Christian Högfors

BIOMECHANICAL MODEL OF THE HUMAN SHOULDER JOINT-II. THE SHOULDER RHYTHM

A method to investigate the rhythm of the human shoulder, i.e. the interplay between the motion of constituent parts of the shoulder, has been devised and tested. The method is based upon numerical evaluation of low dose roentgenstereophotogrammetric motion pictures of subjects equipped with radiation dense implantations in the bones. Evaluation of the method shows that it may be used in determining motion patterns and that the employed interpolation techniques can be used to simulate motions not actually performed in the laboratory. The shoulder rhythm has been previously poorly investigated and quantified results published pertain to one plane only. Our results on motion patterns correlate with previous investigations. With this method, we show that the absolute position of the bones varies significantly between individuals while the relative displacement of the bones during motion exhibit similarities. In particular the results show that, under normal conditions, the individual rhythm is very stable and insensitive to small hand-loads.

Biomechanical model of the human shoulder—I. Elements

In a first step of a biomechanical modelling of the human shoulder the points of application of muscle forces to the bones were determined in a dissection study on four human shoulder specimens. A kinematic description of the shoulder was presented employing bone fixed coordinate systems. Actions of the shoulder muscles were modelled as a system of forces acting along ideal strings. The shoulder complex was treated as a three rigid body twelve degrees of freedom system.

Voluntary redistribution of muscle activity in human shoulder muscles

Four shoulder muscles (the supraspinatus, the infraspinatus, the anterior and middle portion of the deltoid, and the descending part of the trapezius) were examined with electromyography in abducted arm positions. By using feedback techniques, we found that the subjects could reduce the EMG activity voluntarily by 22-47% in the trapezius muscle while keeping different static postures. This was not true for any other muscle investigated. When the trapezius activity was reduced there was a tendency towards an increase of EMG activity in some other shoulder muscles, particularly the infraspinatus. The findings may be related to relaxation from an initial overstabilization of the shoulder, or redistribution of load among synergists. It is suggested that the possibility of reducing trapezius activity may be of ergonomic significance. It is also noted that EMG trapezius activity may not serve as a universal descriptor of total muscular load in the shoulder.

Autobalancing of Rotors

It is shown that plane rotors with one or two particles free to move, subject to viscous damping,, in a groove on the rotor exhibit autobalancing, a property attributed to non-autonomous systems possessing an hyperbolic stable fixed point in the (non-extended) phase space for an open domain in parameter space. The analysis is also extended to non-plane rotors. Dynamical systems of this kind may be represented by perturbed Hamiltonian systems and averaging methods may be applied. Here the driving is considered in the form of a constraint and the particular structure emerging is found to be most easily analyzed by using the method of multiple scales. With this method it is also shown that non-plane rotors are autobalancing. The first order solutions showing how this physically six-dimensional problem approaches the fixed point are also given.

Structure and internal consistency of a shoulder model

A three-dimensional biomechanical model of the shoulder is developed for force predictions in 46 shoulder structures. The model is directed towards the analysis of static working situations where the load is low or moderate. Arbitrary static arm postures in the natural shoulder range may be considered, as well as different kinds of external loads including different force and moment directions. The model can predict internal forces for the shoulder muscles, for the glenohumeral, the acromioclavicular and the sternoclavicular joint as well as for the coracohumeral ligament. A solution to the statistically indeterminate force system is obtained by minimising an objective function. The default function chosen for this is the sum of the squared muscle stresses, but other objective functions may be used as well. The structure of the model is described and its ingredients discussed. The internal consistency of the model, its structural stability and the compatibility of the elements that go into it, is investigated.