Improved longitudinal [(18)F]-AV45 amyloid PET by white matter reference and VOI-based partial volume effect correction.

Amyloid positron-emission-tomography (PET) offers an important research and diagnostic tool for investigating Alzheimer's disease (AD). The majority of amyloid PET studies have used the cerebellum as a reference region, and clinical studies have not accounted for atrophy-based partial volume effects (PVE). Longitudinal studies using cerebellum as reference tissue have revealed only small mean increases and high inter-subject variability in amyloid binding. We aimed to test the effects of different reference regions and PVE-correction (PVEC) on the discriminatory power and longitudinal performance of amyloid PET. We analyzed [(18)F]-AV45 PET and T1-weighted MRI data of 962 subjects at baseline and two-year follow-up data of 258 subjects. Cortical composite volume-of-interest (VOI) values (COMP) for tracer uptake were generated using either full brain atlas VOIs, gray matter segmented VOIs or gray matter segmented VOIs after VOI-based PVEC. Standard-uptake-value ratios (SUVR) were calculated by scaling the COMP values to uptake in cerebellum (SUVRCBL), brainstem (SUVRBST) or white matter (SUVRWM). Mean SUV, SUVR, and changes after PVEC were compared at baseline between diagnostic groups of healthy controls (HC; N=316), mild cognitive impairment (MCI; N=483) and AD (N=163). Receiver operating characteristics (ROC) were calculated for the discriminations between HC, MCI and AD, and expressed as area under the curve (AUC). Finally, the longitudinal [(18)F]-AV45-PET data were used to analyze the impact of quantitation procedures on apparent changes in amyloid load over time. Reference region SUV was most constant between diagnosis groups for the white matter. PVEC led to decreases of COMP-SUV in HC (-18%) and MCI (-10%), but increases in AD (+7%). Highest AUCs were found when using PVEC with white matter scaling for the contrast between HC/AD (0.907) or with brainstem scaling for the contrast between HC/MCI (0.658). Longitudinal increases were greatest in all diagnosis groups with application of PVEC, and inter-subject variability was lowest for the white matter reference. Thus, discriminatory power of [(18)F]-AV45-PET was improved by use of a VOI-based PVEC and white matter or brainstem rather than cerebellum reference region. Detection of longitudinal amyloid increases was optimized with PVEC and white matter reference tissue.

Impact of MRI-based Segmentation Artifacts on Amyloid- and FDG-PET Quantitation.

INTRODUCTION: Magnet resonance image (MRI)-based segmentations are widely used for clinical brain research, especially in conjunction with positron-emission-tomography (PET). Although artifacts due to segmentation errors arise commonly, the impact of these artifacts on PET quantitation has not yet been investigated systematically. Therefore, the aim of this study was to assess the effect of segmentation errors on [(18)F]-AV45 and [(18)F]-FDG PET quantitation, with and without correction for partial volume effects (PVE). MATERIAL AND METHODS: 119 subjects with both [(18)F]-AV45, and [(18)F]-FDG PET as well as T1-weighted MRI at baseline and at two-year follow-up were selected from the ADNI cohort, and their MRI brain images were segmented using PMOD 3.5. MRIs with segmentation artifacts were masked with the corresponding [(18)F]-FDG PET standard-uptake-value (SUV) images to elucidate and quantify the impact of artifacts on PET analyses for six defined volumes-of-interest (VOI). Artifact volumes were calculated for each VOI, together with error-[%] and root-mean-square-errors (RMSE) in uncorrected and PVE corrected SUV results for the two PET tracers. We also assessed the bias in longitudinal PET data. RESULTS: Artifacts occurred most frequently in the parietal cortex VOI. For [(18)F]-AV45 and [(18)F]-FDG PET, the percentage-errors were dependent on artifact volumes. PVEC SUVs were consequently more distorted than were their uncorrected counterparts. In static and longitudinal assessment, a small subgroup of subjects with large artifacts (≥1500 voxels; ≙5.06 cm³) accounted for much of the PET quantitation bias. CONCLUSION: Large segmentation artifacts need to be detected and resolved as they considerably bias PET quantitation, especially when PVEC is applied to PET data.