Chaos and synchrony in a model of a hypercolumn in visual cortex

Area of Science:

• Computational Neuroscience
• Systems Neuroscience
• Neural Network Dynamics

Background:

• Cortical neurons in slices show regular firing, contrasting with irregular firing in vivo.
• Cortical activity exhibits synchronized temporal fluctuations.
• Existing models do not fully capture the irregular and synchronous nature of in vivo cortical activity.

Purpose of the Study:

• To investigate the hypothesis that synchronized chaos in local cortical networks generates observed neural activity patterns.
• To model a hypercolumn in the visual cortex to understand network dynamics.
• To explore cortical mechanisms for orientation selectivity using a computational model.

Main Methods:

• Developed a Hodgkin-Huxley type model of excitable cells for excitatory and inhibitory neuronal populations.
• Incorporated a slow potassium current for adaptation and spatially structured connectivity.
• Performed numerical simulations to analyze network states, auto-correlation, and cross-correlation functions.

Main Results:

• The network model settled into a synchronous chaotic state with temporally variable and correlated neural activity, stabilized by strong inhibitory feedback.
• Identified synchronized oscillations and an asynchronous state in different parameter regimes.
• Demonstrated that the model network exhibits sharp orientation tuning when weakly stimulated, consistent with experimental observations.

Conclusions:

• Cooperative dynamics of neuronal networks can generate variability and synchrony similar to that observed in the cortex.
• The model provides a framework for understanding neural variability, synchrony, and orientation selectivity in the visual cortex.
• Results align with mean-field theory predictions for simplified neural models.