The parameters used in segmentation were: alpha=0.04, a parameter that controls the weight of smoothing item of the image features; and iteration was set to 60, such values were sufficient for detection of (R)-(+)-BAY K 8644 kidney outline in this experiment. Afterwards, dynamic images of the segmented regions were registrated to reduce motion. Then, a slice which covering the largest possible parenchyma during corticomedullary phase was used for cortical ROI drawing to obtain the tissue signal intensity curves. The size of ROIs for all the kidneys was 590 pixels in average. The arterial input function was determined by drawing ROI in the slice that a branch of renal artery was obviously seen, that is, the ROI was placed within the aorta distal to the branch of the renal artery. To improve the robustness of AIF and reduce the inflow artifacts, one more slice above was also selected for averaging. A 363 pixel size ROI was used. Furthermore, the tail of the AIF was fit to a biexponential decay to reduce respiratory motion related noise. In present study, a modified two-compartment model with impulse residue function is implemented, and Monte Carlo simulation results indicate the reliability of the proposed model in parameters estimates. The main result of this study is that the GFR and RPF measurements are in close agreement with literature results in rabbits. Moreover, the new model is valid in detecting GFR alterations in diseased kidneys. The close agreement with literature and the valid assessments of discriminating healthy and dysfunction kidneys indicate that the proposed 2C-IRF model provides a useful AZD3965 method for quantitative measurements in kidneys, and the difference in accuracy is relevant. Our results show that the new model is feasible and not only achieves reliable measurements of renal function, but also be capable in detecting of GFR alterations in dysfunction kidneys. Using the predefined impulse residue function, which considers the transit time and dispersion effect of contrast agent in kidney, our new model could characterize the distribution of contrast and elucidate physiological meaning better, as well as improves the model reliability with reduced bias in measurements of parameters.