Comparison of an EMG-controlled prosthesis and the normal human biceps brachii muscle

Abstract

An electromyographically-controlled elbow prosthesis, the "Boston arm", was tested with constant and sinusoidally-modulated inputs under different loading conditions. These tests were used to determine the frequency response of the EMG amplifier, the range of motor speeds, the force-velocity curve, and the frequency response of the prosthesis for cyclic movements. Human biceps muscle was tested during movements under similar loads, and under isometric conditions using random, electrical stimuli to determine its frequency response. The results were compared to evaluate the performance of the prosthesis and contrast its characteristics to those of the physiological system it replaces. The prosthesis minimizes quiescent power drain, and the possibilities of instability or "tremor" by employing velocity control, an overdamped mechanical system, and an electrical "dead-band". These factors limit the ability of the prosthesis to follow rapidly changing inputs or to move rapidly and accurately to a desired position. These comparisons should provide a useful supplement to the usual clinical evaluations, and a starting point for further improvements in prosthesis design.