Decoding of Force from Neural Signals in Motor Cortex

描述（由申请人提供）：

DESCRIPTION (provided by applicant): The advent of brain-machine interface (BMI) technology has brought with it the expectation that it will be a realistic and practical treatment option for patients with limb amputation or spinal crd injury. However, significant limitations in current approaches that result in a relatively low leve of functionality of prosthetic control need to be addressed before this expectation can be realized. In the current application, we outline three specific aims that tackle some of the major obstacles standing in the way of BMI utility, and propose related hypotheses that address fundamental issues in motor neuroscience. Aim 1: To examine the efficacy and reliability of decoding force output from neural signals in primary motor cortex and dorsal pre-motor cortex during reaching movements in a force field. Our hypothesis is that the neural signals related to the force (kinetic) trajectory of the arm can be decoded in real-time from neural signals in motor and premotor cortex. Aim 2: To test the generalization of force decoding performance across different force fields and over time. Our hypothesis, is that the spatial-temporal patterns of neural activity in multi- channel recordings are specific for the behavioral state (direction of foce trajectory in this case) and that such patterns are robust over time and across different motor environments. Aim 3: To characterize the neural signal associated with self-initiated motor output (movement and force) in motor and pre-motor cortex. Our hypothesis is that the neural signal reflecting self- initiated voluntary motor behavior is detectable in motor areas and can be used to implement 'free-paced' BMI control. The specific aims we propose will be addressed using neural and behavioral data from experiments in non-human primates trained to make reaching movements to spatial targets in the presence of distinct force fields that mimic stiffness, viscosity, and inertia; properties of the motor environment that we deal with on a daily basis. The neural data will be recorded from chronically implanted multi-electrode arrarys and mico-electroencephalography (ECoG) arrays in primary motor (M1) and dorsal premotor (PMd) cortex, We plan to study several different neural signals including single unit activity (SUA), multi-unit activity (MUA), local field potentials (LFP) and ECoG which will give us the added advantage of being able to compare the quality of decoding across the different neural signals