Computational modelling reveals neurobiological contributions to static and dynamic functional connectivity patterns

Area of Science:

• Neuroscience
• Computational Biology
• Brain Imaging

Background:

• Functional connectivity (FC) is crucial for understanding brain function in health and disease.
• The neurobiological basis of static FC (sFC) and dynamic FC (dFC) requires further elucidation.
• Computational modeling offers a non-invasive approach to link brain activity patterns with neurobiology.

Purpose of the Study:

• To investigate the relationship between neurobiological parameters and individual differences in sFC and dFC.
• To model brain activity using empirical resting-state fMRI and DTI data from healthy individuals.
• To identify specific brain regions whose connectivity influences overall network configurations.

Main Methods:

• Developed computational models of brain activity for 200 healthy individuals using fMRI and DTI data.
• Optimized model parameters (global coupling, inhibition, NMDA synaptic coupling, recurrence weight) to replicate empirical sFC and temporal correlation (TC).
• Analyzed associations between brain-wide connectivity features and model parameters using correlation and prediction models; employed perturbation analysis to assess regional coupling effects.

Main Results:

• Models successfully replicated empirical sFC and TC, but not FC variance or node cohesion.
• Global coupling (G) positively correlated with FC features, while NMDA synaptic coupling (J_N) negatively correlated.
• Perturbation analysis identified the left paracentral gyrus and left pars triangularis as key regions impacting sFC and TC fits, respectively.

Conclusions:

• Neurobiological characteristics are significantly associated with individual variability in sFC and dFC.
• sFC and dFC are shaped by distinct, small sets of brain regions.
• This modeling approach provides novel evidence for the role of neurophysiology in establishing brain network configurations.