Correlated neuronal discharge rate and its implications for psychophysical performance

Area of Science:

• Neuroscience
• Computational Neuroscience
• Sensory Perception

Background:

• Single neurons exhibit high fidelity in signaling sensory information, often matching human perceptual sensitivity.
• Pooling neural responses is expected to improve performance by averaging noise, yet this benefit is diminished if noise is correlated.

Purpose of the Study:

• Investigate the impact of correlated noise on the signaling capacity of neuronal populations.
• Explore the relationship between neuronal population activity and psychophysical decisions.

Main Methods:

• Simultaneously recorded responses from pairs of neurons during repeated stimulus presentations.
• Quantified noise correlations using the correlation coefficient of spike counts.
• Developed a theoretical analysis to assess the effect of correlations on pool signaling capacity.

Main Results:

• Observed relatively weak noise covariation between simultaneously recorded neurons, with an average correlation coefficient of 0.12.
• Theoretical analysis demonstrated that even weak correlations can substantially limit the signaling capacity of a neuronal pool.
• Identified a theoretical link between neuronal population responses and psychophysical decisions.

Conclusions:

• Correlated noise, even at weak levels, significantly constrains the information processing capacity of neuronal populations.
• The findings provide a potential explanation for why psychophysical sensitivity often plateaus at the level of single neurons.
• The theoretical framework may aid in identifying neural populations responsible for specific perceptual abilities.