Analysis of Resting-State EEG Features in Alzheimer's Disease and Frontotemporal Dementia with Bayesian Hierarchical Models

Electroencepholography (EEG) biomarkers are crucial for studying neurodegenerative diseases such as Alzheimer's disease (AD) and frontotemporal dementia (FTD). Traditional analyses often require correction for multiple comparisons, reducing sensitivity. We applied Bayesian hierarchical models (BHM) to EEG spectral power and connectivity measures to improve inference. Using EEG data from 88 subjects (36 AD, 23 FTD, 29 controls), we compared BHM and linear models (LM). BHM identified more significant group differences than LM, revealing neurophysiological alterations in AD and FTD and highlighting the advantages of hierarchical modeling for EEG analysis.