Date of Award
Peter J. Laz
Shoulder, Statistical Shape Modeling
Anatomical variability in the shoulder is inherently present and can influence healthy and pathologic biomechanics and ultimately clinical decision-making. Characterizing variation in bony morphology and material properties in the population can support treatment and specifically the design, via shape and sizing, of shoulder implants. Total Shoulder Arthroplasty (TSA) is the treatment of choice for glenohumeral osteoarthritis as well as bone fracture. Complications and poor outcomes in TSA are generally influenced by the inability of the implant to replicate the natural joint biomechanics and by the bone quality around the fixation features. For this reason, knowledge of bony morphology and mechanical properties can support optimal implant design and sizing, and thus improve TSA results. Statistical shape and intensity modeling is a powerful tool to represent the shape and mechanical properties variation in a training set. Accordingly, the objectives of this thesis were: 1) to develop a statistical shape model (SSM) of the proximal humeral cortical and cancellous bone; 2) to develop an SSM and a statistical intensity model (SIM) of the scapular bone. A training set of 85 humeri and 53 scapulae were reconstructed from CT scans and registered to common templates. Principal Component Analysis (PCA) was applied to the registered geometries to quantify morphological and bone properties variation in the population. For both the humerus and the scapula SSM, the first mode of variation accounted for most of the variation and described scaling. Subsequent modes described changes in the scapular plate, acromion process and scapular notch for the scapula, and in the neck angle, head inclination, greater and lesser tubercles for the humerus. Variation in cortical thickness of the humeral diaphysis was largely independent of size and statistically significant differences with ethnicity were noted. Asian subjects showed higher humeral cortical thickness with respect to Caucasians, regardless of gender. The first mode of variation in the scapular SIM described scaling in material properties distribution, with higher bone density located centrally and anteriorly in the glenoid region. The bone property maps developed for the scapular training set realistically captured inter-subject variability and they represent a valuable tool to assess fixation features and screw location and trajectories for TSA glenoid component. The SSMs and SIM developed in this thesis represent a useful infrastructure to support population-based evaluations and assess possible anatomical differences with gender and ethnicity, SSM and SIM can also provide anatomical relationship in support of implant design and sizing.
Sintini, Irene, "Statistical Shape and Intensity Modeling of the Shoulder" (2017). Electronic Theses and Dissertations. 1478.
Recieved from ProQuest