Research And Implementation Of Streaming Technology Framework For Distributed Medical Image

With the development of medical image digitization, medical image subject comes into being for the distributed management and application of large volume medical image data, and medical image information is unavoidable organized in distributed form. In traditional medical image system, the objects are local data and remote image sources must be downloaded through network before operation. This kind of working mode affects the display speed and the doctors need wait a long time before doing forward operation, which brings the inconvenience in medical treatment and becomes the obstacle of the development of medical image information system. How to meet the real-time requirement of remote image sources is an urgent problem for the application of distributed medical image system. Moreover, the variety of the medical image data increases the complexity of this problem. .This thesis, starting from analyses of this problem, brings forward a whole set of solutions for the real time requirement of remote medical image data according to all kinds of medical image format and network environment, which includes the following aspects: It analyzes different types of modern medical image in different points of view, then puts forward the problem of the real-time requirement for the large volume distributed image source data. It introduces the streaming transmission technology applied successfully in the multi-media research area for the network real-time player, the progressive transmission technology and the JPEG2000 specification for large size image display optimization. It also describes the functions of these technologies in medical treatment area and points out the existed problems. Based on the parallel structure and the layer's design principle, this thesis brings forward the Medical Image Streaming Technology Framework. It designs and implements a series of streaming schemes based on the Streaming technology framework for all kinds of medical images (including static and dynamic DICOM image, dynamic MPEG image) under different kinds of network environments. To solve the bottleneck of every layer in medical image Streaming Technology Framework, this thesis puts forward the solution for improving the transmit speed, which includes the design and optimization for FTP client side in translate layer, the MMX and SSE technology employed in decode and display layer, and the DirectDraw technology introduced in presentation layer. Finally, based on the idea of component technology, this thesis designs the streaming component platform and realizes a set of components with key functions. Founded on this platform, it designs and implements the medical image diagnosis workstation. With the software development of medical image diagnosis workstation, as well as its clinical application and testing result, this thesis validates the scheme of the medical image Streaming Technology Framework in practice.