Partial-volume effect and a partial-volume correction for the NanoPET/CT™ preclinical PET/CT scanner

The partial-volume effect (PVE) can compromise the quantitative accuracy of preclinical positron emission tomogprahy (PET) images. We investigated this effect by simulating a NEMA NU4 mouse Image Quality Phantom (including calculation of a spatially variant position-dependent point spread function (PSF) ) as well as measuring it. Three different types of deconvolution-based partial- volume corrections were applied: the Lucy-Richardson (LR), Van Cittert (VC) and Wiener (WNR) algorithms. For simulation and phantom measurement, partial-volume correction was applied. Based on optimal parameters the same methods were used to correct data of a mouse study acquired with the NanoPET/CT™ preclinical scanner (Mediso Ltd. Budapest, Hungary and and Bioscan Inc., Washington DC, USA). The algorithms used successfully improve the values of recovery coefficients to varying extents both in simulation and measurement. The preliminary mouse study shows improvement in quantification. Full validation is subject to further investigation.