Learning to attend: modeling the shaping of selectivity in infero-temporal cortex in a categorization task

Area of Science:

• Neuroscience
• Computational Neuroscience
• Cognitive Neuroscience

Background:

• Neuronal selectivity in the infero-temporal cortex (ITC) is crucial for visual processing.
• Previous studies show task learning enhances ITC neuron selectivity for task-relevant features.

Purpose of the Study:

• To investigate the neural mechanisms underlying enhanced ITC selectivity during visual categorization.
• To test the hypothesis that prefrontal cortex (PFC) interactions modulate ITC selectivity.

Main Methods:

• Developed a biologically inspired computational model of spiking neurons with plastic synapses.
• Employed a reward-based Hebbian learning rule to simulate synaptic plasticity.
• Simulated interactions between ITC and PFC neuronal populations.

Main Results:

• The model successfully replicated experimental findings of enhanced ITC selectivity after learning.
• Simulated top-down influence from PFC neurons modulated ITC selectivity for behaviorally relevant features.
• Demonstrated that Hebbian learning strengthens connections between ITC and category-specific PFC neurons.

Conclusions:

• Interaction between ITC and PFC is critical for modulating visual perceptual representations.
• Higher cognitive functions, mediated by PFC, significantly influence visual processing in areas like ITC.
• The proposed model provides a plausible mechanism for experience-dependent changes in visual cortex.