Reproducibility of the aortic input function (AIF) derived from dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) of the kidneys in a volunteer study

Eur J Radiol. 2009 Sep;71(3):576-81. doi: 10.1016/j.ejrad.2008.09.025. Epub 2008 Nov 11.


Purpose: The aim of this study was to investigate the maximum height, area under the curve (AUC) and full width at half maximum (FWHM) of the aortic input function (AIF) in renal dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI) studies. We evaluated the significance of choice of size for regions of interest (ROI) in the aorta, reproducibility and inter-observer agreement of AIF measurements in healthy volunteers for renal DCE-MRI studies.

Methods: Fifteen healthy volunteers (nine males, six females), mean age 28.8 years (range 23-36 years), underwent two DCE-MRI kidney studies under similar 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 degrees, acquisition matrix=128 x 104 voxels, strong fat saturation, PAT factor=2 (GRAPPA) and 400 mm x 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 x 3.1 mm, acquired every 2.5 s for >5 min. 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). For each DCE-MRI study two observers each drew two ROIs in the abdominal aorta. Both ROIs were 3 voxels in width and had the same inferior limit (just above the emergence of the renal arteries from the aorta) but had different heights (4 voxels for one ROI and 10 voxels for the other). The dimensions, position and time of drawing the ROIs in the dynamic study were standardised between observers prior to data analysis. Mean signal intensities measured in the ROIs were plotted over time, representing the AIF. For each study, AIF 1 was derived from ROI 1 and AIF 2 was derived from ROI 2.

Results and conclusion: Paired t-tests for inter-observer comparison on the pooled 30 DCE-MRI studies, showed good correlations (correlation coefficients >0.85) with no significant differences (p-values >0.82) when comparing the peak value, AUC and FWHM of the AIFs. Thus the results were operator independent. The size of the aortic ROIs significantly affected all measured parameters of the AIF (p-values <0.039). However, correlation coefficients when comparing AIF 1 and AIF 2 were high for all evaluated AIF parameters (correlation coefficients >0.88), indicating a similar shape and temporal dynamic of the passage of the contrast agent through the aorta. When comparing the intra-individual DCE-MRI studies for each volunteer all AIF parameters had p-values >0.22 and correlation coefficients <0.82, with the exception of the FWHM, which had a correlation coefficient of 0.96 showing a significant variation in AIF parameters in the same volunteer on different days.

MeSH terms

  • Adult
  • Algorithms
  • Aorta, Abdominal
  • Computer Simulation
  • Contrast Media / pharmacokinetics
  • Female
  • Gadolinium DTPA / pharmacokinetics*
  • Humans
  • Image Enhancement / methods
  • Image Interpretation, Computer-Assisted / methods
  • Information Storage and Retrieval / methods*
  • Kidney / anatomy & histology*
  • Kidney / metabolism*
  • Magnetic Resonance Imaging / methods*
  • Male
  • Models, Biological
  • Perfusion Imaging / methods*
  • Reproducibility of Results
  • Sensitivity and Specificity
  • Young Adult


  • Contrast Media
  • Gadolinium DTPA