Research On The Key Techniques Of Virtual Endoscopy

Virtual Endoscopy (VE) is an advanced medical image post-processing system emerged only several years ago. It is an integration of medical imaging, image processing, computer graphics and virtual reality techniques. Since it is noninvasive, untouched and free of risks, VE is widely used in diagnostics and anatomic teaching recently.Some key techniques and realizations of VE are systematically investigated in this thesis. Firstly, the medical image pre-processing techniques are surveyed, and a novel algorithm is proposed to segment tissues efficiently in the 3D volume data using the fuzzy connectedness. Secondly, in order to realize the path-planning in VE system, a new centerline extraction algorithm for the tubular organs based on the distance from boundary field is presented, which used fast distance transform and dynamic programming to create the navigation path. Thirdly, in the study of scene rendering in VE system, to overcome the surface noise introduced by normal vectors in the standard Marching Cubes algorithm, a new normal vector smoothing algorithm is proposed, which further improves rendering effect of the VE system. Finally, based on the above key techniques, a VE system with friendly interactive functions is developed, and the architecture and user interface of the VE system is also provided.The experimental results with the real clinic image data demonstrate the effectiveness and practicability of the proposed algorithms.