Spinal Cord Machine Interface for Neuroprosthetic Control
Abstract
The proposed work relates to "Neuroprosthetics" Topic Area for the Fiscal Year 2014 Peer Reviewed Medical Research Program. Innovations in this proposal will enable the design of neuroprosthetics based on a novel spinal cord machine interface. The goal of this research grant is to study the feasibility of generating control signals from the spinal cord neurons involved in the voluntary control of forelimb movements above the level of injury as a means of constructing a spinal cord machine interface (SCMI). The underlying hypothesis is that spinal cord can serve as a better site for recording control signals as compared to the motor cortex in a brain-machine interface (BMI) that can be used to control an external device such as a prosthetic arm or a robotic device. Spinal cord injury (SCI) occurs as a result of trauma or disease. There are over 270,000 cases of SCI patients in the United States alone, with approximately 12,000 new cases every year. During SCI, the descending fibers in the spinal cord (axons from the brain) are damaged and therefore, they lose their connections with the lower motor neurons (in the spinal cord), which control the skeletal muscles. If the injury is at the cervical level, it results in quadriplegia with little or no function remaining in the four limbs. In these individuals, there is a great need for voluntary command generation. A quadriplegic person can benefit from a voluntary command generation system in different ways. S/he can use this system to control her/his environment (like the control of a robot arm), gain self-mobility (wheelchair control), or have access to a computer to improve independence and quality of life. Current BMIs, which have been developed as a method of voluntary command generation, suffer from several issues such as signal stability, large electrode numbers to sample vast neuronal populations, and the need to sample more than one brain region to increase the BMI repertoire. These challenges have limited the translation of technology from a research setting to activities of daily living. Signals responsible for causing voluntary movements would be better encoded in the spinal cord: (1) as the descending signals have already undergone several levels of information processing at different levels of motor control and (2) due to the close proximity of the spinal cord to the motor output (muscles). As an alternative, the goal of this proposal is to record descending signals in the intact regions of the spinal cord above the point of injury to construct a neural interface for voluntary command generation. Such a neural interface is called a spinal cord machine interface (SCMI) for neural interfacing in paralyzed individuals. The feasibility, long-term stability, and decoding of signals recorded from the marmoset spinal cord will be investigated in this study. Signals will be recorded from devices implanted in the spinal cord while the animals perform a reaching task for food rewards. Various signal processing techniques will be used to demonstrate the information content and characteristics of the neural signals necessary for control of neuroprosthetics. This study will also improve our understanding of how the motor cortex interacts with the spinal cord in producing voluntary movements. Importantly, the data from these initial studies will lay the groundwork for the design of a broader, non-duplicative, set of behavioral and computational approaches to be tested under a National Institutes of Health DP2 Innovator Award. Utilizing a multidisciplinary approach to treat SCI in future may help in improving a spinal cord injured individual s quality of life and lower the medical costs incurred by families, hospitals, and rehabilitation centers.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Apr 04, 2016
- Source ID
- W81XWH1510332
Entities
People
- Abhishek Prasad
Organizations
- United States Army
- University of Miami