Measuring glomerular blood transfer rate in kidney using diffusion-weighted arterial spin labeling

To propose a two-compartment renal perfusion model for calculating glomerular blood transfer rate ( $$ ) as a new measure of renal function.

The renal perfusion signal was divided into preglomerular and postglomerular flows according to flow velocity. By analyzing perfusion signals acquired with and without diffusion gradients, we estimated $$ , the blood transfer rate from the afferent arterioles into the glomerulus.

A multislice multidelay diffusion-weighted arterial spin labeling sequence was applied to subjects with no history of renal dysfunctions. In the multiple b-value experiment, images were acquired with seven b-values to validate the bi-exponential decays of the renal perfusion signal and to determine the appropriate b-value for suppressing preglomerular flow. In the caffeine challenge, six subjects were scanned twice on the caffeine day and the control day. The $$ values of the two dates were compared.

The perfusion signal showed a bi-exponential decay with b-values. There was no significant difference in renal blood flow and arterial transit time between caffeine and control days. In contrast, cortical $$ was significantly higher on the caffeine day (caffeine day: $$ min $$ control day: $$ min $$ ). These results were consistent with those from the literature.

We showed that the perfusion signal consists of two compartments of preglomerular flow and postglomerular flow. The proposed diffusion-weighted arterial spin labeling could measure the glomerular blood transfer rate ( $$ ), which was sensitive enough to noninvasively monitor the caffeine-induced vasodilation of afferent arterioles.