Medical Image Based On Gpu-accelerated 3d Real-time Rendering Technology

Medical image 3D visualization is the one of popular research projects, It covers many subjects such as computer graphics, image process and medical knowledge, and it is mainly constituted by the pre-processing of medical image.The prime contents of this thesis are 3D visualization of medical images needs much computation, in traditional platforms, CPU is used to compute and hard to achieve real-time display. In this paper, we make deep researches on several visualization methods on GPU including: 3D texture algorithm based on proxy geometries; Ray-tracing algorithm on GPU shader; Marching Cube algorithm implemented by GPU Geometry shader. These methods solve some disadvantages on current CPU version implementations.The key work and innovations of this thesis mainly include:1. Research of the actuality of the pre-processing of medical image, image 3D visualization and the background and programming method of GPU. Implemented visualization methods to make it more flexible to render, different shader files make different rendering effects.2. Research and implement three dimensional texture algorithm based on proxy geometry rendering. This algorithm loads medical volume data as three dimensional texture in the video memory, and sample volume data by displaying proxy geometries with texture coordinates and using Alpha blending to composite the output color.3. Research the implementation of Ray-tracing algorithm on GPU shaders. This method computes each ray direction and position on each Pixel Shader core, and resizes steps by specified sample rates, and sample volume data on current ray position, then step forward until ray over volume data.4. Research classic Marching Cube algorithm for iso-surface extraction and use newest GPU Geometry shader to implement it. It can generate iso-surface realtime.