Frequencies contributing to functional connectivity in the cerebral cortex in "resting-state" data

Area of Science:

• Neuroimaging
• Systems Neuroscience
• Physiology

Background:

• Resting-state functional connectivity (FC) is characterized by high cross-correlation coefficients between voxels in functionally connected brain regions.
• Understanding the sources of these correlations, particularly low frequencies and physiological noise, is crucial for accurate FC mapping.

Purpose of the Study:

• To quantify the contributions of low-frequency fluctuations and physiological noise to cross-correlation maps in resting-state functional connectivity.
• To differentiate the frequency components influencing functional connectivity in various brain regions.

Main Methods:

• Acquisition of task-activation and resting-state functional magnetic resonance imaging (fMRI) data in four healthy volunteers.
• Analysis of four contiguous slices at a high sampling rate, focusing on regions of interest (ROIs) of four contiguous voxels.
• Calculation of correlation coefficients and separation into frequency contributions (low-frequency, respiratory, and cardiac ranges).

Main Results:

• Resting-state FC maps closely resembled task-activation maps across all volunteers.
• Low-frequency fluctuations (<0.1 Hz) accounted for over 90% of the correlation coefficient in cortical regions.
• Physiological noise (respiratory and cardiac) contributed less than 10% to cortical FC maps, but was more prominent in blood vessels and cerebrospinal fluid.

Conclusions:

• Functional connectivity in auditory, visual, and sensorimotor cortices is predominantly driven by low frequencies (<0.1 Hz), which are slower than cardiac and respiratory cycles.
• These low-frequency fluctuations exhibit high temporal coherence within functionally connected brain regions.
• The findings suggest that low-frequency BOLD signal fluctuations are the primary drivers of resting-state functional connectivity in the human brain.