Electronic Thesis and Dissertation Repository


Doctor of Philosophy


Biomedical Engineering


Tutunea-Fatan, Ovidiu-Remus

2nd Supervisor

Eagleson, Roy



Glenoid reaming is a bone machining operation in Total Shoulder Arthroplasty (TSA) in which the glenoid bone is resurfaced to make intimate contact with implant undersurface. While this step is crucial for the longevity of TSA, many surgeons find it technically challenging. With the recent advances in Virtual Reality (VR) simulations, it has become possible to realistically replicate complicated operations without any need for patients or cadavers, and at the same time, provide quantitative feedback to improve surgeons' psycho-motor skills. In light of these advantages, the current thesis intends to develop tools and methods required for construction of a VR simulator for glenoid reaming, in an attempt to construct a reliable tool for preoperative training and planning for surgeons involved with TSA.

Towards the end, this thesis presents computational algorithms to appropriately represent surgery tool and bone in the VR environment, determine their intersection and compute realistic haptic feedback based on the intersections. The core of the computations is constituted by sampled geometrical representations of both objects. In particular, point cloud model of the tool and voxelized model of bone - that is derived from Computed Tomography (CT) images - are employed. The thesis shows how to efficiently construct these models and adequately represent them in memory. It also elucidates how to effectively use these models to rapidly determine tool-bone collisions and account for bone removal momentarily. Furthermore, the thesis applies cadaveric experimental data to study the mechanics of glenoid reaming and proposes a realistic model for haptic computations. The proposed model integrates well with the developed computational tools, enabling real-time haptic and graphic simulation of glenoid reaming.

Throughout the thesis, a particular emphasis is placed upon computational efficiency, especially on the use of parallel computing using Graphics Processing Units (GPUs). Extensive implementation results are also presented to verify the effectiveness of the developments. Not only do the results of this thesis advance the knowledge in the simulation of glenoid reaming, but they also rigorously contribute to the broader area of surgery simulation, and can serve as a step forward to the wider implementation of VR technology in surgeon training programs.