Frequency Clustering Analysis for Resting State Functional Magnetic Resonance Imaging Based on Hilbert-Huang...

Area of Science:

• Neuroscience
• Brain Imaging
• Signal Processing

Background:

• Resting-state functional magnetic resonance imaging (fMRI) is crucial for understanding brain functions.
• Frequency dependence in blood oxygen level dependent (BOLD) signals may reveal brain region interactions.
• Existing methods may not fully capture the dynamic, frequency-specific nature of brain activity.

Purpose of the Study:

• To introduce a novel frequency clustering analysis method for resting-state fMRI data.
• To leverage the Hilbert-Huang Transform (HHT) and a label-replacement procedure for brain network analysis.
• To provide a new measure for brain functional segregation based on time-frequency characteristics.

Main Methods:

• Extraction of time series from predefined regions of interest (ROIs).
• Decomposition of time series into intrinsic mode functions (IMFs) using HHT.
• Application of an improved k-means clustering with a label-replacement procedure to classify ROIs.

Main Results:

• Consistent and stable clustering results were observed across different runs and datasets.
• The method demonstrated efficacy in analyzing two independent resting-state fMRI datasets.
• Identified clusters reflect stable functional segregation patterns in the brain.

Conclusions:

• The developed framework offers a novel approach to analyzing brain functional networks.
• Time-frequency characteristics of resting-state BOLD signals provide valuable insights into brain organization.
• This method enhances the understanding of functional segregation in the brain.