Merging PET and MEG Neuroimaging Graphs to Predict Alzheimer’s Conversion in Preclinical Patients

Abstract

Decades before the onset of Alzheimer’s disease, individuals may already be unknowingly accumulating harmful amyloid proteins associated with the disease. This study aims to identify the differences in how neurochemical systems shape brain connectivity between preclinical Alzheimer's disease patients accumulating these harmful proteins and healthy agers. By aligning PET neurotransmitter atlases onto MEG neurophysiological connectivity graphs, we will model the influence of neurochemistry on brain signaling for each individual. This analysis will involve linear mixed-effects modelling and non-parametric statistical methods to identify these alignments, as well as to predict which asymptomatic individuals will develop cognitive impairments attributed to Alzheimer’s disease in the future.