Decorrelation of spiking variability and improved information transfer through feedforward divisive normalization.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Visual Cortex Research

Background:

• Correlated response variability is common in similarly tuned visual cortex neurons.
• This variability has been thought to hinder reliable stimulus estimation through averaging.

Purpose of the Study:

• To investigate the role of feedforward divisive normalization in decorrelating neural variability.
• To assess the impact of normalization on the accuracy of stimulus parameter estimation.

Main Methods:

• Simulated feedforward divisive normalization of neuronal inputs.
• Analysis of optimal linear stimulus estimation based on normalized and non-normalized inputs.
• Evaluation of estimation accuracy with varying population sizes.

Main Results:

• Divisive normalization effectively decorrelates the variability of neuronal inputs.
• Normalized inputs lead to more accurate stimulus estimates compared to non-normalized inputs.
• Estimation accuracy improves with increasing population size, up to thousands of neurons.

Conclusions:

• Neurons may utilize divisive normalization as a mechanism to decorrelate input noise.
• This decorrelation enhances the reliability of stimulus information encoded by neuronal populations.
• Further research is needed to reconcile these findings with existing evidence on correlated noise perception.