Monte Carlo simulation of a GE eXplore VISTA system for quantitative small animal PET imaging

We aim to establish an evaluation platform for the GE eXplore VISTA small animal positron emission tomography (PET) scanner, a dual layer phoswich system, by using Monte Carlo simulation.

We developed a detection model based on the Geant4 Application for Tomographic Emission to realistically reproduce the physics of PET, the scanner configuration, and the data collecting system of an eXplore VISTA system. For verification purpose, several different physical phantoms were simulated to perform evaluation tests, including sensitivity, spatial resolution, scatter fraction, and count rate performance, which were compared with an actual scanner. After the experimental validation, our detection model was applied to assess the quantification loss in the reconstructed images associated with photon attenuation, photon scatter, and random coincidences.

A simulated sensitivity profile as a function of 18F point source axial position was fitted to the measured results. In terms of spatial resolution, agreement was within 10-18% for the point source at various locations. The simulated and measured scatter fractions differed by less than 4.3 and 5.2% for the physical mouse and rat phantoms, respectively. The count rate performance of our model was matched by the measured results, up to the peak activity concentration of 455 kBq/ml for the mouse-sized phantom and 141 kBq/ml for the rat-sized phantom. Finally, we found that photon attenuation is the dominant physical degrading factor in quantitative analysis (> 13.4%).

These results suggested that the proposed detection model is able to produce realistic data from the eXplore VISTA system with knowing the ground truth, thus facilitating its evaluation for small animal PET studies.