Optimizing the quantitative in vivo imaging for longitudinal studies in rat brain using FDG and microPET

FDG-PET imaging was performed in rats using a small animal PET scanner. The purpose was to improve the accuracy and precision of quantitative rat brain PET imaging and make longitudinal studies feasible. We evaluated bed position(s), scan time, movement correction, image reconstruction algorithms, input function (IF) derivation and methods for quantification of cerebral metabolic rate of glucose (CMRG, mumol/min/100g). The optimum protocol includes: (1) acquisition of PET data in two-bed positions: 0-45 minutes centered at the heart for determination of IF and 45-65 minutes centered at the brain; (2) minimization of head movements with a restraining device; (3) a single tail vein blood sample to increase accuracy of the image-derived IF; (4) application of FBP reconstruction to the heart data and both FBP and MAP reconstructions to the brain data and (5) application of a CT-based attenuation correction. We generated CMRG parametric images by applying the Sokoloff´s operational equation to the FBP brain images. The IF required by the operational equation was derived from the FBP heart images using a modified factor analysis program. Regions of brain were drawn on the MAP-reconstructed ´anatomical images´ and superimposed on the parametric images to yield mean CMRG values. CMRG values were comparable to those from a FDG 3-compartmental model fit to the time activity curves of the brain regions.