Internal models engaged by brain-computer interface control.

Area of Science:

• Neuroscience
• Motor Control
• Brain-Computer Interfaces

Background:

• The brain predicts sensory consequences of movement to handle feedback delays.
• Internal models are evidenced in oculomotor, vestibulo-ocular, and skeletomotor systems.
• Neural population activity in closed-loop brain-computer interface (BCI) control offers a new avenue to study internal models.

Purpose of the Study:

• To investigate evidence of internal models in simultaneously recorded neural population activity during closed-loop BCI control.
• To determine if the brain predicts upcoming cursor positions to compensate for visual feedback delays.
• To examine the adaptation of internal models following changes in the BCI mapping.

Main Methods:

• Recorded population activity from nonhuman primate motor cortex during cursor-based BCI control.
• Utilized a novel BCI task to measure visual feedback processing delay (approx. 130 ms).
• Analyzed population activity at different time lags to assess prediction of cursor positions and internal model adaptation.

Main Results:

• Evidence suggests the brain predicts upcoming cursor positions to compensate for a ~130 ms visual feedback delay.
• Population activity analysis revealed characteristics consistent with an internal forward model.
• The study observed the time course of internal model adaptation after altering the BCI control mapping.

Conclusions:

• Closed-loop BCI experiments provide insights into the neural basis of feedback motor control.
• Internal forward models are utilized in motor cortex to predict sensory consequences and compensate for feedback delays.
• This approach offers a powerful tool for studying motor learning and neural control mechanisms.