European Journal of Radiology 54 (2005) 124–135 Parametric and quantitative analysis of MR renographic curves for assessing the functional behaviour of the kidney N. Michoux a,∗ , X. Montet a , A. Pechère b , M.K. Ivancevic a , P.-Y. Martin c , A. Keller a , D. Didier a , F. Terrier a , J.-P. Vallée a a Unité d’imagerie Numérique, Département de Radiologie, Hˆ opital Universitaire de Genève, Rue Micheli-du-Crest 24, CH-1211 Geneva 14, Switzerland b Policlinique de Médecine et Service d’Endocrinologie, Hˆ opital Universitaire de Genève, Rue Micheli-du-Crest 24, CH-1211 Geneva 14, Switzerland c Service de Néphrologie, Hˆ opital Universitaire de Genève, Rue Micheli-du-Crest 24, CH-1211 Geneva 14, Switzerland Received 7 January 2004; received in revised form 8 April 2004; accepted 13 April 2004 Abstract The aim of this study was to refine the description of the renal function based on MR images and through transit-time curve analysis on a normal population and on a population with renal failure, using the quantitative model of the up-slope. Thirty patients referred for a kidney MR exam were divided in a first population with well-functioning kidneys and in a second population with renal failure from ischaemic kidney disease. The perfusion sequence consisted of an intravenous injection of Gd-DTPA and of a fast GRE sequence T1-TFE with 90 ◦ magnetisation preparation (Intera 1.5 T MR System, Philips Medical System). To convert the signal intensity into 1/T1, which is proportional to the contrast media concentration, a flow-corrected calibration procedure was used. Following segmentation of regions of interest in the cortex and medulla of the kidney and in the abdominal aorta, outflow curves were obtained and filtered to remove the high frequency fluctuations. The model of the up-slope method was then applied. Significant reduction of the cortical perfusion (Q c = 0.057 ± 0.030 ml/(s 100 g) to Q c = 0.030 ± 0.017 ml/(s 100 g), P< 0.013), of the medullary perfusion (Q m = 0.023 ± 0.018 ml/(s 100 g) to Q m = 0.011 ± 0.006 ml/(s 100 g), P< 0.046) and of the accumulation of contrast media in the medulla (Q a = 0.005 ± 0.003 ml/(s 100 g) to Q a = 0.0009 ± 0.0008 ml/(s 100 g), P< 0.001) were found in presence of renal failure. High correlations were found between the creatinine level and the accumulation Q a in the medulla (r 2 = 0.72, P< 0.05), and between the perfusion ratio Q c /Q m and the accumulation Q a in the medulla (r 2 = 0.81, P< 0.05). No significant difference was found in times to peak between both populations despite a trend showing T a the time to the end of the increasing contrast accumulation period in the medulla, arriving later for renal failure. Advances in MR signal calibration with the building of quantitative model such as the up-slope allow to assess kinetic and haemodynamic and functional parameters of the diseased kidney. © 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: MR dynamic images; Renal circulation; Contrast media; Perfusion; Morphological analysis 1. Introduction Using dynamic MRI, sequential images of the kidney can be acquired during the injection of intravenous MR contrast media, offering a unique tool for assessing the state of the kidney. Indeed, anatomy [1] as well as function [2] can be investigated in a single exam with a spatial and tem- poral resolution out-performing that of nuclear medicine techniques. Particularly, the high spatial resolution and the high temporal resolution of new MRI sequences permit a ∗ Corresponding author. Tel.: +41 22 3723311; fax: +41 22 3727072. E-mail address: nicolas.michoux@sim.hcuge.ch (N. Michoux). regional analysis of the kidney function by a separate inves- tigation of the cortical and medullary perfusion. Numerous studies examined the potential of dynamic MRI to assess the normal kidney perfusion [3–6], the effect of renal ob- struction [7], of kidney transplant rejection [8] and renal artery stenosis [9–12]. Nevertheless, relatively few groups studied the effect of renal failure on the renal transit-time curves. Vallée et al. [13] found in seven patients with an advanced renal failure a decrease of the cortical perfusion by comparison to patients with a normal renal function. Dalla-Palma et al. [14] observed a delay of the cortical time to peak in 33 patients with renal failure. However, no infor- mation on the cortical and medullary perfusion as well as on the shape of the transit-time curves was available in this 0720-048X/$ – see front matter © 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2004.04.010