Theory for the development of neuron selectivity: orientation specificity and binocular interaction in visual cortex

Area of Science:

• Neuroscience
• Computational Neuroscience
• Mathematical Biology

Background:

• The development of stimulus selectivity in the primary sensory cortex is crucial for sensory processing in higher vertebrates.
• Existing models often focus on converging afferents or neuronal nonlinearity, but a general framework is lacking.

Purpose of the Study:

• To propose a novel mathematical framework for understanding stimulus selectivity development in the primary sensory cortex.
• To introduce a new synaptic evolution scheme where incoming patterns compete, influencing synaptic efficacy based on both instantaneous and time-averaged postsynaptic activity.

Main Methods:

• A general mathematical framework was developed to model synaptic evolution.
• The model incorporates a new synaptic evolution scheme dependent on instantaneous and time-averaged postsynaptic activity.
• Simulations were performed using simplified and complex sensory environments, including those mimicking visual cortex development.

Main Results:

• The proposed model demonstrates the development of stimulus selectivity under general conditions, without requiring neuronal integrative nonlinearity or specific intracortical circuitry.
• Convergence to maximum selectivity was shown to occur with probability 1 in simplified environments.
• Simulations accurately reproduced orientation tuning curves and ocular dominance patterns observed in experimental data for visual cortex development.

Conclusions:

• The novel synaptic evolution scheme provides a general and effective mechanism for developing stimulus selectivity in sensory cortices.
• The model's success in simulating visual cortex development suggests its broad applicability to understanding neural development.
• Further experiments are proposed to validate the theoretical predictions of this new framework.