Biomarkers.

Area of Science:

• Neuroimaging
• Biomarkers
• Alzheimer's Disease Research

Background:

• Amyloid plaque deposition in Alzheimer's Disease (AD) is typically assessed using PET imaging and the Centiloid scale for global amyloid burden.
• Tracer uptake in specific brain subregions may offer additional diagnostic information beyond global measures.

Purpose of the Study:

• To identify brain regions where tracer uptake can serve as effective markers for AD diagnosis.
• To find key variables associated with AD status using advanced statistical methods.

Main Methods:

• Utilized Pittsburgh Compound B (PIB) PET imaging data from the OASIS-3 dataset (44 controls, 35 AD patients).
• Calculated Standardized Uptake Value Ratio (SUVR) in brain subregions after co-registration with MRI scans and partial volume correction.
• Employed Bayesian methods and Markov blanket analysis to identify a minimal, highly predictive subset of features linked to disease status.

Main Results:

• Identified 23 brain regions with SUVRs as a minimal marker subset highly predictive of AD diagnosis.
• Six of these regional SUVRs, in addition to the Centiloid measure, were directly linked to AD status.
• Key predictive regions included the Precuneus, Inferior temporal, lateral occipital, posterior cingulate, supramarginal, and temporal regions.

Conclusions:

• Regional SUVRs from PIB-PET scans can identify key brain regions associated with AD diagnosis.
• Tracer uptake in 13 specific regions, combined with global amyloid deposition, strongly correlates with AD status.
• These regional markers could aid in assessing AD likelihood and warrant further investigation in larger cohorts.