Content area
Abstract
A major problem facing those trying to solve the forces within the human musculoskeletal system is deciding how to distribute the joint moments amongst muscular protagonists. In this thesis a hierarchical physiological model is developed which can resolve individual muscle forces. The lowest level is an anatomical model which consists of muscle geometry, fibre type and line of action. The data which is necessary to implement the model has been gleaned from various published sources. The middle level is a biomechanical description of a muscle unit which is used to estimate the force each muscle can exert at a certain stimulation given its previous activation and current kinematics. The data defining the mechanical properties of the muscle unit were obtained from the literature and include fibre types, force per unit cross-sectional area, the force-length and force-velocity relationships, activation, elastic and geometrical considerations. At the top of the hierarchy the control level provides a singular solution of the muscle forces which is based on a neurophysiological model of muscle recruitment. At this level muscle protagonists and antagonists are respectively facilitated or inhibited according to their ability to generate the required net muscle moments about those joint axes for which a large range of movement exists.
The model predicts the individual muscle force-time profiles strongly suggested by EMG data. This is in contrast to the reduction and optimization methods reported in the literature. The model also predicts peak muscle forces well within the imposed physiological constraints to generate the measured net muscle moments (peak muscle forces, normalized to isometric, of 0.21 (+OR-) 0.14). Muscle recruitment was confined almost entirely to the slow twitch fibres during the walk and antagonistic muscle activity was demonstrated.
The results partially demonstrate that it is feasible to construct a hierarchical physiological model to evaluate individual muscle forces. This statement must be tempered somewhat knowing that better anatomical and neurophysiological data must be made available and that the model be rigourously validated to use it as a muscle force predictor.
