Abstract
1873
Objectives A mathematic model for kinetics of radiopharmaceutical of 18F-fluorodeoxyglucose (FDG) in humans was developed and was used to estimate the radiation absorbed dose for patients in nuclear medicine by the MIRD Committee. However, due to the residence time used in the model was derived from different ethnic subjects who were administered intravenously by 18F-FDG, a large variation of the model parameters evaluated propagates a high uncertainty in dose estimation. An uncertainty of the absorbed dose of 18F-FDG using current biokinetic/dosimetric models and the new voxel phantom S values was analyzed.
Methods The sources of uncertainty of all parameters of 18F-FDG model were analyzed, and the ranges and the types of distributions of each parameter were evaluated. The Latin hypercube sampling technique was used to sample the parameters. The absorbed dose was calculated by using the residence times modeled, coupling with the S values calculated by using the voxel phantoms for reference female and male. A computer program was written to calculate doses. Parameter sensitivity was analyzed by combining the model input and output.
Results The uncertainty of the organ absorbed dose was given. The sensitivity of each model parameter on the model prediction at different times, the residence times and the organ absorbed doses after intravenous administration were analyzed.
Conclusions The most influence parameters on the organ absorbed doses in the 18F-FDG model were indicated. The exploitation of the parameter uncertainty and sensitivity of the model for the biokinetic data acquisition in clinic practice and for the individual dosimetry of patients is discussed.
Research Support This work is supported by the EU Project MADEIRA
- © 2009 by Society of Nuclear Medicine