̽��ϵ��

Topic

Development and In-Vivo Validation of a Subject-Specific Musculoskeletal Modelling Framework for Reverse Total Shoulder Arthroplasty

Department of Mechanical Engineering

Date & location

• Tuesday, April 14, 2026

• 8:30 A.M.

• Virtual Defence

Reviewers

Supervisory Committee

• Dr. Joshua Giles, Department of Mechanical Engineering, ̽��ϵ�� (Supervisor)

• Dr. Marianna Black, Department of Mechanical Engineering, UVic (Member) 

External Examiner

• Dr. Brandon Haworth, Department of Computer Science, ̽��ϵ�� 

Chair of Oral Examination

• Dr. Sandra Marquis, School of Public Health and Social Policy, UVic 

Abstract

Degenerative upper limb pathologies such as rotator cuff arthropathy lead to severely inhibited joint function. To restore strength and range of motion in end-stage disease, Reverse Total Shoulder Arthroplasty (RTSA) may be performed in which the scapula is implanted with a ball-shaped glenosphere, while the humeral head is implanted with a cup-shaped tray. As a result of the drastic change to joint structure, a departure from normative biomechanics is observed in muscle coordination, movement patterns, and joint loading. It is thought that these outcomes are a result of compensatory mechanisms due to complex interactions between individual bony morphology, and implant configuration. 

In-silico methods, specifically Musculoskeletal (MSk) modelling, are well suited for non- invasively quantifying human motion as a multi-body dynamics system with physiologically relevant muscle actuators. Typical modelling workflows approximate individual anatomy through linear scaling, and therefore neglect the full range of anatomic variability. This work aimed to 1) develop a semi-automatic workflow for the development of highly subject specific MSk models; 2) perform inverse simulation to compute biomechanical values of interest, with validation of muscle activity measures; and 3) collect a unique in-vivo dataset useful for model creation and validation. 

Participants (N=3) with RTSA performed two motion tasks, forward elevation and scapula abduction, while synchronized biplane fluoroscopy and optical motion capture kinematics, and Electromyography (EMG) muscle activity were recorded. A previously validated Open Sim model was modified for each subject using MATLAB scripting. Computed Tomography (CT)-derived surface anatomy defined model bony and muscle geometry, while functional joint centres were determined from experimental kinematics. Muscle actuated dynamic simulations were performed that consider the freely moving nature of the scapulothoracic joint, and simulated muscle activations were compared to experimental EMG. Muscle moment arms, muscle forces, and joint loads were compared to those from literature. 

Simulation results broadly suggest the utility of MSk models to consider interactions between individual anatomy, implant configuration, and kinematics. The framework introduced in this thesis rapidly and deterministically generates highly subject-specific models, without the ambiguity and labour-intensivity of manual, often heuristic model creation. The modular processes are modifiable, readily allowing for correcting possible sources of discrepancy, or quantifying model sensitivity to the methodology. These findings may direct future research efforts to achieve the goal of clinically relevant upper limb MSk modelling and simulation.