Critical avalanches and subsampling in map-based neural networks coupled with noisy synapses.

Area of Science:

• Computational neuroscience
• Neurodynamics
• Complex systems

Background:

• The brain exhibits various forms of noise, impacting neural activity.
• Understanding neural network dynamics is crucial for deciphering brain function.
• Investigating critical states may reveal fundamental principles of neural computation.

Purpose of the Study:

• To model noisy chemical synapses and identify critical avalanches in neural networks.
• To explore neuronal and synaptic properties necessary for achieving a critical state.
• To inform experimental approaches for detecting self-organized criticality in neural systems.

Main Methods:

• Utilizing dynamical maps to model neurons and synapses.
• Simulating neural networks with specific neuronal and synaptic properties.
• Analyzing spatiotemporal activity for avalanche dynamics.
• Employing data subsampling techniques to mimic experimental measurements.

Main Results:

• Demonstrated that networks of excitable neurons with fast synapses exhibit power-law avalanches.
• Identified rebound spiking dynamics as a key mechanism for generating these critical avalanches.
• Showcased the feasibility of detecting neuronal avalanches through subsampling.

Conclusions:

• Noisy chemical synapses can lead to critical avalanches in neural networks.
• Specific neuronal (excitable) and synaptic (fast) properties are essential for criticality.
• The findings provide a framework for experimental searches for self-organized criticality in the brain.