Performance Evaluation of a PET Insert for Preclinical MRI in Standalone PET and Simultaneous PET-MRI Modes
Abstract Background: This study aimed to evaluate the performance of a preclinical PET insert in three configurations: as a standalone unit outside the MRI bore, inside the bore of a cryogen-free 3T MRI and, finally, while performing simultaneous PET/MRI studies.Methods: The PET insert consists of 2 rings of 6 detectors, each detector comprising 8x12 SiPMs reading out dual offset layers of pixelated LYSO crystals with a 1.4mm pitch. The inner diameter is 60mm, transaxial Field of View (FoV) 40mm and axial FoV 98mm. Evaluation was based on NEMA NU 4-2008 guidelines with appropriate modifications. Spatial resolution and sensitivity were measured inside and outside the MR bore. Image quality, count rate and quantitative performance were measured in all three configurations. The effect of temperature stability on PET sensitivity during fast spin echo sequences was also evaluated. B0 field homogeneity, T1 and T2 relaxation times were measured using a water-filled phantom, with and without simultaneous PET operation. Finally, PET and MRI scans of a mouse injected with 10 MBq [18F]NaF and a mouse injected with 16 MBq [18F]FDG were performed in sequential and simultaneous modes. Results: Peak absolute sensitivity was 10.15% with an energy window of 250-750 keV. Absolute sensitivity values outside and inside the MR bore with MR idle agreed to within 0.1%. Outside the MR bore spatial resolution was 1.21/1.59 mm FWHM (radial/tangential) 5mm from the centre of the FoV which compared well with 1.19/1.26mm FWHM inside the MR bore. There were no substantial differences between all three scan configurations in terms of peak NEC rate (175 kcps at 17MBq), scatter or random fractions. Uniformity and recovery coefficients were also consistent between scanning modes. B0 field homogeneity and T1 and T2 relaxation times were unaltered by the presence of the PET insert. No significant differences were observed between sequential and simultaneous scans of the animals.Conclusion: We conclude that performance of the PET insert and MRI system are not significantly affected by scanning mode.