You are here:  Home » Research » Neural Interfaces Program » Research Projects » 

Cortical Control of Neural Prostheses

Skip secondary menu

THIS PAGE IS ARCHIVED MATERIAL.

Principal Investigator Affiliation Contract Number Link
John Donahue, Ph.D. Brown University NS9-2322
Andy Schwartz, Ph.D. Arizona State Univ. NS9-2321

DESCRIPTION/SPECIFICATION/WORK STATEMENT

ARTICLE C.1 BACKGROUND

The Neural Prosthesis Program of the National Institute of Neurological Disorders and Stroke is committed to research and development on functional neuromuscular stimulation (FNS) to restore hand and arm function in quadriplegic individuals at the highest functional level possible. These FNS systems operate under the voluntary control of the individual. Generating control signals for a prosthesis that are integrated with the individual's uninjured, voluntary, cortical motor system is a critical part of any potential FNS system. Presently, control signals are generated by voluntary movements of unparalyzed muscles such as the contralateral shoulder. The goal of this research is to establish the feasibility of generating control signals by direct voluntary control of neurons in the central nervous system (CNS).

To demonstrate that cortical neural activity can be used to control a prosthesis, we need to demonstrate that the neural activity of selected cell populations in the CNS can be reliably recorded for extended periods of time. Then it must be shown that the signals from this population of neurons can, under voluntary control, reliably control an electromechanical device. During the current contract period, it has been demonstrated that a rat can modulate neural activity in its motor cortex to control a mechanical arm to retrieve a food reward. Efforts are underway to duplicate these results in animals with a gyrencephalic brain, but thus far, chronic recording has been difficult to achieve in larger brained animals.

The focus of this research will be on the development of chronic microelectrode recording techniques in a primate with a gyrencephalic brain. There will also be research on extracting control signals from the recorded neural activity. To facilitate communication between investigators and to permit signal processing developed by one investigator to be used on neurophysiological data obtained by other investigators, a standard data format will be used for the exchange of multiunit experimental data among investigators.

The results of these animal studies are needed by the Neural Prosthesis Program to provide information on microelectrode recording array design, recording site selection, number of cells required to provide stable control signals, and the ability to adapt to changes in recorded cell populations. These factors are all critical to establishing reliable control signals from neural activity. This information is also needed before a decision can be made to initiate feasibility studies in spinal cord injured individuals. This RFP represents a competitive renewal of two contracts one at the Arizona State University and the other at Allegheny University of the Health Sciences that will expire in September, 1999. A bibliography listing publications from current and prior NPP supported studies is available on the Neural Prosthesis Program Home Page: www.ninds.nih.gov/npp.

ARTICLE C.2. STATEMENT OF WORK

Independently and not as an agent of the Government, the contractor shall exert its best efforts to demonstrate in a suitable animal model the feasibility of generating control signals directly from CNS neural activity and shall determine the factors that contribute to a reliable neural control system.

I. Specifically the contractor shall:

A. Select appropriate animal model(s) (excluding chimpanzees) and CNS areas where chronic control signals for an FNS system can be obtained taking into consideration factors that shall include:

1. the relationship of the brain area to limb movements,

2. the accessibility of the brain area for microelectrode implantation,

3. the survival of the brain area after various spinal cord lesions, and

4. the existence of a gyrencephalic brain.

B. Obtain or develop an electromechanical arm that can be controlled in at least one degree-of-freedom by neuronal activity from the selected brain areas. The proposed arm shall be capable of delivering rewards to the animal. It shall have a graded response under neural control, and shall be capable of being switched on and off under neural control.

C. Obtain or develop the necessary recording microelectrodes and instrumentation capable of chronically recording populations of single units in the CNS.

D. Obtain or develop a real-time interface between the recorded neural data and the electromechanical arm such that a variety of neural control methods may be evaluated. This interface shall be capable of transforming the neural activity into a control signal with a time delay of not greater than 200 milliseconds.

E. Using the animal model(s) and resources in A through D, chronically implant arrays of recording microelectrodes into the selected CNS area and evaluate:

1. The stability of the neural recordings and the control signals derived from them over periods of at least 4 months giving consideration to:

a. the number of neurons required to provide stable control signals,

b. the effects of natural electrode movement on control signal stability, and the ability of the real-time interface, the coding scheme, and the animal to adapt for electrode movement, and

c. the advantages and disadvantages of using pooled multi-unit neuronal activity recorded from single electrodes compared with separating the activity into single unit activity for the control signals.

2. The ability of the animal to control the mechanical arm giving consideration to:

a. accuracy and rate of information transfer,

b. the ability to use the arm in retrieving rewards,

c. the ability of the animal to learn new patterns of neural activity to optimize control of the arm, and

d. the ability of the animal to control the electromechanical arm after functional loss of its real arms by lesion or reversible block of the spinal cord or peripheral nerves.

F. At the conclusion of the above experiments, sacrifice the animals and determine histopathologically the extent of damage produced by the chronic microelectrodes.

G. Based on the results of the studies in E and F, modify the microelectrode arrays and signal processing system as appropriate to improve the performance of the system.

H. Coordinate your efforts, through the Project Officer, with other investigators in the Neural Prosthesis Program.

Last updated November 24, 2008