Kidney Modeling Using a Polynomial Function

In the analysis methods for dynamic positron emission tomography (dPET) data, compartment model has been recognized as the gold standard. But, the studies on the development and validation of a model are relatively complex and time-consuming. Hence, fast algorithms become frequently used for analysing dPET data and parametric images. The purpose of this study is to simulate the 18F-Fluorodeoxyglucose (18F-FDG) concentration in the kidneys from different individuals by applying polynomial regression function and quantitatively describe the fitted data with R-squared correlation coefficient. Four subjects had been injected intravenously with 18F-FDG dose of 302.29 ± 18.75 MBq prior to dPET/CT kidney scanning. The time–activity curve (TAC) of abdominal aorta and kidneys were plotted based on the drawn region of interest (ROI) in each frame of image acquisitions. The 18F-FDG concentration was measured by averaging the values of entire voxel within the ROI. Four sets of PET data were entered into the polynomial function of MATLAB R2015a software to implement and analyse the model for fitting the observed data. The best fit was stated by a 15th-degree polynomial function for both sides of the kidneys. The mean R-squared for the right kidney is 0.85 while 0.86 is for the left kidney. Therefore, the model developed can simulate the distribution of 18F-FDG concentration in the kidneys using dPET data.