Abstract
Introduction
Advanced medical imaging techniques have allowed the understanding of the patterns of relative bone motions at human joints1. However, poor imaging contrasts of soft tissues have not allowed the full understanding of various glenohumeral ligaments (GHL) functions during glenohumeral joint (GHJ) manoeuvres. This is presently a significant limitation to research as these structures are said to be responsible for the passive stability of the GHJ2. Furthermore, the repairs of GHJ instability often take recourse to these structures3. Earlier studies have presented a model that numerically reconstructs or simulates GHJ motions4 and how the locus of bony attachment points of the GHLs on a dynamic GHJ could be numerically tagged and trailed5. The aim of this study was to advance these previous findings by developing an algorithm that allows the quantification of GHL lengths at any instantaneous position of the GHJ.
Materials and Method
CT scan of a set of humerus and scapula was reconstructed into two individual surface meshes of interconnected nodes, each node having a unique vectorial identification in space. The two attachment nodes (a and b) of a GHL were identified on the bones5. Least squares geometric sphere was fitted upon the humeral head (HH) and its centre (c) and radius (r) quantified6. Vectors a, b and c were applied to represent the ‘dominant ligament plane’ concomitant with the 2D ‘dominant plane’ of Runciman (1993)7. This plane defined the path through which the ligament wrapped on the HH. The point of initial or end of contact of GHL on the HH was defined as the point on HH where a line from c intercepts the ligament at 90°. Total GHL length was calculated as the sum of its three segments, namely: (1) Proximal segment – a straight line from its glenoid attachment node to the point of initial contact (2) Wrap segment – an arc of (r) radius of curvature from initial to end contact points (3) Distal segment – a straight line from end contact point to the humeral node of attachment. The wrap segment was further refined by adjusting ligament contacts along this path to the actual surface contour of the HH by integrating all the surface nodes along the path. The algorithm was tested for short incremental steps of GHJ abduction, flexion, rotation and translations on the Amadi et al's kinematics simulation model4.
Results
From plotted graphs of 5 simulated GHL, lengths increased or decreased smoothly as the rotations and translations were increased or decreased at a constant rate, respectively. Some GHJ motion directions resulted in contrasting stretching or folding effects on different ligaments in a mathematically reasonable manner.
Conclusion
This numerical application would allow the quantification of functional loading of each GHL during simulated or reconstructed GHJ motion and hence provide understanding of how the various GHL may be treated during surgical repairs.
