Abstract
1376
Objectives We have developed a novel scatter compensation that uses individual photons’ position and energy information (ESN-ESC) and compared its accuracy to that of the standard scatter correction using only the position information (SN-SC) in realistic Monte Carlo simulations of a list-mode dynamic cardiac PET study of a large patient.
Methods Our scatter correction incorporates the position and energy probability density functions of primary (unscattered) and scatter coincidences in the scatter sinogram normalization process and the iterative reconstruction algorithm. We simulated dynamic list-mode PET acquisitions of a large patient (BMI~50) using a realistic Monte Carlo simulation of the Gemini TF PET scanner and the XCAT phantom. We modeled the 18F-based BMS747158 cardiac tracer. Ten noise realizations and two energy resolutions (11.5% and 20% at 511 keV) were simulated, resulting in 20 dynamic studies. These were reconstructed and corrected for scatter independently using the SN-SC and the ESN-ESC strategies as well as the primary coincidences alone (reference). For every dynamic volume the left ventricle (LV) and myocardium (MYO) time activity curves were extracted using masks of the LV and the MYO, which were known exactly in these simulations, and were fitted with a 2-compartment kinetic model in order to assess the impact of the scatter correction and energy resolution on the absolute quantification of kinetic parameters.
Results Average errors in the estimation of kinetic parameters, computed with respect to primaries, were 12%, 12% and 25% (NS-CS) and 4%, 5% and -6% (NSE-CSE) at 11.5% at 511 keV for K1, k2 and k3, respectively. Errors were 22%, 25% and -41% (NS-CS) and 5%, 8% and -8% (NSE-CSE) at 20% at 511 keV for K1, k2 and k3, respectively.
Conclusions Using the energy in addition to the position information in the scatter correction of dynamic list-mode PET data significantly improves the accuracy of the absolute quantification of kinetic parameters in cardiac imaging of large patients.
Research Support AHA-0655909T, R01-HL9507