Research On Image Processing In Virtual Endoscopy

Virtual Endoscopy incorporates with Computer Graphics, Image Processing, Scientific Computing Visualization, Virtual Reality and etc, which is an application of virtual reality in modern medicine simulating optics endoscopy by use of medical images. It is a complete noninvaded, untouched medical diagnosing method avoiding the risk of invading body while the actual optical endoscopy is used. The research objective of virtual endoscopy system is to provide trusty means for medical diagnoses, and to spread the new application fields into assistant diagnosing, surgery planning and training.Several key techniques existing in virtual endoscopy system are studied in the dissertation work.1. With a thorough investigation into the VE technique, an introduction is made on the advantages and features of VE together with its major technical components as well as technical difficulties in image pre-processing, image segmentation, center path planning, reconstructing of 3D and visualization..2. In the phase of image pre-processing, techniques such as image restoration, filtering, strengthening, and interpolating are adopted to the pre-process of CT and MRI slice. The slice contour interpolation algorithm on the basis of distance transform is used to better eliminate steps phenomenon making the target outline smooth.3. Image segmentation algorithm and it's merits and defects as well as its application condition are analyzed. The level set image segmentation algorithm on the basis of simplified Mumford-Shah model is proposed. By energy function with image global features, the evolution curve is made to stay on the border of the objective. It segments images well with noise or with discontinuous border.4. In center path planning, surveys are made on definition and features of center path as well as its extract technology. Center path extract algorithm is proposed on the basis of distance transformation. The center path extraction is fulfilled through distance coding of discrete voxel in 3D data set. Besides, it proposes the VE center path search algorithm by using cellular automata model.5. In 3D reconstruction and visualization, an introduction is made on merits and defects of surface render and volume render representative algorithm in 3D regular data set. A reformed dynamic volume render algorithm is proposed, which has been utilized in VE to gain receiving good effects, Besides, technologies to produce and smooth animation are studied and realized through key frame.6. The validity and feasibility of the algorithm in a human brain virtual endoscopy have been verified with major techniques mention above as well. Finally, prospects have been made on several aspects in further study.