Open Collections

Grzybowski, Georgia Rose

University of British Columbia

Advanced head and neck cancer in the mandible and accompanying soft tissue requires aggressive resection of large segments of the oral cavity including the mandible bone. Accurate reconstruction is essential after these resections to rebuild the proper geometry of the mandible to restore form and function. To reconstruct the proper geometry of the mandible, vascularized donor bone and tissue, commonly from the fibula, needs to be harvested and accurately segmented. The conventional approach is to do this in an iterative, unguided process, which is difficult, time intensive and can produce insufficient accuracy. Virtual planning and 3D printed cutting guides have been developed to improve accuracy, decrease operative time and decrease difficulty. However, these 3D printed guides take a substantial amount of time to create and need to be designed and manufactured 2-4 weeks prior to surgery. Aggressive tumours can grow in that time period, making the plan obsolete. Therefore, we aim to develop an intraoperative optical tracking navigation system and associated tracked surgical tools to guide the execution of virtual reconstruction plans in an accurate and time efficient manner. This thesis will specifically cover the aims of developing and verifying a registration protocol, developing the tracked surgical tools and testing the system on cadaver specimens. Using a combined paired-point and fiducial-model registration method we found we could achieve a Target Registration Error of 1.25 ± 0.06mm on the mandible and 1.86 ± 0.07mm on the fibula on a porcine model. Two optically tracked devices were developed: a mandible fixator device to maintain alignment of mandible fragments and a fibula cutting and placement device to guide the surgeon in executing fibula cuts and subsequently placing those fibula segments into position according to the virtual plan. We successfully used the developed system to conduct 5 guided reconstructions on cadaver specimens with an average accuracy of 1.15 ± 1.17mm in width, 1.47 ± 1.61mm in projection and a Dice score of 0.80. This is comparable to the current 3D printed guide approach, indicating the feasibility of utilizing image guided technology in this surgery.

Text; Other

2022-10-31

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/80102

Biomedical Engineering

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/