Cortical Hierarchy Collapse in Alzheimer's Disease: Connectome Gradient Compression as a Potential Biomarker

Abstract

This investigation centered on Alzheimer's disease (AD), a progressive neurodegenerative disorder characterized by memory impairment and cognitive decline. Functional connectome gradient analysis was utilized to investigate alterations in the hierarchical architecture of brain networks in AD. The study cohort consisted of 222 subjects, encompassing 111 AD patients and 111 normal controls (NC). Connectome gradients were computed via a dimensionality reduction technique based on diffusion map embedding and analyzed at both the region of interest (ROI) and network levels. Additional connectome gradient metrics, including network median distance and gradient eccentricity, were calculated, and the relationship between connectome gradients and rich-club organization was assessed. These connectome gradient values were subsequently correlated with clinical cognitive scores. The results demonstrated a significant reduction in the principal gradient range in AD patients. At the network level, gradient values exhibited an increase in the somatomotor (SMN) and visual networks (VIS), while decreasing in the default mode (DMN) and frontoparietal networks (FPN) relative to controls. Analyzes of network mean distance and gradient eccentricity further revealed compression of the brain cortical hierarchy in AD patients. Furthermore, rich-club analyzes indicated a reduction in the gradient value difference between hub and peripheral nodes in AD patients. Finally, clinical correlation analysis revealed a positive correlation between the degree of cognitive impairment and the degree of compression of the brain cortical hierarchy. These findings provide a novel perspective on the study of brain network organization in AD patients, contributing to a more comprehensive understanding of the neural mechanisms underlying Alzheimer's disease.