A strategy for removing the bias in the graphical analysis method.

摘要: Top of pageAbstract The graphical analysis method, which transforms multiple time measurements plasma and tissue uptake data into a linear plot, is useful tool for rapidly obtaining information about the binding radioligands used in PET studies. The strength method that it does not require particular model structure. However, bias introduced case noisy resulting underestimation distribution volume (DV), slope obtained from method. To remove bias, modification developed by Feng et al. (1993), generalized least squares (GLLS) provides unbiased estimates compartment models was used. one GLLS has relatively simple form, to estimate DV directly as smoothing technique more general classes structures. In latter case, applied two parts, is, set parameters determined times 0 T1 second end time. curve generated these sets then input This been tested using simulations similar ligand [11C]-d-threo-methylphenidate (MP, = 35 mL/mL) 11C raclopride (RAC, 1.92 compared with examples image same tracers. noise based on counting statistics through half-life isotope scanning Five hundred at each level were analyzed. Results (DV) (DVG), nonlinear (NLS) (DVNLS), one-tissue (DVF), part followed (DVFG) compared. DVFG found increase somewhat increasing some high levels no could be obtained. intermediate provided good estimation true DV. extended use reference place function generate ratio (DVR) region. A linearized form simplified Lammertsma Hume (1996) DVR (DVRFL) DVRFG (determined "smoothed" DVFG) method. Keywords: Positron emission tomography, Kinetic modeling, Graphical analysis, Distribution