Location of Thesis Examination

Room 3102 Spencer Engineering Building

Degree

Doctor of Philosophy

Program

Biomedical Engineering

Supervisor

Dr James Johnson

Abstract

Post-traumatic osteoarthritis commonly occurs as a result of a traumatic event to the articulation. Although the majority of this type of arthritis is preventable, the sequence and mechanism of the interaction between joint injury and the development of osteoarthritis (OA) is not well understood. It is hypothesized that alterations to the joint alignment can cause excessive and damaging wear to the cartilage surfaces resulting in OA. The lack of understanding of both the cause and progression of OA has contributed to the slow development of interventions which can modify the course of the disease. Currently, there have been no reported techniques that have been developed to examine the relationship between joint injury and joint alignment. Therefore, the objective of this thesis was to develop a non-invasive image-based technique that can be used to assess joint congruency and alignment of joints undergoing physiologic motion. An inter-bone distance algorithm was developed and validated to measure joint congruency at the ulnohumeral joint of the elbow. Subsequently, a registration algorithm was created and its accuracy was assessed. This registration algorithm registered 3D reconstructed bone models obtained using x-ray CT to motion capture data of cadaveric upper extremities undergoing simulated elbow flexion. In this way, the relative position and orientation of the 3D bone models could be visualized for any frame of motion. The effect of radial head arthroplasty was used to illustrate the utility of this technique. Once this registration was refined, the inter-bone distance algorithm was integrated to be able to visualize the joint congruency of the ulnohumeral joint undergoing simulated elbow flexion. The effect of collateral ligament repair was examined. This technique proved to be sensitive enough to detect large changes in joint congruency in spite of only small changes in the motion pathways of the ulnohumeral joint following simulated ligament repair. Efforts were also made in this thesis to translate this research into a clinical environment by examining CT scanning protocols that could reduce the amount of radiation exposure required to image patient’s joints. For this study, the glenohumeral joint of the shoulder was examined as this joint is particularly sensitive to potential harmful effects of radiation due to its proximity to highly radiosensitive organs. Using the CT scanning techniques examined in this thesis, the effective dose applied to the shoulder was reduced by almost 90% compared to standard clinical CT imaging.

In summary, these studies introduced a technique that can be used to non-invasively and three-dimensionally examine joint congruency. The accuracy of this technique was assessed and its ability to predict regions of joint surface interactions was validated against a gold standard casting approach. Using the techniques developed in this thesis the complex relationship between injury, loading and mal-alignment as contributors to the development and progression of osteoarthritis in the upper extremity can be examined.