The importance of AIF ROI selection in DCE-MRI renography: Reproducibility and variability of renal perfusion and filtration

PurposeThe aim of this study was to investigate (a) the effect the choice of the region of interest (ROI) defining the aortic input function (AIF) has on the estimation of renal perfusion and filtration in dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI) renography, and (b) the reproducibility of these parameters. Using renal DCE-MRI and a three-compartment model analysis, this work evaluated the effect two different AIFs, derived from variable sized ROIs in the aorta, has on calculating DCE-MRI renal perfusion and filtration values in a group of healthy adult volunteers who underwent two consecutive renal DCE-MRI studies.MethodsFifteen healthy volunteers underwent two DCE-MRI studies under similar physiological conditions. Oblique-coronal DCE-MRI data volumes were acquired on a 1.5 T Siemens Avanto scanner with a 3D-FLASH pulse-sequence (TE/TR = 0.53/1.63 ms, flip angle = 17°, acquisition matrix = 128 × 104 voxels, strong fat saturation, PAT factor = 2 (GRAPPA) and 400 mm × 325 mm FOV). Each dynamic dataset consisted of 18 slices of 7.5 mm thickness (no gap) and an in-plane resolution of 3.1 mm × 3.1 mm, acquired every 2.5 s for not less than 5 minutes. During the MR scan a dose of 0.05 mmol (0.1 mL) kg−1 body weight of dimeglumine gadopentetate (Magnevist) was injected intravenously (2 mL s−1 injection rate), followed by a 15 mL saline flush at the same rate, using a MR-compatible automated injector (Spectris).Two AIFs were defined for each volunteer by drawing two ROIs in the aorta for each study. Renal perfusion and glomerular filtration rate (GFR) values were then calculated for each of the AIFs using a modified Tofts Renal Model (TRM). Both renal perfusion and GFR were expressed in mL min−1 100 mL−1 of tissue.Results and conclusionInter-individual reproducibility tests for renal perfusion and glomerular filtration rate showed that the size of AIF ROIs significantly affects calculated values of perfusion and GFR (p-values <0.02). No significant differences were observed when comparing perfusion and GFR values in the same volunteer between scans performed on different days (p-values >0.22). From our study we conclude that while DCE-MRI derived indices of renal function are reproducible in the same individual when imaged on different days, the size of the aortic ROI and hence the AIF has a significant influence on calculated renal perfusion and GFR values. Currently there is no accepted standard for drawing the aortic ROI and no standardized approach for the AIF definition in renal DCE-MRI studies.