Skip to Main Content
Physiologic parameters such as glucose metabolic rate (GMR), perfusion and cardiac output (CO) can be estimated by performing quantitative analysis using PET dynamic images. The measurement of the image derived arterial input function (IF) and the tissue time activity curve (TAC) can be affected by partial volume effect (PVE). Because of partial volume, the estimate of different physiological parameters can be severely biased. The main goal of this work was to evaluate the effects of an image reconstruction based partial volume correction (PVC) method of small animal PET images on metabolic rate and perfusion on a pixel-based analysis and on cardiac output. The proposed PVC method is based on Point Spread Function (PSF) modeling in the reconstruction scheme. Dynamic mouse heart images were created using the Moby phantom. IF and TAC were simulated for one and two compartmental models and different radiotracers in order to take into account the different positron range. Images were simulated using different sets of rate constants and different noise levels. In order to obtain an estimate of the probability distribution of each kinetic parameter from each pixel value, bootstrap resampling with replacement was applied. The coefficient of variation and the bias of the mean of the distribution with respect to the theoretical value were estimated. Pre and post-correction parametric images of GMR and perfusion and the relative errors show that the image reconstruction PVC reduces errors with respect to the theoretical values in each pixel. The percentage error of CO from uncorrected and corrected images with respect to the theoretical value was 13.5% and 2.3%, respectively, for 18F-FDG study. Regarding images acquired using 82Rb, CO estimate was equal to 34.6 % and 23.2 % for non PVC and PVC images, respectively. In conclusion, pixel based PVC allows to obtain a better measure of the radiotracer concentration in the heart and a more accurate estimation- - of the phisiological parameters.