An introduction to the application of system identification techniques to elements of the neuromuscular system

Abstract

A simplified account of the organisation of the neuromuscular system is presented. An important component of this system, the 'muscle spindle', is a complex sensor that responds to imposed length changes and is thought to provide information which is important in the control of movement and in the maintenance of posture. The structure of the muscle spindle and its response to various inputs are described. The multi-input and multi-output nature and non-linear behaviour of the muscle spindle are emphasised, both in terms of continuous and pulse- coded quantities. The methods of pre-processing of pulse-coded signals from nerves, and the identification techniques currently applied by neurophysiologists for the estimation of the parameters of linearised models of the muscle spindle, are presented and assessed. New developments in the identification of point-process systems are shown to be applicable to some aspects of muscle-spindle behaviour. This approach, based upon point-process parameters, is applied to an analysis of the output signal of a muscle spindle under a variety of input conditions and to the estimation of a transfer function of the muscle spindle subjected to a point-process input. A description of one non-linearity and the effects of an additional input are also presented.