Validation of a GATE Model for the Simulation of a Trionix TRIAD SPECT Camera

Abstract

The aim of this study was to validate a GATE simulation for the Trionix TRIAD triple head SPECT camera. Experimental and simulated physical characteristics, such as the energy spectrum and resolution, the system sensitivity, and the spatial resolution, were compared. The energy spectrum of the TRIAD was analyzed by using averaged screen-captures of the screen out MCA window and a linear interpolation of the graph. The energy resolution of 140 keV gamma rays was measured without the collimator. To measure the system sensitivity, we measured counts from a Tc-99m point source five times and averaged the results. The spatial resolution was measured from the profile of the line source aligned along the x-axis by using the bi-linear interpolation method. The tests were performed at various distances with a LEUHR parallel-hole collimator for further research on brain scans. The energy window was set to 126 similar to 154 keV for the simulation and to 140 keV +/- 20% for the real detector. The shapes of the normalized energy spectra were identical. The energy resolution was 10.1% in the experiment and 9.9% in the simulation, for a 10.1% resolution setting. The sensitivities at various distances (5 cm, 10 cm, 15 cm, and 20 cm) were almost identical. Simulation results (66 counts/sec/MBq) were higher than real experiment results (63 counts/sec/MBq). The spatial resolutions of the simulation data (4.7, 6.3, 7.9, and 9.7 mm) were also comparable to those of real experiments (4.4, 5.8, 7.5, and 9.1 mm). Overall the comparisons showed good agreements between the experimental and the simulation studies. The error levels were roughly 5%, what would be small enough to use the GATE simulation for further MonteCarlo-simulation-based investigations.