Comparison of point spread function variations across the field of view of a PET/MR scanner with a standard resolution PET/CT

The point spread function (PSF) over the scanning field of view (FOV) of a PET scanner can be measured using printed sources. This provides a valuable extension to NEMA resolution measurements and can be used to enable resolution modelling image reconstruction. In this work the PSF of a multimodality whole body PET/MR scanner has been measured using fluorine-18 printed point sources and compared to measurements on a standard resolution PET/CT. A double 3D Gaussian function was fitted to each printed source and used to model the PSF response. Differences between the measured PSFs were quantified using full-width-at-half-maximum (FWHM) and full-width-at-tenth-maximum (FWTM) in the radial, tangential and axial directions. For the PET/MR in comparison to the PET/CT, greater deterioration of the PSFs were observed visually and quantitatively as the radial offset position increases from 10 - 230 mm (FWHM range values of 3.54 - 8.31 mm compared to 4.71 - 5.69 mm on the PET/CT). Additionally, the resolution of the PET/MR was better transaxially and worse axially than the PET/CT close to the scanner centres. These observations are likely to be consequence of: the effect that the magnetic field has on positron range (Lorentz force), shortening the transaxial positron range and increasing the axial range); differences in the implementations of the reconstruction algorithms; differences in scanner geometries where PET detectors on the PET/MR are closer to the patient (resulting in bigger parallax errors as the radial offset increases); and on the larger axial crystal size of the PET/MR. A preliminary assessment of image quality on the PET/MR scanner using an Esser phantom at different radial offset positions is also presented in this work.