Quantitative evaluation of simulated functional brain networks in graph theoretical analysis

Area of Science:

• Neuroscience
• Computational Biology
• Network Science

Background:

• Increasing interest in whole-brain computational models for mechanistic insights into resting-state brain networks.
• Importance of validating computational models against empirical data for topological accuracy.

Purpose of the Study:

• To assess the fidelity of computational models in reproducing topological features of empirical functional brain networks.
• To evaluate the quantitative agreement between graph theory metrics from simulated and empirical network data.

Main Methods:

• Utilized diffusion spectrum and resting-state functional magnetic resonance imaging data from healthy individuals.
• Constructed empirical and simulated functional networks (66 nodes) constrained by structural connectivity.
• Employed the Kuramoto model for simulating functional data with anatomical regions as phase oscillators.
• Analyzed network topology using graph theory and estimated relative error between empirical and simulated measures.

Main Results:

• Simulated data can reliably model global network organization across dynamic states.
• Solutions from simulated data show sensitive dependence on specified connection densities.
• Quantitative evaluation method developed for validating graph theory metrics from simulated data.

Conclusions:

• Whole-brain computational models, when constrained by structural connectivity, can confidently model topological features of resting-state brain networks.
• The study provides a framework for the quantitative evaluation and external validation of simulated network data.
• Findings inform future study designs for computational neuroscience and network analysis.