Lateral neural model of binocular rivalry

Area of Science:

• Neuroscience
• Computational Neuroscience
• Vision Science

Background:

• Binocular rivalry is a phenomenon where perception alternates between two competing images presented to each eye.
• Existing models often lack spatial extension and detailed neuronal mechanisms.

Purpose of the Study:

• To introduce a novel two-dimensionally extended, neuron-based computational model for binocular rivalry.
• To investigate the role of lateral coupling and correlated noise in simulating rivalry dynamics.
• To explore the model's ability to reproduce experimental findings and predict new perceptual phenomena.

Main Methods:

• Development of a 2D network model composed of coupled astable multivibrator blocks (excitatory and inhibitory neurons).
• Incorporation of laterally correlated noise to mimic biological systems.
• Simulation of varying contrast ratios to observe changes in perceptual dominance.
• Analysis of model responses to inhomogeneous stimuli.

Main Results:

• The model successfully reproduces the known changes in perceptual time share with varying contrast ratios, attributed to lateral coupling.
• Spatial extension and correlated noise are crucial for generating stochastic oscillations.
• The model predicts periodically moving perceptions, such as propagating fronts and breathing spots, with specific stimuli.
• A bifurcation from static to moving perceptions is predicted under certain conditions.

Conclusions:

• The 2D neuron-based model provides a robust framework for understanding binocular rivalry.
• Lateral coupling is essential for explaining dynamic perceptual shifts observed in experiments.
• The model's predictions offer avenues for new experimental investigations into visual perception and neural processing.
• The model offers a potential interpretation for recent observations of different neuron classes in visual cortical areas.