Research On Several Key Issues In Image-Guided Surgery System

Image-guided surgery has revolutionized traditional surgical techniques. The technology aids in shortening operating times, decreasing the size of the patient's incision, reducing the procedure's invasiveness -all of which can lead to better patient outcomes and faster recoveries. Image-guided surgery system requires image acquisition, image segmentation, creating 3D(three dimension) models, presurgical planning, image-to-physical-space registration, three-dimensional tracking, surgical navigation, archival data storage and etc. This thesis is devoted to several key issues in image-guided surgery system, which are introduced as following. 1)Volume rendering based on histogram volume. In order to avoid the time-consuming processes of manual segmentation and reduce the complexity of setting parameters during segmentation of medical images, this thesis adopts the method of histogram volume to generate an opacity function. Volume rendering is performed with the opacity function. At present, the technology provides the more preferable result with less adjustment and less time than most technologies. 2)Clipping 3D models freely with implicit modeling To meet the requirement of widening the view of surgeons during presurgical planning and surgical navigation, and let they see precisely what they need to see, clipping 3D models freely will be a necessary technology. This thesis adopts the technique of implicit modeling in VTK(Visualization Toolkit) to clipping 3D models. 3)Image-to-physical-space registration A number of classifications of the image registration methods have been proposed. The used classification criteria includes dimensionality, domain of the transformation, type of the transformation, tightness of feature coupling, measure of the registration quality, method of parameter determination, type of data, source of features. The thesis adopts method of point-based registration with four extrinsic markers.