Detection of functional connectivity using temporal correlations in MR images

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Functional Neuroimaging

Background:

• Functional connectivity analysis in neuroscience typically uses task-based fMRI.
• Resting-state analysis is less prone to task-related confounds but requires robust methodologies.
• Understanding brain network dynamics is crucial for cognitive function.

Purpose of the Study:

• To develop and validate a novel method for assessing functional brain connectivity using independent datasets.
• To investigate functional connections within the human language system at rest and during auditory processing.
• To confirm and quantify the strength of known and discover novel interregional correlations.

Main Methods:

• Utilized a cross-validation approach with two independent resting-state fMRI datasets.
• Defined regions of interest (ROIs) and generated connectivity hypotheses in one dataset.
• Evaluated hypotheses by analyzing low-frequency temporal correlations in the independent dataset, reversing roles iteratively.
• Applied the method to the language system, including Broca's area and Wernicke's area.

Main Results:

• Confirmed functional connectivity between Broca's area and Wernicke's area in healthy subjects at rest.
• Demonstrated an increase in this connectivity during active language processing (listening to narrative text).
• Identified a significant resting-state correlation between Broca's area and left premotor cortex, which also increased during listening.

Conclusions:

• The proposed methodology reliably identifies and quantifies functional connections in high-level cognitive systems.
• The cross-validation approach enhances the robustness of functional connectivity findings.
• The study provides evidence for dynamic changes in brain network connectivity related to cognitive load.