Abstract
BACKGROUND: Developing biomarkers for early stage AD patients is crucial. Glucose metabolism measured by (18)F-FDG PET is the most common biomarker for evaluating cellular energy metabolism to diagnose AD. Arterial spin labeling (ASL) MRI can potentially provide comparable diagnostic information to (18)F-FDG PET in patients with neurodegenerative disorders. However, the conclusions about the diagnostic performance of AD are still controversial between (18)F-FDG PET and ASL. This study aims to compare quantitative cerebral blood flow (CBF) and glucose metabolism measured by (18)F-FDG PET diagnostic values in patients with Alzheimer's disease (AD) and amnestic mild cognitive impairment (aMCI) using integrated PET/MR. RESULTS: Analyses revealed overlapping between decreased regional rCBF and (18)F-FDG PET SUVR in patients with AD compared with NC participants in the bilateral parietotemporal regions, frontal cortex, and cingulate cortex. Compared with NC participants, patients with aMCI exclusively demonstrated lower (18)F-FDG PET SUVR in the bilateral temporal cortex, insula cortex, and inferior frontal cortex. Comparison of the rCBF in patients with aMCI and NC participants revealed no significant difference (P > 0.05). The ROC analysis of rCBF in the meta-ROI could diagnose patients with AD (AUC, 0.87) but not aMCI (AUC, 0.61). The specificity of diagnosing aMCI has been improved to 75.56% when combining rCBF and (18)F-FDG PET SUVR. CONCLUSION: ASL could detect similar aberrant patterns of abnormalities compared to (18)F-FDG PET in patients with AD compared with NC participants but not in aMCI. The diagnostic efficiency of (18)F-FDG-PET for AD and aMCI patients remained higher to ASL. Our findings support that applying (18)F-FDG PET may be preferable for diagnosing AD and aMCI.