Restoring hand function following spinal cord injury has tremendous potential to improve the quality of life and independence of 11,000 patients annually. The goal of this proposal is to develop a novel Brain Machine Interface (BMI) that bypasses the injury by using the brain's natural signals to control electrical stimulation of the paralyzed muscles. BMIs have the potential to restore motor function lost due to disease or injury by recording neural signals directly from the brain and using a computer to interpret the signals to drive an actuator. Most current BMI research focuses on using primary motor cortical (M1) signals for kinematic control of a computer cursor or a robotic arm. However, it is increasingly clear that M1 includes a heterogeneous representation of the kinematics (position, velocity, etc) and dynamics (force, torque, etc) of limb movements. If M1 signals are more comparable to muscle activity than kinematics, a BMI to drive a dynamic actuator might be easier to use and require less time to learn. By using signals related to the dynamics of natural movement to cause artificial contraction of the paralyzed muscles, it should be possible to restore a wide repertoire of hand movements to spinal cord injured patients with minimal adaptation required on the part of the patient. Functional electrical stimulation (FES) is the application of electrical current to nerves or muscle to augment or restore function in neurologically impaired individuals. Toward this end, monkeys in our lab were trained to use a unique BMI designed to control FES, allowing them to exert isometric flexion and extension torque about the wrist. Multi-electrode recordings from monkey M1 and an isometric and movement version of a 2D wrist center-out task, that requires the monkey to control a computer cursor from a central target to one of eight peripheral targets, will be used to characterize M1 neurons and compare BMI decoders. The goal of aim 1 is to characterize M1 neurons based on how their tuning curves change in response to a postural rotation of the forearm that changes the relationship between the muscle activity and the direction of torque. The time course of behavioral adaptation will be compared to any underlying neural adaptation within single sessions and across days, with the expectation that stable subpopulations encoding a range of kinematic and dynamic information exist in M1. The goal of aim 2 is to develop a 2D version of a BMI controlled FES wrist task. BMI control of the FES task will be compared to a BMI that directly controls the computer cursor. The neurons characterized in aim 1 will guide further experiments, to determine the optimal input neurons for FES and cursor control. Neurons will be selected whose characteristics best match the dynamics of the decoder. This process is expected to reduce the required adaptation time for both dynamic and kinematic decoders. These two aims provide the means to characterize a mode of BMI control which resembles natural motor control signals, and to use such a BMI to control FES to restore 2D wrist control to temporarily paralyzed monkeys.
Restoring hand function following spinal cord injury has tremendous potential to improve the quality of life and independence of 11,000 patients annually. The goal of this proposal is to develop a novel Brain Machine Interface (BMI) that bypasses the injury by using the brain's natural signals to control electrical stimulation of the paralyzed muscles. By developing a mode of control which closely approximates natural movement, restoration of a wide repertoire of hand movements to spinal cord injury patients should be feasible with minimal adaptation required on the part of the patient.
|Oby, Emily R; Ethier, Christian; Miller, Lee E (2013) Movement representation in the primary motor cortex and its contribution to generalizable EMG predictions. J Neurophysiol 109:666-78|