Purpose To determine the accuracy of quantitative 124I PET imaging in the presence of therapeutic levels of 131I.
Material and Methods Multiple PET images were acquired using a NEMA IEC phantom with spheres containing 0.4 MBq/cc of 124I and increasing amount of 131I activity in the phantom background (0 to 3.76 GBq). Acquisitions were performed on a GE Discovery 710 PET/CT scanner. At each 131I activity level two scans were acquired, one with the phantom at the center of the field of view (FOV) and one 11 cm off-center. Images were reconstructed with an ordered subset expectation maximization (OSEM) algorithm using between 1 and 25 iterations of 16 subsets. Results were evaluated visually and by comparing the 124I activity relative to the baseline PET performed in the absence of 131I.
Results The presence of 131I within the PET FOV added to the random coincidence rate, to dead-time and to pile-up within the PET detectors. Using our standard clinical reconstruction parameters, the image quality and quantitative accuracy suffered at 131I background activities above 1.4 GBq. However, increasing the number of iterations resulted in dramatic improvements in image quality and quantitative accuracy. Projection space measurements suggest that the dead time corrections implemented on the scanner perform well even at the highest singles count rate tested (52 Mcps).
Conclusion This study shows that 124I quantitative PET is feasible in the presence of large amounts of 131I on a GE D710. The high random coincidence fraction slows the reconstruction convergence rate, therefore iterations equivalent to at least 8x16 are recommended.