Evaluating EEG complexity and spectral signatures in Alzheimer’s disease and frontotemporal dementia: evidence for rostrocaudal asymmetry
Kassra Ghassemkhani, Kevin S. Saroka, Blake T. Dotta
Abstract
Accurate classification of neurodegenerative disorders remains a challenge in neuroscience. Using open-source electroencephalography (EEG) data, we investigated electrophysiological signatures to differentiate frontotemporal dementia (FTD) from Alzheimer's disease (AD) via complexity measures. Traditional relative band power analysis showed consistent increases in lower-frequency activity but did not distinguish the two disorders after correction. In contrast, fractal dimension and long-range temporal correlations (LRTCs) revealed distinct topographical differences: AD exhibited rostral dominance in fractal dimension, whereas FTD showed caudal dominance. Both disorders demonstrated reduced LRTCs, particularly in caudal regions, indicating disrupted large-scale neural dynamics. These findings suggest that complexity-based EEG features may offer a reliable, cost-effective tool for distinguishing neurodegenerative conditions, complementing traditional neuroimaging approaches.