumors. However, perfusion imaging of the portal vein is challenging due to the unique dual blood supply system of the liver.

Methods

We developed a novel method for specific perfusion imaging of the portal vein and downstream vessels, which was validated on healthy mice (n = 4). The right lobe of the liver in healthy mice was sequentially imaged using ultrafast plane-wave Doppler imaging and vascular labeling. In each experiment, mice were first injected with phase-change nanodroplets (PCNDs), followed immediately by ultrafast Doppler imaging to determine the imaging section and locate portal vein branches. Through an interactive process, portal vein branches were selected by mouse click for data acquisition of vessel-labeling ultrasound (VLUS) based on PCNDs. Subsequent arrival time calculations and super-resolution ultrasound (SRUS) imaging were performed offline. To demonstrate the specificity of the proposed method for vascular imaging, one mouse was injected with Sonovue microbubbles for plane-wave ultrasound data acquisition and microbubble-based VLUS data acquisition. All imaging experiments were conducted on the Verasonics (Kirkland, WA, USA) Vantage 256 ultrasound system using an L22-8v linear array transducer with a center frequency of 15.625 MHz. The multi-angle coherent compounding plane-wave acquisition frame rate was 500 Hz.

Results

Imaging results from healthy mice (n = 4) demonstrated that VLUS was able to label different branches of the hepatic portal vein and specifically image downstream vessels. Analysis of the in vivo results at different spatial scales showed that the brightness of the downstream perfusion area was significantly enhanced after labeling started, while there was no significant difference in image brightness before the labeling started and after it ended. By analyzing the acoustic field distribution at the focal point, the full width at half maximum in the x₁ and z₁ directions were 98.56 μm and 526.68 μm, respectively. Along the propagation path of the focused beam (outside the labeling area), no significant activation of the PCNDs was observed (p < 0.0001). Combined with SRUS technology, the resolution of the VLUS portal vein imaging results was further enhanced. The time-intensity curves of the downstream regions of interest indicated that VLUS provided a step input signal to the downstream vessels. Based on the arrival time of the step point in the time-intensity curves, the arrival time distribution map of the downstream vessels relative to the labeling point could be calculated.

Conclusion

We propose a novel method for hepatic portal vein perfusion imaging based on VLUS. In vivo experiments, simulation results and statistical analysis demonstrate that this method is able to accurately label portal vein vessels with millimeter-level precision, enabling specific high-resolution imaging and precise, non-invasive measurement of the downstream perfusion area. By combining VLUS with SRUS technology, the resolution of the portal vein imaging results can be further enhanced.