| The 3D visualization of medical image is to display the medical information in 3D style,it provide more comprehensive medical information directly, and it has more practical value in clinical diagnosis, treatment and basic medical study. This paper has studied the theories during the liver 3D visualization, mostly on live image segmentation based on abdomen CT images and liver 3D reconstruction.After the summarization of the significance about medical image visualization and the progress of research, the paper discusses the problems including image segmentation, image registering, interpolation, 3D reconstruction. Sequentially, the author analyzes technologies about image segmentation, 3D reconstruction.The technology of image segmentation means extracting interested objects from image, and it is one of key issues in medical image 3D visualization. Based on Mumford-Shah model, Chan-Vese model is a research hot point in Curve evolution combining geometric contour model. Because traditional Chan-Vese model fails to distinguish two regions with same mean but different standard deviation, the thesis take the Gaussian probability density function as the probability density functions of regions interior, calculates the sign of normal force and push the contour move toward the right direction according the difference of Gaussian probability density value. The experiment results of abdomen CT image segmentation demonstrate the improved model can line out the contour of liver, but augment the time of calculate at the same time.The volume rendering algorithm renders the 3D image by simulating a series of optical phenomenon in the object interior. The high-quality image can be obtained by the Ray-Casting algorithm. VTK is a general development toolkit for visualization of data. The efficiency of development software of liver 3D reconstruction can be increased greatly by using the VTK. This paper implements the 3D reconstruction based on the Ray-Casting algorithm using the VTK. |
PDF Full Text Request