Constraints on neural redundancy.

Area of Science:

• Neuroscience
• Motor Control
• Computational Neuroscience

Background:

• Millions of neurons control muscles, allowing numerous neural activity patterns for identical movements.
• Redundant neural activity patterns, or behaviorally equivalent ones, may offer computational benefits.
• However, constraints limiting the selection of these redundant patterns remain unclear.

Purpose of the Study:

• To investigate the constraints that limit the selection of redundant neural activity patterns.
• To determine if principles from muscular redundancy apply to neural activity.
• To explore the relationship between redundant and task-relevant neural activity.

Main Methods:

• Utilized a brain-computer interface (BCI) to precisely define redundant neural activity patterns.
• Recorded neural activity in the primary motor cortex of rhesus monkeys during cursor movements.
• Compared observed redundant neural activity distributions with predictions based on muscular redundancy principles.

Main Results:

• Principles from muscular redundancy did not accurately predict the observed distributions of redundant neural activity.
• A significant coupling was discovered between the distributions of redundant neural activity and task-relevant neural activity.
• This coupling enabled accurate predictions of redundant neural activity distributions.

Conclusions:

• The brain's exploitation of neural redundancy for computation is constrained.
• Redundant and task-relevant neural activity are not independent but coupled.
• Findings suggest limitations on the extent to which neural redundancy can be leveraged for computational purposes.