A Review of the Criteria For Setting Switching Levels in Myoelectric Prostheses


This paper is a response to the observation that adjustment of myoelectric control systems in clinical practice rarely if ever corresponds to what should be the optimum adjustment based upon analysis of such systems. The purpose is to encourage dialogue, in order to achieve better adjustment and/or to identify and correct defects in the analysis.

Most myoelectric prostheses use an estimate of the amount of myoelectric activity to determine which, if any, function to activate. This is true of systems such as those provided by Otto Bock and others, in which a single muscle causes either an "on" or an "off " response, as well as those in which a single muscle may cause activation of more than one function depending upon signal level. In practice the estimate is quite imprecise, because to obtain a very precise estimate would require an unacceptably long time. As well, the amputee cannot be expected to generate precisely a desired level of myoelectric activity. The errors assocated with the estimation process and with the amputee's signal generation are referred to as system error and operator error respectively.

The difference between the maximum signal that the amputee can maintain comfortably and the signal prevailing in the absence of intentional contraction is known as the dynamic range of the myoelectric signal. An analytical solution exits to minimize the chance of having the wrong function activated, (or having a function activated when none was desired), for any specific dynamic range of the myoelectric signal. This solution has been modified to take account of operator error, which has been shown to be significant in comparison with system error.

It is considered likely that the primary difficulty in adjustment of myoelectric systems lies in determining the maximum and minimum signal levels. Accordingly, this issue occupies much of the presentation.

P.O. Box 4400, University of New Brunswick, Fredericton, NB E3B 5A3