Causal network in a deafferented non-human primate brain.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Robotics

Background:

• De-afferented neural ensembles can adapt when interfaced with neuroprosthetic devices.
• Adaptation involves changes in neuronal recruitment, isolation, functional connectivity, and roles (excitatory/inhibitory).

Purpose of the Study:

• To demonstrate the emergence of a causal network and dynamic changes in a deafferented brain region during brain-machine interface (BMI) learning.
• To investigate neural plasticity in response to neuroprosthetic control of robotic systems.

Main Methods:

• Utilized a brain-machine interface (BMI) for robotic reach-and-grasp control.
• De-afferented motor cortical regions from chronic amputation were used.
• Employed a generalized linear model-framework based Granger causality (GLM-GC) technique for ensemble connectivity estimation.
• Model selection was guided by the Akaike Information Criterion (AIC).

Main Results:

• Demonstrated the emergence of a causal network in the deafferented brain region after BMI learning.
• Observed significant changes in neural ensemble dynamics during the learning process.
• Successfully decoded neuronal ensembles for velocity control of a multi-degree-of-freedom (DOF) robot.

Conclusions:

• Neuroprosthetic interfacing can induce causal network formation and dynamic alterations in deafferented neural ensembles.
• BMI learning facilitates adaptation and functional reorganization in motor cortical regions post-amputation.
• This research highlights the potential for neuroprosthetics to restore or enhance motor function through neural plasticity.