Eigenvector alignment: Assessing functional network changes in amnestic mild cognitive impairment and Alzheimer's...

Area of Science:

• Neuroscience
• Network Science
• Medical Imaging Analysis

Background:

• Established methods for brain network analysis, like eigenvector centrality, offer limited insights into regional relationships.
• Understanding functional connectivity requires assessing pairwise connections, often missing holistic network dynamics.

Purpose of the Study:

• To introduce and validate eigenvector alignment as a novel method for analyzing human brain functional networks.
• To identify and characterize differences in brain networks among healthy controls (HC), amnestic Mild Cognitive Impairment (aMCI), and Alzheimer's disease (AD) subjects.

Main Methods:

• Adapted eigenvector alignment from network science to analyze human brain functional networks.
• Compared the placement of brain regions in a Euclidean space defined by dominant eigenvectors.
• Assessed functional connectivity changes and their impact on regional relationships.

Main Results:

• Eigenvector alignment identified degradation of bilateral cortical connectivity in AD subjects.
• Significant alignment changes were observed in the auditory cortex starting in aMCI, becoming prominent in AD.
• Notable alignment differences were detected in hippocampal regions between aMCI and AD subjects.

Conclusions:

• Eigenvector alignment provides a holistic approach to brain network analysis, complementing existing methods.
• The method reveals subtle yet significant structural changes in brain networks affected by neurodegenerative diseases.
• Eigenvector alignment can capture the adaptive changes in functional brain networks during disease progression.