Control of spiking regularity in a noisy complex neural network

Area of Science:

• Computational Neuroscience
• Complex Systems
• Nonlinear Dynamics

Background:

• Neural networks exhibit complex dynamics influenced by noise.
• Spiking oscillations are fundamental to neural information processing.
• Understanding noise effects is crucial for neural network function.

Purpose of the Study:

• To investigate the impact of spatiotemporally correlated noise on neural spiking regularity.
• To explore coherence resonance (CR) in Fitz-Hugh-Nagumo neural networks.
• To determine optimal network topologies and noise parameters for enhanced spiking regularity.

Main Methods:

• Simulations of a network of Fitz-Hugh-Nagumo neurons.
• Analysis of spiking regularity under varying noise intensity and correlation.
• Investigation of Watts-Strogatz small-world networks with added long-range connections.

Main Results:

• Optimal spiking regularity observed at moderate noise intensity, indicating coherence resonance.
• Coherence resonance is enhanced in Watts-Strogatz networks with a small fraction of long-range connections.
• An optimal network topology randomness exists for maximizing spiking regularity, dependent on noise temporal (tau) and spatial (lambda) correlations.

Conclusions:

• Spatiotemporally correlated noise can optimize neural spiking regularity through coherence resonance.
• Network topology, specifically the balance of local and long-range connections, significantly modulates CR.
• The interplay between noise characteristics and network structure is key for robust neural signal processing.