Co-activation patterns across multiple tasks reveal robust anti-correlated functional networks

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Functional Brain Imaging

Background:

• The existence and significance of antagonistic brain states (anti-correlations) in functional brain organization remain debated.
• Some researchers consider anti-correlations an artifact of data preprocessing, while others view them as biologically meaningful predictors of brain activity.

Purpose of the Study:

• To investigate task-based co-activation patterns across the whole brain.
• To determine if brain regions exhibit anti-correlated activity during tasks, similar to resting-state observations.
• To explore the clinical relevance of anti-correlations in patients with depression.

Main Methods:

• Analysis of 47 task contrasts from the Human Connectome Project (N=680) to examine whole-brain co-activation patterns.
• Comparison of task-based anti-correlations with resting-state functional connectivity data, including analysis with and without global signal regression (GSR).
• Assessment of task-based anti-correlations between the dorsolateral prefrontal cortex (DLPFC) and subgenual anterior cingulate cortex in 25 patients with depression.

Main Results:

• Robust antagonistic interactions were found between brain networks during various tasks.
• The default mode network showed the highest degree of cortex-wide negative connectivity.
• Task-based anti-correlations corresponded well with resting-state data processed using GSR, which also proved a better predictor of task-induced modulation.
• Task-based anti-correlations between DLPFC and subgenual anterior cingulate cortex were associated with treatment efficacy in depression patients.

Conclusions:

• Anti-correlations represent a biologically meaningful phenomenon in brain function.
• These antagonistic interactions may constitute a fundamental principle of functional brain organization.
• GSR-processed resting-state data enhances the prediction of task-induced brain activity.