A unified approach for characterizing static/dynamic connectivity frequency profiles using filter banks.

Area of Science:

• Neuroscience
• Functional Neuroimaging
• Network Science

Background:

• Static and dynamic functional network connectivity (FNC) are typically analyzed independently, limiting a comprehensive understanding of brain connectivity.
• Existing methods, such as sliding-window approaches, may miss crucial frequency-dependent connectivity information.

Purpose of the Study:

• To introduce filter-banked connectivity (FBC) as a unified method for analyzing both static and dynamic FNC across the full frequency spectrum.
• To investigate the utility of FBC in identifying distinct network states and characterizing frequency profiles in brain activity.

Main Methods:

• Developed and applied the FBC approach to estimate connectivity across multiple frequency bands.
• Utilized a resting-state fMRI dataset from schizophrenia patients (SZ) and typical controls (TC).
• Clustered FBC results into distinct network states and analyzed group differences in state occupancy based on frequency.

Main Results:

• FBC successfully estimated connectivity across a wider frequency range compared to sliding-window methods.
• Identified distinct network states, some characterized by low-frequency patterns not captured by traditional analyses.
• Schizophrenia patients exhibited a tendency to occupy higher-frequency network states more than typical controls.

Conclusions:

• FBC provides a novel, unified framework for analyzing static and dynamic functional network connectivity.
• This approach offers valuable insights into the frequency-specific characteristics of brain connectivity patterns.
• FBC can reveal group differences in dynamic connectivity, as demonstrated in the comparison between schizophrenia patients and typical controls.