Constructing representative group networks from tractography: lessons from a dynamical approach.

Area of Science:

• Neuroscience
• Network Science
• Computational Biology

Background:

• Human group connectome analysis combines individual data for representative networks.
• Current methods for selecting group network features lack clear evaluation criteria.
• The suitability of structural features for group-level analysis remains uncertain.

Purpose of the Study:

• To investigate the impact of including dynamical behavior in defining a representative group connectome.
• To evaluate if a dynamically representative connectome better recapitulates individual network structures and dynamics.
• To compare a dynamically optimized group connectome with those derived from structural optimization methods.

Main Methods:

• Applied consensus thresholding to individual structural connectomes from healthy adults.
• Constructed group networks across various thresholds.
• Measured dynamical behavior using a coupled oscillator model.
• Identified the most dynamically representative group connectome based on minimal deviation from individual dynamics.

Main Results:

• The dynamically representative network effectively recaptured structural aspects without explicit optimization.
• No significant structural differences were found between the dynamically optimized network and those optimized for specific metrics.
• Other structurally optimized networks were either equally or less dynamically representative.

Conclusions:

• Dynamical behavior should be a key criterion for defining representative group connectomes.
• The study highlights the importance of functional dynamics in understanding brain organization.
• Further research is needed to identify valid and testable criteria for representativeness in connectome analysis.