NEURAL NETWORK INTERFACE FOR UPPER EXTREMITY PROSTHESIS CONTROL

A NATURAL FUNCTION PROSTHESIS INTERFACE FOR AMPUTEES AND VICTIMS OF UPPER EXTREMITY DYSFUNCTION CAN BE ACCOMPLISHED USING AN INTELLIGENT CONTROLLER TO INTERPRET THE AVAILABLE NEURAL SIGNAL DATA FOR CONTROL OF A ROBOTIC ARM. THE GOAL IS TO DEVELOP A SUFFICIENTLY SOPHISTICATED NATURAL FUNCTION INTERFACE WHICH USES NEURAL NETWORKS TO PROVIDE SIMULATED MUSCULAR CONTROL AND REFLEX. THE OBJECTIVES OF THIS PROJECTINCLUDE EVALUATION OF AVAILABLE NERVE SIGNAL SENSORY INTERFACES AND SENSORY FEEDBACK MECHANISMS, DEVELOPMENT OF A SOFTWARE BASED NEURAL SIGNAL SIMULATOR, DESIGN OF A NEURAL NETWORK CONTROL INTERFACE SYSTEM, IMPLEMENTATION OF NEURAL NETWORK SOFTWARE TO PERFORM TRANSLATIONS FROM SIMULATED SUBJECT NEURAL SIGNALS TO PROSTHETIC CONTROL SIGNALS, DEMONSTRATION AND OPERATIONAL TESTING OF THE CONTROL INTERFACE SYSTEM, AND OPTIMIZATION OF THE INTEGRATEDCONTROL INTERFACE SYSTEM TO PROVIDE THE GREATEST CAPABILITY AND MOST NATURAL FUNCTION. QUESTIONS OF ERGONOMICS, PROGRAMMED REFLEXES AND THE APPLICATION OF NEURAL NETWORK COMPUTERS IN AN INTELLIGENT CONTROLLER WILL BE ADDRESSED. IN PHASE II, THE DESIGN OF THE INTELLIGENT CONTROL INTERFACEBEGUN IN PHASE I WILL BE CONTINUED AND WILL RESULT IN A PORTABLE SYSTEM CAPABLE OF PROVIDING FINE MOTION CONTROL OF UPPER EXTREMITY PROSTHETICS. THIS SYSTEM WILL BE MADE COMMERCIALLY AVAILABLE UNDER A PHASE III PROGRAM BACKED BY PRIVATE OR VENTURE CAPITAL FUNDING.