Dynamic Changes in Brain Networks of Alzheimer's Disease Based on Co-Activation Patterns

The dynamic changes in brain networks during the progression from normal controls (NC) to mild cognitive impairment (MCI) and Alzheimer's Disease (AD) remain unclear. This study investigates the dynamic alterations in brain networks as the disease progresses from NC to MCI and ultimately to AD. Using the Alzheimer's Disease Neuroimaging Initiative (ADNI) data, we performed the co-activation pattern (CAP) analysis on resting-state functional magnetic resonance imaging (rs-fMRI), focusing on the attentional network (AN), the default mode network (DMN), the frontoparietal network (FPN), the limbic network (LN), the motor-sensory network (MN), the salience network (SN), and the visual network (VN). The clustering analysis identified eight distinct CAPs, which are classified into three categories: primary sensory network (PSN) dominant, high-order cognitive network (HOCN) dominant, and PSN-HOCN co-dominant. We further analyzed CAP transition matrices and CAP occurrence rates across different disease groups. Results revealed that during the transition from NC to MCI, CAP transitions were primarily unidirectional and confined to the same CAP category, with increased dynamic activity of HOCN-dominant CAPs. In contrast, the transition from MCI to AD involved bidirectional CAP transitions with frequent inter-network transitions, while the AD stage exhibited complex interactions across multiple networks. These findings provide novel insights into the dynamic brain network changes during AD progression, potentially paving the way for future non-invasive brain stimulation therapies based on brain functional dynamics.