Cross-comparison of PET/MRI Scanner and microPET Linearity with Autoradiography and Direct Assay To enable quantitative comparison of images across PET and QAR

PET/MRI scanner images of syringe phantoms were decay corrected and normalized for scan duration prior to analysis. The first linearity phantom image acquired during a given study session was used to convert image intensities to units of absolute activity concentration (kBq/mL). Rod image intensities from the PET/MRI image, obtained from cylindrical volumes of interest axially centered on each rod (diameter = 40% of the rod diameter, length = 8.3 mm), were plotted against activity concentrations measured by well counting of samples taken from the phantom. Plots were fitted by linear regression, taking into account the random variability of both the measured independent and dependent variables (1). The resulting regression coefficients were used to calculate activity concentrations for all images acquired during the corresponding study session. This calibration procedure incorporates phantom-specific compensation for attenuation, scatter and partial volume effects into the calibration coefficients.