Synaptic information transfer in computer models of neocortical columns

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• Understanding information flow in neuronal networks is crucial for neuroscience.
• Brains must balance storing external information with reacting to it.
• The impact of network activity on input-output information transfer remains an active research area.

Purpose of the Study:

• To quantify how neuronal network activity modulates information flow from external inputs to output patterns.
• To investigate the relationship between internal network connectivity and information processing.
• To explore the role of intrinsic network dynamics and oscillations in information handling.

Main Methods:

• Simulations of neocortical column neuronal networks were employed.
• Input-output correlations were measured using Kendall's tau correlation.
• Information flow was quantified using normalized transfer entropy (nTE).

Main Results:

• Increased internal connectivity reduced correlations between excitatory and inhibitory synapses.
• Higher connectivity strength led to a decrease in information flow (nTE).
• Network dynamics contributed additional information, but excessive connectivity corrupted external information.
• Increased gamma power correlated with greater information retrieved from the network.

Conclusions:

• Internal connectivity strength dictates how neuronal networks handle external information.
• Recurrent networks integrate intrinsic dynamics, adding stored information but potentially corrupting external inputs at high connectivity.
• Gamma oscillations may play a significant role in processing information within neuronal networks.