Association between partial volume corrected longitudinal tau measures and cognitive decline

Objectives: PET imaging of tau aggregates in the brain plays a pivotal role in image-based biomarker development for aging and Alzheimer’s disease. However, partial volume effects arising from the limited spatial resolution of PET pose a challenge to quantitation. We have applied an image deblurring technique that utilizes high-resolution MR-based anatomical information to correct for partial volume effects in PET to longitudinal [18F]Flortaucipir datasets. Our objective is to assess the impact of partial volume correction on longitudinal tau quantitation as revealed by associations with clinical indices. Methods: Serial [18F]Flortaucipir (also known as [18F]T807 and [18F]AV1451) PET data were acquired from an elderly cohort (63-90 years) consisting of 109 Harvard Aging Brain Study participants. The time gap between baseline and follow-up scans was 6-63 months. The images were deblurred based on 1) deconvolution using the actual spatially-variant point spread function (PSF) of the PET scanner measured in the image space and 2) an MR-based information theoretic framework that minimizes the joint entropy between the PET and a high-resolution T1-MPRAGE MR. The following cognitive and clinical measures were available for these subjects: Mini-Mental State Examination (MMSE), Boston Naming Test (BNT), Logical Memory Test (LM), and Clinical Dementia Rating (CDR). Additionally, amyloid measures based on [11C]PiB retention in the frontal, lateral temporal and retrosplenial cortices (PiB-FLR) was computed for each subject. FreeSurfer-defined regions-of-interest (ROI) means were computed from the original and deblurred standardized update value ratios (SUVRs) with white matter reference.

Results: Pearson’s correlation coefficients were computed between the clinical measures and the original and deblurred mean PET SUVRs at two time points (PET-t1, PET-t2, dbPET-t1, and dbPET-t2) for four ROIs: inferior temporal cortex (ITC), fusiform gyrus (FG), parahippocampal gyrus (PHG), and entorhinal cortex (EC). Qualitatively, PET image deblurring successfully recovered edges and structural details in the PET images. Quantitatively, deblurring led to stronger associations of mean tau SUVRs with almost all clinical measures and all ROIs. Deblurring led to stronger negative correlations for MMSE, BNT, and LM scores and stronger positive correlations with CDR and PiB-FLR.

Conclusions: High-resolution MR-based PET image deblurring led to contrast improvement at both baseline and follow up in several clinically meaningful ROIs. This translated to improved correlation of tau measures with cognitive testing results, thereby demonstrating the clinical utility of image deblurring in a sizable human cohort.