Study And Implementation On The Three-Dimensional Visualization Of CT Sequence Image

The way of traditional medical diagnosis has made a revolutionary change since the mergence of the devices of medical images such as CT and MRI. With the help of these devices, the doctors can observe the inner tissue of a human body by using 2D images. It improves the diagnost efficiency and diagnost accuracy at certain extent. Nowdays, the doctors tends to diagnose by observing a series of slices of CT, MRI images. It mainly depends on the doctors' experience. How to get more intuitionistic and accurate information about conditions of patient? The technology of three-dimensional visualization was put on the research agenda.This paper first introduces the structure of the three-dimensional data, the visualizaion process and visualization models of VTK (Visualization Toolkit) and the related knowledge of CT image.The three-dimensional visualization of the volume data has become possible because of the rapid development of computer technology. The visualization of medical images is a cross-cutting reaserch areas of different disciplines. It is the important application of the computer graphics and imge processing in bioledical engineering. The study focuses on the three-dimensional vlisualiztion technology. In this paper, the technology of surface rendering, volume rendering and virtual cutting are reasearched and discussed.Surface rendering technology includes many methods, this article introduces the principles of the algorithm of outline connection, Marching Cubes and Dividing Cubes. At the basis of studying the methods, we presented an improved MC algorithm against the classic MC which produces lots of triangular patches and uses lots of time during the rendering process. the improvede MC algorithm reduce the vertexes which meet the constrain condition to reduce the number of triangular patches, and the experiment proves that the improved MC can reduce the rendering time by reducing the trianglar patches at the premise of not obvous reducing the effect of rendering result. This article also disscusses several algorithms of volume redering, such as Shearp-warp, splating and Ray-casting. Ray-casting used widely is the typical algorithm. Its theoy and implementation, as well as the pseudo-color mapping to the result are the foucus of the paper. In the article, the CT images (512×512×118) are used to implement the algorithms of surface rendering and volume rendering in personal computer with VTK (Visualization Toolkit) and C++.In many cases, the reconstruction of a three-dimensional model is far from enough to meet the needs of the diagosis. In order to obtain more image information in to diagnosis, the article continues to study the virtual cutting technology after the successful implementation of the surface rendering and volume rendering. It includes the virtual cutting of orthogonal palnes, virtual cutting of arbitrary palne and box-shaped cutting. The virtual cutting of orthogonal palnes means cutting the rendering results along the important anatomical direction, while the virtual cutting of arbitrary plane is more flexible. It can cut the results from different angles and different direction to obtain more information to meet doctors' need. Box-shaped cutting is a more complex virtual cutting technology. The six planes of the cubic box have the function of virtual cutting. Then VTK is used to simulate the virtual cutting after the detailed study of the principle of the above-mentioned virtual cutting technology, and we get the satisfactory experiment results.This papper theoretically discusses the principles of visualization algorithm. More importantly, the algorithms are implemented on the computer with VTK. It laid the founfation for assisted diagnosis, has important practical significance.