Inter-scanner Aβ-amyloid PET harmonization using barrel phantom spatial resolution matching

Abstract

AbstractThe standardized uptake value ratio (SUVR) is used to measure Aβ uptake in PET images of the brain. Variations in PET scanner technologies and image reconstruction techniques can lead to variability in images acquired from different scanners. This poses a challenge for Aβ-PET studies conducted across multiple centers. The aim of harmonization is to achieve consistent Aβ-PET measurements across scanners. In this study, the procedure of matching the spatial resolution of a barrel phantom measured in each PET scanner is proposed as a method of Aβ-PET harmonization, validated using subject data.MethodsThree different PET scanners were used: the Siemens Biograph Vision 600, Siemens Biograph mCT, and Philips Gemini TF64. A total of five, eight, and five subjects were each scanned twice with [18F]-NAV4694 across Vision-mCT, mCT-Philips, and Vision-Philips scanner pairs. The Vision and mCT scans were reconstructed using various iterations, subsets, and post-reconstruction Gaussian smoothing, while one reconstruction configuration was used for the Philips scans. The full-width at half-maximum (FWHM) of each reconstruction configuration was calculated using [18F]-filled barrel phantom scans with the SNMMI phantom analysis toolkit. Regional SUVRs were calculated from 72 brain regions using the AAL3 atlas for each subject and reconstruction configuration. Statistical similarity between SUVRs was assessed using paired (within subject) t-tests for each pair of reconstructions across scanners; the higher the p-value, the greater the similarity between the SUVRs.ResultsVision-mCT harmonizationmaximal statistical similarity (i.e.,p-value) between regional SUVRs was achieved using a 4.10 mm FWHM Vision reconstruction with a 4.30 mm FWHM mCT reconstruction.Philips-mCT harmonizationmaximal statistical similarity between regional SUVRs was achieved using an 8.2 mm FWHM Philips reconstruction with a 9.35 mm FWHM mCT reconstruction.Philips-Vision harmonizationa 9.1 mm FWHM Vision reconstruction had maximum statistical similarity with regional SUVRs from an 8.2mm FWHM Philips reconstruction. Reconstruction pairs that maximized statistical similarity, and supported a null hypothesis of being drawn from the same distribution, were selected as harmonised for each scanner pair.ConclusionUsing data obtained from three sets of participants, each scanned on a different pair of PET scanners, using reconstruction configurations with matched barrel phantom spatial resolutions, we have demonstrated that Aβ-PET quantitation can be harmonised across scanners, producing SUVR values statistically likely to be drawn from the same distribution. This finding is encouraging for the use of different PET scanners in multi-centre trials, or updates during longitudinal studies.