A Subtyping Study of Patients with Alzheimer’s Disease Based on Brain Functional Dynamics

Abstract

Alzheimer’s disease (AD) is a neurodegenerative disorder primarily affecting the an older population and the most common cause of dementia. The disease exhibits substantial heterogeneity in genetic basis, pathological manifestations, and clinical phenotypes, collectively posing major challenges for precise diagnosis, effective treatment, and drug development. To disentangle the heterogeneity of AD from both brain structural and functional dimensions, this study included 215 subjects from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) cohort and applied the Subtype and Stage Inference (SuStaIn) algorithm for disease subtyping and staging based on gray matter volume derived from structural magnetic resonance imaging and dynamic features of co‐activation patterns (CAPs) from resting‐state functional magnetic resonance imaging. The results identified three structural subtypes and four functional subtypes. Each subtype exhibited a specific progression trajectory of brain atrophy or functional network abnormalities. There existed anatomical‐functional correspondences between structural and functional subtypes, but they did not constitute a simple one‐to‐one mapping. By systematically revealing the heterogeneity of AD from a structure‐function coupling perspective, this study provides new imaging biomarkers and a theoretical basis for precise subtyping, early detection, and individualized intervention of AD.