Medical Image Visualization

Visualization in Scientific Computing is being applied in the field of Biomedical Science more and more. It can reconstruct the 3D models from a series of 2D images, which can enhance the diagnosis ability of the doctors and improve the quality of the therapy. In this thesis, the extraction of iso-surfaces from the 3D data sets is studied, which combines the threshold segmentation and Marching Cubes algorithm, and 3D display is implemented by OpenGL.After comparing some popular threshold segmentation methods, such as double peaks, iterative method, otsu algorithm and multi-thresholds image segment method based on potential function clustering, the last one is chosen to segment the multi-component CT image into different parts. 3D Reconstruction is applied to the segmented images based on the classic iso-surface extraction algorithm Marching Cubes. A new method, which is based on searching through the edges of a voxel, is adopted to extract the iso-surfaces. This method needn't construct the iso-surfaces in each kind of voxel in advance and a uniform method can be used to extract iso-surface for different voxel, consequently, the complexity of the MC algorithm can be reduced. At last, the iso-surfaces extracted by the algorithm are rendered in a window with dynamical interaction using OpenGL APIs.