Generation of Functional Images of The Brain Trapping Constant for α-[11C]Methyl-L-Tryptophan Using Constrained Linear Regression

Abstract

The main objective of described work was a generation of the functional images of the brain trapping constant (K*; μL/g/min) of -methyl-L-tryptophan, an index of 5-HT synthesis, which under some assumptions is related to the serotonin synthesis. Comparisons of the regional K* calculated by the Patlak approximation, non-linear fitting and the linearized form of the non-linear operational equation were made and discussed. In addition a contrast between the white and gray matter K* values was evaluated by different methods. Results presented suggest that the linearized form of the operational equation yields the best gray to white matter contrast. It was also shown that with this calculation approach, as was shown before for the Patlak approximation, the use of the venous sinus-venous blood normalized input function instead of the arterial input function is satisfactory. Also results show that the error of the K* estimates is smaller than those in the Patlak estimates, when the linearized solution of the model equation is used. Simulation results indicate that the coefficient of variation for K* is smaller than some errors for the equation parameters.