PT - JOURNAL ARTICLE AU - Arnd Sörensen AU - Peter Pieperhoff AU - Ganna Blazhenets AU - Lars Frings AU - Katrin Amunts AU - Philipp Meyer TI - Deformation-based morphometry applied to FDG PET shows atrophy patterns suggestive of Alzheimer’s disease in patients with cognitive impairment and amyloid pathology DP - 2019 May 01 TA - Journal of Nuclear Medicine PG - 1456--1456 VI - 60 IP - supplement 1 4099 - http://jnm.snmjournals.org/content/60/supplement_1/1456.short 4100 - http://jnm.snmjournals.org/content/60/supplement_1/1456.full SO - J Nucl Med2019 May 01; 60 AB - 1456Aim: Deformation-based morphometry (DBM) is a powerful tool to reveal anatomical differences (e.g., on MRI) in patients compared to healthy controls, e.g., to depict atrophy patterns as a biomarker of neurodegeneration in Alzheimer’s disease (AD). The advent of fully-digital detector technology for PET/CT enables high-resolution FDG PET brain imaging in clinical routine. In the present study, the feasibility of applying DBM to fully-digital FDG PET data of patients with cognitive impairment was investigated. Methods: Twenty-five patients with cognitive impairment underwent FDG and amyloid PET during clinical routine. The FDG PET scans of Aβ+ (n=13) and Aβ- subjects (n=12; according to clinical reads of amyloid PET) were spatially transformed onto an FDG PET template which was created by non-linear registration of PET scans from nine healthy controls, similar to the procedures in Fonov et al. (2011). Each patient’s FDG PET scan (source image) was non-linearly registered to this template, yielding the transformed FDG PET scan and a 3D-deformation field that encodes structural differences between source and template brains. Maps of voxel-wise volume differences were computed to estimate the degree of local atrophy by voxel-wise ANCOVA comparing the Aβ+ and Aβ- groups, adjusted for age, sex and intracranial volume (assessed by low-dose CT) using SPM 12. In parallel, differences in normalized regional glucose metabolism were calculated (i.e., conventional FDG PET analysis using SPM 12 with proportional scaling to the whole brain FDG uptake; ANCOVA adjusted for age and sex). Results: The group comparison revealed reduced volumes in Aβ+ compared to Aβ- patients in the hippocampi and para-hippocampal gyri bilaterally as well as the right temporal pole (p < 0.005, k ≥ 30 voxels). Additional regions of volume loss included the lateral temporal and parietal lobes including the precuneus, which overlapped with extensive areas of regional hypometabolism on conventional analysis of regional hypometabolism. The areas of volume loss covered a total volume of 22.7 mL and the regions of hypometabolism a total volume of 31.5 mL, whereas the overlap between the two was only 0.8 mL. Conclusions: The present proof-of-principle study suggests that DBM is capable to reveal morphological changes by analysis of cerebral FDG PET scans in patients with cognitive impairment and Alzheimer’s pathologic change. This may constitute a biomarker of neurodegeneration that is complementary to conventional FDG PET analysis (regional hypometabolism). Future studies need to investigate its additional diagnostic value. References V. Fonov et al. (2011) NeuroImage 54(1):313-327 "Unbiased average age-appropriate atlases for pediatric studies"