Decoding of muscle activity from the sensorimotor cortex in freely behaving monkeys

Area of Science:

• Neuroscience
• Biomedical Engineering
• Robotics

Background:

• Brain-Machine Interface (BMI) technologies show promise for restoring motor function.
• Current BMI applications are limited by studies on restrained subjects, hindering natural movement.
• Decoding motor intent in unrestrained, freely behaving subjects is a key challenge.

Purpose of the Study:

• To demonstrate accurate decoding of muscle activity from electrocorticogram (ECoG) signals in unrestrained, freely behaving monkeys.
• To investigate the neural correlates of movement in both constrained and naturalistic conditions.
• To assess the generalizability of decoding models across different movement types.

Main Methods:

• Recorded ECoG signals from the sensorimotor cortex and electromyogram (EMG) signals from upper arm muscles in monkeys.
• Monkeys performed both forced forelimb movements (lever-pull task) and natural whole-body movements (free cage movement).
• Analyzed high-gamma activity in sensorimotor cortex regions to predict muscle activity.

Main Results:

• Accurate prediction of muscle activity from ECoG signals was achieved during both forced and natural movements.
• High-gamma activity in the primary motor cortex was crucial for both movement types.
• Premotor and somatosensory cortex activity became more important during natural whole-body movement.
• Decoding models trained on forced movements did not generalize to natural movements.

Conclusions:

• Accurate BMI decoding of muscle activity is feasible in unrestrained, freely behaving subjects.
• Naturalistic whole-body movements require broader sensorimotor cortex engagement than constrained movements.
• Behavior-specific decoding models are necessary for effective BMI applications in unrestrained individuals.