Font Size: a A A

On The Technical Keys Of Medical Image Database

Posted on:2003-10-31Degree:DoctorType:Dissertation
Country:ChinaCandidate:G H GengFull Text:PDF
GTID:1118360062485389Subject:Computer software and theory
Abstract/Summary:PDF Full Text Request
The database technology has become one of the kernels in both the modern computer information system and the development of its application system, It also supports the development of Internet. The combining of Multi-technology with database is a main trend at present. The Picture Archival and Communications Systems is a field emerging in recent years as a result of the development of medical information management. For the PACS how to effectively store .control and retrieve the mass data is an urgent problem needed to be solved. The research of Medical Image Database(Mid-Base) has integrated the techniques of multimedia medical images processing with database .This thesis focuses on the keys of Mid-Base design and its implementation.This thesis presents the structural model of medical information management system and medical images retrieval model and proposes the designing scheme of medical image database from the point of view of the software architecture and system function, which is based on the research of multimedia database and the analysis of the task and function of the PACS. Then through analyzing the feature of the medical image data, we present the concept of using model structure to denote image, and propose the description of processing medical archiving space data surrounding medical image archiving structure and arithmetic. We have also improved the storage structure of the triangle mesh in the surface protract and the arithmetic of fast reconstructing 3D model .We have proposed and realized the algorithms of the color optimization based on the octree, realized the algorithm of data compressing using hafruman code and octree structure.By comparing the technical specialty .working flow, image feature picked-up, computing distance and comparability measurement, we have realized color-based .. shape-based % texture-based retrieval techniques . Considering the actuality of medical image database, we adopted color correlogram image retrieval algorithm combined with Color and Spatial Information; shape based retrieval using invariant moments and edge direction; statistic-based texture description method including gray-level co-occurrence matrix, Gray-Grad- level co-occurrence matrix and autocorrelation function. We also present the results and analysis of related algorithms.One of the characteristics of the Mid-base is the application of the remote access system and 3-D medical visualization system in WAN. Based on CORBA and Web, we set up a remote access system to Mid-Base and its Web-retrival. We have realized the Web-based visualization of 3-D medical image, which provides a foundation for the application of 3-D medical image in WAN and makes the medical image database system universal and invariant to the platform.We have designed and developed an medical image database system as an application of our research in medical image database. This system has been tested and appraised by Science and Technology Bureau of Shannxi province. It has been put into use and proved to be effective...
Keywords/Search Tags:image database, Pacs (Picture Archival and Communications Systems), medical image, image retrieval model, image processing, content-based image retrieval, three layes, Client/Server structure Web/Server
PDF Full Text Request
Related items