Structural, geometric and genetic factors predict interregional brain connectivity patterns probed by...

Area of Science:

• Neuroscience
• Network Science
• Computational Biology

Background:

• Electrocorticography (ECoG) offers insights into brain connectivity but faces challenges due to invasiveness and individual variability.
• Analyzing whole-brain ECoG networks is crucial for understanding brain function and dysfunction.

Purpose of the Study:

• To investigate the architecture of whole-brain ECoG networks and the factors influencing their organization.
• To develop predictive models for brain connectivity patterns using network science and linear modeling.
• To explore the potential clinical utility of these models, such as predicting surgical outcomes.

Main Methods:

• Analysis of whole-brain, interregional, and band-limited ECoG networks from a large cohort of individuals with medication-resistant epilepsy.
• Application of network science tools to characterize network organization, including modularity and distance-dependent connectivity.
• Development of linear models incorporating white matter pathways, Euclidean distance, and gene expression to explain connection strengths.

Main Results:

• Characterization of frequency-specific ECoG network architecture and modular organization.
• Identification of white matter pathways, interregional distance, and gene expression as key factors shaping brain-wide connectivity.
• Successful prediction of out-of-sample, single-subject ECoG network data.

Conclusions:

• Whole-brain ECoG network analysis is feasible and reveals fundamental organizational principles.
• A combination of anatomical, spatial, and genetic factors jointly shapes brain network architecture.
• Predictive models derived from ECoG data hold promise for clinical applications, including surgical planning.