Super-Resolution Ultrasound Imaging of Rat Kidneys before and after Ischemia-Reperfusion

Abstract

In vivo visualization of the microvasculature is feasible with super-resolution ultrasound imaging (SRI), but the method needs more affirmative data before clinical use. The kidneys have a rich vasculature, and microvascular dysfunction decreases the kidney function. Therefore, detection of subtle renal microvascular changes could benefit patients with renal disease. We hypothesized that our SRI setup can visualize the microvascular network of two healthy rat kidneys and subsequently demonstrate microvascular flow changes immediately after ischemiareperfusion. The left kidney of two male Sprague-Dawley rats was scanned during laparotomy using a customized BK5000 scanner, an X18L5s transducer, and interleaved contrast and B-mode sequences with focused beam transmission. Images were acquired over 10 min using SonoVue (1:10) as the contrast agent. After a baseline scan, one rat had the renal vein clamped, while the other rat had the renal artery clamped, both for 45 min. The kidneys were rescanned immediately after clamp release and after 60 min of reperfusion. Motion correction was applied before microbubble (MB) detection. The characteristic renal microvascular structure was visualized with anatomical distinction between the dense cortical vascular network and the straight vessels of the medulla. Immediately after vein clamp release, almost no MBs perfused the medullary vessels. After 60 min of reperfusion, MBs refilled the renal vascular bed, but with a slower velocity compared with the baseline scan. The results after artery clamping were subtler, with more MBs perfusing the medulla immediately after clamp release compared with the vein clamping. The results imply that this SRI setup can evaluate different stages of ischemic kidney disease in rats as it can visualize the entire renal vascular bed and differentiate the pattern of reperfusion in two types of ischemic injury.

Current Development and Applications of Super-Resolution Ultrasound Imaging

“…The results showed that the blood flow speed in the injured rat kidney (<10 mm/s) was much lower than that in the healthy kidney (~30 mm/s) [ 80 ]. Similar results was achieved by Andersen et al, suggesting that blood flow in the renal microvasculature was measured to be slower after ischemia and reperfusion by SRU imaging [ 81 ]. Studies that demonstrated the feasibility of SRU for identifying microvascular alterations during the disease progression, with a larger group of animals and histological verifications, were performed by Chen et al [ 45 ].…”

Three-Dimensional Super-Resolution Imaging Using a Row–Column Array

“…This is sufficient to employ speckle tracking [31] in 3-D to yield and compensate for the motion as described for SA flow imaging [32,33]. Although many schemes use very high frame rates with thousands images per second [6,9] , it has been shown that a conventional linear array scan with frame rates at 54 Hz can yield excellent super resolution images with both motion estimation [34] and quantification of flow [35]. The 154 Hz volume rate should, thus, be sufficient for in-vivo imaging.…”

“…Such arrays can directly be used on modern ultrasound consoles with few modifications in the beamforming. The approach can fairly easily be translated to clinical use by modifying our current 2-D super resolution pipeline to include searches and localizations in 3-D [34,35]. The motion correction schemes developed for 2-D imaging and needed for in-vivo imaging can then also be applied [34].…”

“…A first in-vivo target would be to scan a rat kidney as performed in [34,35]. The acquisition time was between 1 and 10 minutes for a 2-D image, with 1 minute giving an overall rough view of the vasculature and 10 minutes giving precise quantitative data for the blood flow.…”

Deep Learning Models for Fast Ultrasound Localization Microscopy

“…ULM is achieved by localizing gas-filled microbubbles (MBs) that are injected into the bloodstream and accumulating their centroids from multiple frames in an image. The resulting super-resolution images can be used to diagnose early-stage cancer [7], ischemic kidney disease [8], and diabetes [9].…”