Changes of brain functional network in Alzheimer's disease and frontotemporal dementia: a graph-theoretic analysis
Shijing Wu,Ping Zhan,Guojing Wang,Xiaohua Yu,Hongyun Liu,Weidong Wang
DOI: https://doi.org/10.1186/s12868-024-00877-w
2024-07-04
Abstract:Background: Alzheimer's disease (AD) and frontotemporal dementia (FTD) are the two most common neurodegenerative dementias, presenting with similar clinical features that challenge accurate diagnosis. Despite extensive research, the underlying pathophysiological mechanisms remain unclear, and effective treatments are limited. This study aims to investigate the alterations in brain network connectivity associated with AD and FTD to enhance our understanding of their pathophysiology and establish a scientific foundation for their diagnosis and treatment. Methods: We analyzed preprocessed electroencephalogram (EEG) data from the OpenNeuro public dataset, comprising 36 patients with AD, 23 patients with FTD, and 29 healthy controls (HC). Participants were in a resting state with eyes closed. We estimated the average functional connectivity using the Phase Lag Index (PLI) for lower frequencies (delta and theta) and the Amplitude Envelope Correlation with leakage correction (AEC-c) for higher frequencies (alpha, beta, and gamma). Graph theory was applied to calculate topological parameters, including mean node degree, clustering coefficient, characteristic path length, global and local efficiency. A permutation test was then utilized to assess changes in brain network connectivity in AD and FTD based on these parameters. Results: Both AD and FTD patients showed increased mean PLI values in the theta frequency band, along with increases in average node degree, clustering coefficient, global efficiency, and local efficiency. Conversely, mean AEC-c values in the alpha frequency band were notably diminished, which was accompanied by decreases average node degree, clustering coefficient, global efficiency, and local efficiency. Furthermore, AD patients in the occipital region showed an increase in theta band node degree and decreased alpha band clustering coefficient and local efficiency, a pattern not observed in FTD. Conclusions: Our findings reveal distinct abnormalities in the functional network topology and connectivity in AD and FTD, which may contribute to a better understanding of the pathophysiological mechanisms of these diseases. Specifically, patients with AD demonstrated a more widespread change in functional connectivity, while those with FTD retained connectivity in the occipital lobe. These observations could provide valuable insights for developing electrophysiological markers to differentiate between the two diseases.