Abstract
Kinetic analysis based on compartment model is one of the effective pharmacokinetics methods for quantitative physiologic analyses of nuclear medicine data, such as positron emission tomography data. A typical tracer kinetic modeling with an arterial input function is to represent tracer up take in brain tissue as a three compartment model including rate constants. In this study, the effect of the approximation of the input function on estimating the rate constants is numerically investigated by using FDG-PET data. Imaging the regional metabolism of glucose in the human brain uses ^<18>F-FDG, because FDG is transported and phosphorylated by the same pathways as glucose. Numerical simulation results indicate that the estimation of rate constants are depend on the approximation of the input function.
