Simulation of plasticity in the adult visual cortex

Area of Science:

• Neuroscience
• Computational Neuroscience
• Visual System Plasticity

Background:

• The adult mammalian visual nervous system demonstrates retinal plasticity.
• Retinal lesions (scotomas) induce reorganization of cortical receptive fields.
• Cortical neurons adapt to represent new visual field areas post-lesion.

Purpose of the Study:

• To investigate retinal plasticity and cortical reorganization using a self-organizing neural network model.
• To simulate the effects of a scotoma on neuronal selectivity and connectivity.
• To elucidate the mechanisms underlying rapid receptive field reorganization.

Main Methods:

• A self-organizing neural network was developed to model normal visual cortex development.
• A simulated retinal lesion (scotoma) was introduced by perturbing the input layer.
• The network's connectivity patterns, particularly intra-cortical and geniculo-cortical, were analyzed post-perturbation.

Main Results:

• The neural network successfully modeled normal development of neuronal selectivity.
• Simulated scotomas led to a reorganized receptive field distribution.
• Reorganization occurred primarily via changes in intra-cortical connectivity, with minimal alterations in geniculo-cortical connections.

Conclusions:

• Self-organizing neural networks can effectively model retinal plasticity and cortical reorganization.
• The study supports the hypothesis that intra-cortical rewiring is the main driver of rapid receptive field map changes after retinal lesions.
• The limited change in geniculo-cortical connectivity may explain the observed short time scale of plasticity.