Anticorrelated networks in resting-state fMRI-BOLD data

Area of Science:

• Neuroimaging
• Brain Activity Analysis

Background:

• Resting-state functional magnetic resonance imaging (fMRI) measures spontaneous brain activity.
• Negative BOLD (Blood-Oxygen-Level-Dependent) signals are crucial for understanding brain function but are complex to characterize.

Purpose of the Study:

• To detect and characterize anticorrelated networks in resting-state fMRI data.
• To investigate the consistency of these networks across subjects.

Main Methods:

• Applied spatial independent component analysis (sICA) to fMRI data from 20 subjects at individual and group levels.
• Defined positive signals (PS) and negative signals (NS) using Z-score mapping (>5 and <-5).
• Calculated correlation coefficients between PS and NS time series to identify anticorrelated networks (correlation coefficient <-0.3).

Main Results:

• Detected abundant anticorrelated networks in 36.5% of independent components at the individual level.
• Observed significant subject-to-subject variability in the spatiotemporal patterns of most anticorrelated networks.
• Identified three specific networks with comparably consistent spatial patterns across subjects.
• Found no anticorrelated networks at the group level of analysis.

Conclusions:

• Anticorrelated networks are prevalent at the individual level in resting-state fMRI.
• The findings highlight the importance of considering both positive and negative BOLD signals for a comprehensive understanding of brain function.
• Future research should adopt broader approaches to characterize negative BOLD signals and their role in neural processing.