Repeatability of renal arterial spin labelling MRI in healthy subjects

Arterial spin labelling (ASL) can be used to measure renal perfusion non-invasively. The aim of this study was to determine the repeatability of this technique in healthy kidneys to vindicate its use in clinic. Two groups of healthy volunteers were imaged two different days to assess intra- and inter-session repeatability. Oblique-coronal data volumes were acquired on a 1.5 T scanner with a dedicated abdominal 32-channel body phased array coil. ASL was performed using a multi-TI FAIR labelling scheme and 3D GRASE imaging module. Background suppression and respiratory triggering were used. T-1 maps of the kidney were acquired using the same sequence with background suppression disabled. For the group with multiple intra-session ASL measurements, the average cortical perfusion was 197 mL min(-1)100 g(-1) and average cortical T-1 was 1265 ms. For both perfusion and T-1 the variation shown by the within-subject standard deviation (SDws) (14.6 mL min(-1)100 g(-1) and 33.4 ms) and coefficient of variation (CVws) (7.52 and 2.69%, respectively) was small for all the analyses carried out. Bland-Altman plots were also used to visualise the variation between the same parameters collected from the different scanning sessions in both groups, and demonstrated good reproducibility. We have shown that in healthy volunteers, ASL parameters are repeatable over a short and long period. This supports the overall aim of using ASL in the clinic to assess longitudinal renal perfusion changes in patients.