Performance evaluation and optimization of the MiniPET-II scanner

This paper presents results of the performance of a small animal PET system (MiniPET-II) installed at our Institute. MiniPET-II is a full ring camera that includes 12 detector modules in a single ring comprised of 1.27×1.27×12mm3 LYSO scintillator crystals. The axial field of view and the inner ring diameter are 48mm and 211mm, respectively. The goal of this study was to determine the NEMA-NU4 performance parameters of the scanner. In addition, we also investigated how the calculated parameters depend on the coincidence time window (τ=2, 3 and 4ns) and the low threshold settings of the energy window (Elt=250, 350 and 450keV). Independent measurements supported optimization of the effective system radius and the coincidence time window of the system. We found that the optimal coincidence time window and low threshold energy window are 3ns and 350keV, respectively. The spatial resolution was close to 1.2mm in the center of the FOV with an increase of 17% at the radial edge. The maximum value of the absolute sensitivity was 1.37% for a point source. Count rate tests resulted in peak values for the noise equivalent count rate (NEC) curve and scatter fraction of 14.2kcps (at 36MBq) and 27.7%, respectively, using the rat phantom. Numerical values of the same parameters obtained for the mouse phantom were 55.1kcps (at 38.8MBq) and 12.3%, respectively. The recovery coefficients of the image quality phantom ranged from 0.1 to 0.87. Altering the τ and Elt resulted in substantial changes in the NEC peak and the sensitivity while the effect on the image quality was negligible. The spatial resolution proved to be, as expected, independent of the τ and Elt. The calculated optimal effective system radius (resulting in the best image quality) was 109mm. Although the NEC peak parameters do not compare favorably with those of other small animal scanners, it can be concluded that under normal counting situations the MiniPET-II imaging capability assures remarkably good image quality, sensitivity and spatial resolution.