3d Medical Image Reconstruction Algorithm Realization

This dissertation mostly studies the technology of three-dimensional (3D) reconstruction from medical images. Medical images 3D reconstruction is an intersection technique of computer graphics and image processing, applied in the field biomedicine engineering.3D reconstruction and visualization of medical images are widely used in diagnostic, surgery planning and simulating, plastic and artificial limb surgery, and teaching in anatomy.Firstly, the paper introduced the way of medical images acquisition, analyzing the decoding technology of CT images based on DICOM standard in detail. After the image acquisition, the image will be pre-processed, includes filtering and interpolating; segmenting and features extraction. Image segmentation, finding the upper and lower threshold values after gray-level histograms analyzing of the whole 3D image and slice images. Mathematical morphology methods are selected according to the features of tissues or regions to be extracted, filling the regions using seed fill algorithm.Secondly, the paper discussed the surface rendering and the direct volume rendering algorithm of 3D reconstruction. By analyzing the characteristic of surface rendering and direct volume rendering technologies, evaluate the performance of these two methods. Combine both methods to build the hybrid rendering algorithm of 3D reconstruction. At last, we compared the different results of the three methods above.Thirdly, the paper discussed the Marching Cubes(MC) algorithm principle and the realization method, analyzed the causes of connections ambiguity and the elimination method in the MC algorithm. For the existence of connection ambiguity in the MC algorithm, discussed the Marching Tetrahedra(MT) algorithm principle and the realization method, analyzes the MT algorithm cuts in half uniformly. It assigns the iso-surface value, and has the MT algorithm structure its 3D model for organization region which divides. Proposed in view of the MT algorithm relevant processing, has avoided the massive duplicated interpolation calculation, has accelerated the reconstruction speed.Finally, by analyzing the class of Visualization Toolkit (VTK) in VC++6.0, I have setup the 3D reconstruction system for medical data surface rendering, interaction operations on the 3D graphics, including translation, rotation and scale, etc..