Behavioral Context Determines Network State and Variability Dynamics in Monkey Motor Cortex.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Motor Control

Background:

• Neuronal spiking variability is common in the brain but its context-dependency is poorly understood.
• Spike count variability (Fano Factor, FF) and spike timing irregularity (Coefficient of Variation, CV) measure variability on different timescales.
• The local Coefficient of Variation (CV2) quantifies within-trial spike train irregularity.

Purpose of the Study:

• To investigate the contextual dependence of neural variability in the monkey motor cortex.
• To compare trial-to-trial spike count variability (FF) and within-trial spike timing irregularity (CV2) during movement preparation ('wait') and execution ('movement').

Main Methods:

• Co-analysis of Fano Factor (FF) and local Coefficient of Variation (CV2) in monkey motor cortex.
• Comparison of neural variability metrics during two distinct behavioral contexts: movement preparation ('wait') and movement execution ('movement').

Main Results:

• Fano Factor (FF) significantly decreased from the 'wait' to the 'movement' condition.
• Local Coefficient of Variation (CV2) increased during movement execution compared to preparation.
• Lower CV2 during 'wait' correlated with increased power of local field potential (LFP) beta oscillations and spike-LFP phase locking.
• The relationship FF ≈ CV2² held during 'movement' but not during 'wait' (FF ≫ CV2²).

Conclusions:

• Movement preparation involves dynamic network states leading to high trial-to-trial variability (high FF).
• Movement execution recruits the motor cortex network for reliable, stereotyped output (low FF, higher CV2).
• Findings align with computational models of non-stationary neural network dynamics.