Notes From The Prosthetics Research Program
This is the first of a series of abstracts from reports issued by prosthetics research laboratories in the United States and Canada. The material which follows is a portion of the April 15, 1967, report from the University of California at Los Angeles Biotechnology Laboratory (Engineering Department Report 67-22).
Exploration of New Harnessing Techniques for Transducers
During the current quarter, work continued on the development of a means for attaching "in harness" strain gauge transducers for prosthesis control. A bilateral upper-extremity amputee was employed as a subject for a series of exploratory investigations of promising transducer locations within conventional harness patterns.
Nine control sites were chosen, involving both conventional and non-conventional hookups to a conventional harness. These sites were first tested singly according to the criteria of signal strength, signal-to-noise ratio, maximum velocity, fatigue, and isolation from inadvertent activation. These one-dimensional tests yielded five motions worthy of further consideration. The motions were arm and shoulder flexion, sensed by a strain gauge across the amputee's back, one connected to a point near the end of the stump and the other near the point of the shoulder; arm and shoulder extension, sensed by a strain gauge spanning the amputee's chest, with similar connections on the stump; and shoulder elevation, sensed by a strain gauge connected between the top of the shoulder and a strap across the back.
These control sites were then tested in pairs for efficiency of combined movements and for isolation. Several combinations were found to be possible. Next, three-dimensional combinations were tested. The best triple-site combination, however, was not derived solely from the previously tested sites. Problems of isolation of the previous sites required the use of a new site-the end of the upper arm stump. Signals were produced by contraction of the stump muscles within the socket against a strain gauge mounted on a piece of rubber and inserted in the bottom of the amputee's socket.
This new site, along with the shoulder extension and shoulder elevation sites, permitted the most reliable triple control. Isolation between the three was good; the combination movements were all good with one exception. The amputee's stump was left reasonably mobile.
Development and Testing of Control for a Practical Arm Prototype
A detailed analysis of the upper-extremity amputee-prosthesis system was carried out using principles of control and information theory. An adaptive control system for an upper-extremity prosthesis was designed and simulated. The system was self-adjusting and selfoptimizing. Pilot tests of the system using normal subjects on a manual control task (end-point control) were successful. Further studies with the system with models of arm prostheses are planned. A report describing the design and test of the system is in preparation.