On null models for temporal small-worldness in brain dynamics

Area of Science:

• Neuroscience
• Network Science
• Computational Biology

Background:

• Brain dynamics are modeled as temporal brain networks using functional magnetic resonance imaging (fMRI) signals.
• Validating temporal network hypotheses requires statistical null models that mimic empirical data features.
• Temporal small-worldness is crucial for efficient information exchange in brain networks.

Purpose of the Study:

• To advance the theory of temporal null models for brain networks.
• To introduce the random temporal hyperbolic (RTH) graph model, an extension of the random hyperbolic (RH) graph.
• To evaluate the RTH model's ability to reproduce temporal small-worldness in brain networks.

Main Methods:

• The study introduces the random temporal hyperbolic (RTH) graph model.
• The RTH model is compared against standard null models for temporal networks.
• The models are evaluated based on their ability to reproduce temporal small-worldness in resting-state fMRI data.

Main Results:

• The RTH graph model is shown to be superior to standard null models.
• The RTH model best reproduces the temporal small-worldness observed in resting brain activity.
• The RTH model captures crucial properties of real-world networks, similar to the RH graph.

Conclusions:

• The RTH graph model is a promising tool for validating hypotheses about temporal brain networks.
• Its ability to replicate key features of brain networks with a single additional parameter makes it advantageous.
• This model offers a more accurate null model for analyzing dynamic brain connectivity.