A Model for Use in Automated Control of Muscle Relaxation During Surgical Procedures

Abstract It is proposed to use an automatated feedback control system to solve the clinical problems associated with the achievement of muscle relaxation during surgical procedures. Due to the variability between individuals a 1 self-tuning' control system would" be required. This paper deals with the problems of determining a suitable model structure and model order for the human response to the muscle relaxant drug curare (or d-Tubocurarine) for use in such a self-tuning control loop. We present a formalism based on linear compartmental theory, and show by examining its structural identiflability how it can be commpared with existing clinical models. We then use off-line non-linear least-squares fitting and the general linear test to obtain a simple bi-compartment model structure, which we show by simulation to be suitable for control purposes and an improvement over existing models.