This thesis presents the development of computer vision techniques for a robot-assisted emergency neurosurgery system that is being developed by the Mechatronics in Medicine group at Loughborough University, UK, and situates them within the context of the overall project. There are two main contributions in this thesis. The first is the development of a registration framework, to establish spatial correspondence between a preoperative plan of a patient (based on computed tomography images) and the patient. The registration is based on the rigid transformation of homologous anatomical soft tissue point landmarks of the head, the medial canthus and tragus, in CT and patient space. As a step towards automating the registration, a computational framework to localise these landmarks in CT space, with performance comparable to manual localisation, has been developed. The second contribution in this thesis is the development of computer vision techniques for a passive intraoperative supervisory system, based on visual cues from the operative site. Specifically, the feasibility of using computer vision to assess the outcome of a surgical intervention was investigated. The ability to mimic and embody part of a surgeon s visual sensory and decision-making capability is aimed at improving the robustness of the robotic system. Low-level image features to distinguish the two possible outcomes, complete and incomplete, were identified. Encouraging results were obtained for the surgical actions under consideration, which have been demonstrated by experiments on cadaveric pig heads. The results obtained are suggestive of the potential use of computer vision to assist surgical robotics in an operating theatre. The computational approaches developed, to provide greater autonomy to the robotic system, have the potential to improve current practice in robotic surgery. It is not inconceivable that the state of the art in surgical robotics can advance to a stage where it is able to emulate the ability and interpretation process of a surgeon.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:551292 |
| Date | January 2011 |
| Creators | Ovinis, Mark |
| Publisher | Loughborough University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | https://dspace.lboro.ac.uk/2134/8502 |
Page generated in 0.0019 seconds