DEVELOPMENT AND APPLICATION OF A CLINICALLY RELEVANT GLENOHUMERAL JOINT MODEL

Abstract

Introduction: Models of shoulder motion differ with intended application and shoulder models often simplify the complex movement. Therefore, the design often negates clinical usage, in which, for example, multidirectional instabilities are present. To aid the work of clinicians in treating articulations without simplifying physiological constraint, a full open-chain 6 Degrees Of Freedom per articulation has been suggested (Inui et al., 2002).

Aim: Develop a spatial linkage model in order to facilitate communication between surgeon and engineer, and to apply this model to image datasets.

Model Design: Modification of Grood and Suntay’s (1983) 3-cylinder open chain model of the Tibiofemoral articulation to faithfully determine spatial parameters throughout a large range of motion, about clinically relevant axes.

Method: A computer program was scripted (Matlab, Mathworks Inc.) to embed orthogonal coordinate frames in both Humerus and Scapula. These were specified in respect of the planes of clinical rotation and well defined anatomical landmarks. A floating axis was defined within the script as the bipolar common perpendicular to both fixed frames. The magnitude of relative rotations, α, β and γ – flexion, abduction and axial rotation respectively – between Scapula and Humeral frames are measured directly, whilst translations occur along the axis about which rotation is measured. Gimbal lock limitations were minimised.

Validation: A physical linkage was made to validate the computations resulting in further model modification to create continuous rotational data throughout the following range: α from −90° – 270°, β from −90° – 270° and γ from −180° – 180°. This model provided an iterative development and examination tool for enhancing the capabilities of the modelling program.

Application: The model was applied to functional images acquired from both Electron Beam Computed Tomography and MRI. Anatomical landmark coordinates were digitised and input into the customised software. The real-time output displays rotations and translations of the humerus relative to the scapula.

Conclusion: The model circumvents a rotational sequence dependent outcome by determining the joint displacements within the modelled system as independent of the order in which segmental translations and rotations occur: 2 axes are fixed within articulating segments, whist a third mutually perpendicular floating axis moves in relation to both. The method facilitates multi-disciplinary communication: the parameters have a rigorous mathematical description and they correspond to clinical measures of position and orientation. Finally, this method accounts for Codman’s paradox with geometric principles.