The Research And Implementation Of Image Processing Key Technology In3D Human Kidney System

In recent years, with the development of medical image processing technologyand computer tomography technology, ultrasound, magnetic resonance imaging visiblehuman project database and so on, more and more countries set focus on"Three-dimensional visualization" research. As an important technology in"Three-dimensional visualization" research, image processing technology has alsobeen a great deal of development. In this project, using human kidney as a researchobject and using Visualization Korean Human dataset (VKH) as the original materials,a three-dimensional virtual human kidney system was builded by image preprocessing,image segmentation techniques, image registration, three-dimensional reconstruction,computer interactive technology and so on. This article focuses on the keytechnologies of image processing. Specific include the following content:1. An introduction about VRKidney platform. After the introduction aboutVRHuman platform architecture, this paper introduces main function modules ofVRKidney platform. These functions make the platform have the processing capacityfor the original VKH images and provide basic materials for the three-dimensionaldrawing of the kidney. They also make the platform have the processing capacity forthree-dimensional kidney and lay foundation for the follow-up study.2. According to the image processing of two-dimensional slice of the sequence inVKH data set. Firstly, three ways are used to achieve the segmentation of the kidney:the segmentation method based on graph theory, the the segmentation method basedon boundary and manual segmentation method. These methods are also compared.Drawn from the results of the comparison, the method based on graph theory issuperior to the other two methods. Secondly, the segmentation of more than oneobject in the same image is also achieved. The achievement of the function providebasis for rebuilding the entire human body. Finally, this paper also introduces theimage registration based on ITK.3. According to the three-dimensional image processing. The3D image whichobtained by surface rendering or volume rendering often have noisy. This paperintroduces how to remove these noises. In order to facilitate the internalorganizational structure of the kidney, a virtual three-dimensional cutting function isdeveloped. In order to avoid drawing process over and over again, a function whichstore3D image by a document is also developed. In addition, the function that can mark physical properties of kidney is developed. So,3D image has the mainphysiological structure of the kidney. It is important for the study of real physiologicalfunction of the kidney.