Neural response to grasp of robot hand from M1 area of Rhesus monkey

Abstract

A brain machine interface (BMI) is a technology that makes direct connections between neural systems and external device. BMI is often used in platform to restore function for patients with paralysis. In preclinical trials, non-human primate is a preferable choice for animal model especially for upper-limb neuroprosthetics. However, there is gap in the decoder from the animal model to apply the result to human. This is because the animal can use their upper limb in normal circumstances but patients cannot. To overcome this gap, we set up task in which a monkey was required only to observe reaching to grasp motion of robot arm and hand while we were collecting single unit activities from the primary motor cortex (M1). We analyzed the neuronal activities focusing on period of early stage, late stage of movement of robot arm, and grasping of robot hand respectively. There were neuronal groups showing increased activities for each time periods. Some of neurons showed increased activities during late stage of movement and grasping. This result has shown that neural decoder based on observation could be used for reaching to grasp type brain-machine interface in human especially for grasping.