The Research Of Volume Data Segmentation And Visualization For Medical Image

The rapid development in the medical imaging field has promoted the progress of modern medicine greatly. Computed tomography (CT), magnetic resonance imaging (MRI) and other imaging modalities have been widely used in clinical diagnosis and therapy. The main thesis of this paper is: in the computer-assisted, extracting regions of interest (ROI) from medical images and displaying them through 3D visualization techniques, which can help doctor observes the pathological regions more directly and clearly, consequently improve the accuracy of diagnosis.In the field of medical image processing and analysis, segmentation has become one of the classic problems, it is the basis of the following processing, such as the 3D visualization of tissue, surgery simulation and image guided planning. The accuracy of the segmentation is vital for the doctor to make a judgment on the diseases and draw out a corresponding curing plan. Therefore, segmentation is meaningful in the medical application, and is the precondition of the visualization in scientific computing. Medical image visualization techniques play an important role in medical image processing and analysis, and it's one of the most successful applications of visualization in scientific computing. There're two main technologies in 3D image visualization, one is surface rendering, the other is volume rendering. Since volume rendering can reserve rich information of the volume data, and can show the internal relationship of the organizations, this paper uses volume rendering algorithm to show the effect of 3D segmentation.In this paper, we introduce the basic knowledge of medical image 3D visualization at first; generalizes several 3D reconstruction algorithms which are commonly used. Secondly, this paper probes into the field of image segmentation. It summarizes the current main methods and techniques of image segmentation, both fully developed and under exploration. Furthermore, it analyzes of how to apply image segmentation algorithm to volume data. Thirdly, volume data is segmented by several algorithms in experiments, and this paper analyzes these algorithms'respective application spheres, advantages and disadvantages. At last, we realize a complete medical image 3D visualization system, aiming at the medical CT fault images of DICOM 3.0 standard. The experimental result shows, this study is practical on the volume data's segmentation and visualization, which provides a good assistant function in the medical teaching and clinic diagnosis.This paper discussed the key technologies of volume data segmentation. In the processing of segmentation, the series of CT images were dealt as a whole volume data for processing. This study introduced several 3D segmentation algorithms to process a sequence of CT images. Furthermore, in order to improve the efficiency of segmentation, a split-and-merge algorithm base on octree was introduced, which can segment volume data accurately and efficiently.The last part of this paper generalizes the work of my graduation topic, and puts forward the further research ideas.