The functionality and user acceptance of prosthetic limbs can be substantially improved by providing sensory feedback. The goal is to develop a sensory system which prosthetists can incorporate (or retrofit) into presently available myoelectric arms. The system is non-invasive and based on an array of vibrating actuators on the skin (vibrotactors) used to input coded sensory information from the prosthesis to the skin of the residual limb. Three categories of sensory information will be targeted for powered arms: grasp force, object slippage and either wrist rotation (i.e. pronation/supination) or span of finger opening. Prior work in developing vibrotactile strategies for sensory substitution has shown it is a promising approach. The implementation of that research into the marketplace has been held up by a lack of vibrotactors that are small, power efficient and that can provide local stimulation with little cross-talk to adjacent units. Further, optimal vibrotactile feedback codes require that the vibration frequency and intensity be independent. Available vibrators such as cell phone rotary motor based vibrators or traditional solenoid based vibrators have strong co-dependence of vibration frequency and intensity. Further, solenoid based vibrators are only efficient at their resonant frequency. A substantial portion of the Phase 1 effort was directed towards developing and testing a novel engineering approach to the actuators which de-coupled intensity from frequency. A prototype actuator and driving technique was developed which provides for independent frequency and intensity, is small, provides focal stimulation, consumes minimal power and will be inexpensive to manufacture. This prototype was tested on several non-amputee and amputee subjects with excellent results. The goal of this PHASE 2 project is to develop a sensory feedback system for users of powered artificial arms to provide reliable, useful position, force, and slippage information. The vibration actuators will be further developed into a readily manufacturable, inexpensive device. These actuators will be smaller in diameter than a stack of 5 US dimes, weigh less than 10 grams, consume minimal power;be inexpensive to manufacture;be reliable for one year of typical operation, and disposable. Joint position sensors (eg wrist rotation) and force sensors (eg pinch force) will be selected for various prostheses. A microprocessor will convert sensory data from the prosthetic sensor to appropriate actuations using the novel sensory coding algorithms demonstrated in Phase1. The sensory feedback system will be tested extensively with 10 upper extremity amputee subjects to: 1) optimize the coding strategy;2) develop a software training aid;3) demonstrate the reliability of the system components and the system;and 4) determine the improvement in functionality of the prostheses when the sensory system is employed.
TO PUBLIC HEALTH: The incorporation of sensory feedback could greatly improve the quality of prosthesis control. Peripheral knowledge of object contact and grip force and object slippage will allow amputees to have increased confidence when using their prosthesis as well as make it possible to handle fragile items such as finger foods, engage in a hand shake greeting, or provide infant or animal care. Such improvements in quality of life are important for amputees.
