A 3D Shoulder Scan Compatible Humeral Coordinate System for Glenohumeral Joint Kinematics

Advanced surgical planning techniques often require modeling the functional characteristics of the affected body region. Most patient-specific modeling in vivo relies on medical image scans that are expensive and may also allow patient’s exposure to ionizing radiation. This poses a challenge for the modeling of the kinematics of the glenohumeral joint (GHJ) based on the tissue geometries of the affected patients. The humeral morphology uniquely presents its canal (HC) and epicondyle (EC) axes as the two longest axes that are nearly orthogonal. This gives them the mathematical advantages as best axes for the definition of humeral coordinate system (HCS), especially from 2D radiographic images. This is however limited in 3D in vivo kinematics as minimization of radiation exposure may not allow medical imaging of the whole volume of interest all the way down to the distal epicondyles. It is therefore necessary that landmarks for use are captured within the field of view (FOV) of standard shoulder scans. This would avoid extra radiation exposure to patients and imaging cost as the scan might have been used earlier for traditional diagnosis. The aims of this study were to (1) confirm that HC-axis quantified from a ‘stack of discs (SOD)’ technique was the most reliable and consistent (2) identify the most closely oriented or most inter-subject related axis to the EC-axis for its replacement or prediction respectively from 3D proximal humeral scan and (3) use these to propose a HCS definition procedure that can be applied to a standard shoulder scan.