The Algorithm Research Of Medical Image 3D Reconstruction And Its Application

Medical image visualization is an important branch of Medical and Visualization in Scientific Computing, and it is always a hotspot of researchers.Surface rendering and volume rendering are the two main methods of medical image 3D reconstruction. This thesis deeply describes the theories of the two rendering methods. Then it uses CT and MRI to reconstruct the 3D image interactively.After deeply describe the Marching Cubes surface rendering method, this thesis proposes two improvements on it. One is to reuse the shared intersected points of the cubes, it can reduce the repeated computation, save the cost of the computation, and boost the running efficiency. The other one is to use multi cores of the modern CPU, it can computes the intersected points parallelly, leading a outstanding speed of surface rendering.On the volume rendering, this thesis first introduces the light model in computer graphics, then analyzes the method of ray casting. At last Shear-Warp acceleration is proposed.The software platform of this thesis is based on Visual C++and an open source image library, using the OpenGL technique to display the 3D model. This thesis proposes some improvements while describe the algorithm. Experiments are carried out under the software with contrast.