Invariant neural dynamics drive commands to control different movements.

Area of Science:

• Neuroscience
• Motor Control
• Computational Neuroscience

Background:

• The nervous system's ability to produce varied movements may stem from reusing invariant neural codes.
• Previous research observed similar neural population activity dynamics across different movements.

Purpose of the Study:

• To investigate if invariant neural population dynamics are actively used to generate motor commands.
• To determine how these dynamics relate to the commands issued for movement control.

Main Methods:

• Utilized a brain-machine interface (BMI) to decode motor cortex activity in rhesus macaques.
• Transformed neural activity into commands for a neuroprosthetic cursor.
• Analyzed neural activity patterns and their temporal dynamics during different movements.

Main Results:

• Identified that distinct neural activity patterns can produce the same motor command.
• Discovered that transitions between neural activity patterns follow consistent, invariant dynamics across movements.
• Demonstrated that these low-dimensional invariant dynamics align with the BMI, predicting command-relevant neural activity.

Conclusions:

• Invariant neural dynamics are crucial for generating a variety of movements.
• These dynamics facilitate the transformation of feedback into generalizable motor commands.
• A model of optimal feedback control (OFC) supports the role of invariant dynamics in reducing neural computational load.