Coherence resonance in neuronal populations: Mean-field versus network model

Area of Science:

• Computational Neuroscience
• Nonlinear Dynamics
• Complex Systems

Background:

• Coherence resonance is a phenomenon where intermediate noise levels enhance the regularity of oscillations in excitable systems.
• FitzHugh-Nagumo (FHN) models are widely used to study neuronal excitability and network dynamics.

Purpose of the Study:

• To analytically investigate coherence resonance in populations of FitzHugh-Nagumo neurons with global and local coupling.
• To compare the accuracy of mean-field models with network simulations for capturing coherence resonance.

Main Methods:

• Development and application of mean-field approaches for globally and locally coupled FHN networks.
• Analytical derivation of mean-field limits, particularly for low noise intensities in locally coupled networks.
• Numerical simulations of both network models and their corresponding mean-field approximations.

Main Results:

• The mean-field approach accurately captures coherence resonance in globally coupled FHN networks.
• A derived mean-field limit approximates locally coupled FHN networks effectively at low noise intensities.
• Good agreement was found between network and mean-field models, validating the analytical approach for studying coherence and anticoherence resonance.

Conclusions:

• Mean-field theory provides a powerful analytical tool for understanding coherence resonance in neuronal networks.
• The study validates the use of mean-field limits to predict network behavior, including complex phenomena like anticoherence resonance.
• Findings contribute to the understanding of how network structure and noise interact to influence oscillatory regularity in biological systems.