Revealing a multiplex brain network through the analysis of recurrences

Area of Science:

• Network Neuroscience
• Computational Neuroscience
• Brain Dynamics Analysis

Background:

• Multilayer approaches are gaining traction in network neuroscience for modeling complex brain dynamics.
• Existing models often simplify the intricate interplay of brain activity across different frequency bands and time scales.
• Understanding functional connectivity's evolution during cognitive tasks is crucial for mapping brain function.

Purpose of the Study:

• To develop and demonstrate a frequency-based multilayer functional network model using recurrence analysis.
• To investigate the dynamic evolution of whole-head functional connectivity during a prolonged stimuli classification task.
• To analyze the relationship between network synchronization, frequency bands, and behavioral response times.

Main Methods:

• Construction of a frequency-based multilayer functional network from nonstationary multivariate electroencephalography (EEG) data.
• Application of recurrence analysis and a recurrence-based index of synchronization to define network edges.
• Analysis of both intralayer (within-frequency) and interlayer (cross-frequency) graph connections.

Main Results:

• Graph edge weights significantly increased throughout the stimuli classification experiment.
• Increased network synchronization negatively correlated with behavioral response times, indicating improved task performance.
• High-frequency brain activity showed synchronization of remote local areas, while low-frequency activity promoted large-scale coupling.

Conclusions:

• The developed multilayer network approach effectively models dynamic functional connectivity changes in the brain.
• Frequency-specific network dynamics play distinct roles in information processing and cognitive task performance.
• This framework provides novel insights into the spatiotemporal organization of brain networks during cognitive tasks.