Evaluation of direct 4D parametric reconstruction with low count human PET data

Previously, we presented a direct 4D EM method for producing kinetic parameter images from list mode PET data, where the time-activity curve for each voxel is described by a one tissue compartment model (1T). The initial evaluation was performed with simulation, without random and scatter included. In addition, a practical single-scatter simulation (SSS) method was proposed so that the direct 4D method can be used for real data. In order to investigate the performance of the direct 4D method for real data of low count level, three datasets were obtained by gating original list mode data event-by-event (gating by 3, 6 and 10). Initial evaluation was performed using [¹¹C]AFM human data. Comparisons with the 2-step approach (frame-based reconstruction followed by voxel-by-voxel parameter estimation) provided encouraging results. Regional analysis showed that K₁ and V_T values for 4D are less variable over different gating reconstructions and the statistical noise for 4D was improved. Consistent with previous results of low count simulated data, the direct 4D has its greatest statistical advantage for low count human brain data.