Neural modes in motor cortex cycle over fast timescales

Area of Science:

• Systems Neuroscience
• Computational Neuroscience
• Motor Control

Background:

• Low-dimensional latent states are embedded in neural population activity.
• Individual neuron activity changes over time, but population-level dynamics on short timescales are less understood.
• Adaptation during learning or after injury requires rapid network changes.

Purpose of the Study:

• Investigate changes in coordinated neuron population activity on short timescales.
• Determine if fast and slow drift timescales share common physiological mechanisms.
• Explore the relationship between neural drift and behavioral performance.

Main Methods:

• Tracked individual neuron contributions to population state dimensions in motor cortex during reaching tasks.
• Analyzed changes in encoding patterns over short blocks of repeated trials.
• Applied causal perturbation (electrical current) to motor cortex to induce drift.

Main Results:

• Distinct encoding patterns were fewer than typical motor cortex dimensionality.
• Population state space and dynamics were conserved, but encoding patterns showed fast switches and slow modifications.
• Electrical perturbation induced drift on both fast and slow timescales.
• Increased slow drift correlated with decreased behavioral performance, while fast switches did not.

Conclusions:

• Revealed an additional timescale of drift in correlated population activity.
• Suggests distinct mechanisms for fast and slow drift.
• Highlights the importance of slow drift for maintaining behavioral performance.
• Provides insights into neural mechanisms for structure maintenance, relearning, and recovery.