A Triple-Network Dynamic Connection Study in Alzheimer's Disease.

Area of Science:

• Neuroscience
• Cognitive Science
• Medical Imaging

Background:

• Alzheimer's disease (AD) is linked to altered brain network organization and function.
• Understanding the dynamic interactions within large-scale brain networks is crucial for diagnosing and treating AD.

Purpose of the Study:

• To investigate dynamic functional network connectivity (dFNC) within the triple-network (saliency network, central executive network, default mode network) in individuals with AD, MCI, and CN.
• To determine if dynamic network interactions can serve as biomarkers for AD and MCI.

Main Methods:

• Group independent component analysis (Group ICA) was used to construct the triple-network.
• Dynamic time-varying triple-network interactions were analyzed using Group ICA with k-means clustering (GDA-k-means).
• Support vector machine (SVM) classification was performed using mean network interaction indices, age, and gender.

Main Results:

• Significant differences in mean brain state-specific network interaction indices (meanNII) were found among AD, MCI, and CN groups via ANOVA.
• SVM classification achieved high accuracy: 95% for AD vs. CN, 94% for AD vs. MCI, and 77% for MCI vs. CN.
• Dynamic functional interactions within the triple-network effectively differentiate cognitive states.

Conclusions:

• The dynamic characteristics of functional interactions within the triple-network are vital for understanding AD pathophysiology.
• Dysregulation of brain dynamics is a key feature of Alzheimer's disease.
• dFNC analysis offers a promising approach for early detection and monitoring of AD.