The effect of partial volume effect correction on PIB and FDG PET data in MCI and AD patients

Objectives Partial volume effects (PVE) in brain PET-imaging hamper the estimation of the true cortical tracer-uptake. The purpose of this study is to investigate effects of MRI-based PVE-correction (PVEC) on PIB and FDG-PET studies in patients with Alzheimer’s disease (AD) and mild cognitive impairment (MCI).

Methods In 8 healthy subjects, 16 MCI patients and 21 AD patients, PIB-PET, FDG-PET and cranial MRI was performed. MRI-based PVE-correction of PET-data was carried out using PMOD-software. Voxel-based comparison (SPM5) was performed between images with and without PVEC (paired t-test) and between patients and healthy subjects (two-sample t-test). Results surviving correction for FDR at a significance threshold of p<0.05 were considered as significant.

Results PVE-correction resulted in a significantly higher PIB-uptake in the temporal lobes, parietal lobes and some areas of the frontal lobe and lead to higher FDG-uptake in the medial temporal lobes and cerebellum. Lower uptake in PIB and FDG-PET was found in the peripheral areas of the cortex. Despite these significant findings, comparison of the patients FDG and PIB-PET data with a healthy control population did not reveal major differences between PVE-corrected and non-corrected data sets.

Conclusions Although the pattern of regional cortical tracer-uptake changed considerably after PVE-correction, PVE-correction did not induce any basic changes in the results of statistical comparison between PET-data of patients and healthy subjects. Consequently, the assessment of PIB-PET data without PVE-correction may lead to regional misinterpretation of tracer uptake but probably will not lead to a clinical misclassification of the examined subject.