| Since the complex structure and diverse morphology of human tissues and organs, the traditional two-dimensional tomographic data can not meet people’s needs in terms of medical diagnosis, making three-dimensional reconstruction technology has become a hot topic in today’s medical research. Computed tomography, magnetic resonance imaging and other medical data can be used two-dimensional reconstruction techniques to form real,interactive three-dimensional image, which can display a full range of multi-angle complex structures inside the body. Therefore, the 3D reconstruction technique with good academic value and application prospect in medical diagnosis, surgical planning.Baed on the complex structure of the human head and the difficulty of the head diseases’ treatment , volume rendering, surface rendering and other three-dimensional reconstruction algorithm has been studied. Then according to the requirement that the head medical image can be reconstructed and the rendering result can be cut, the paper designs a three-dimensional medical reconstruction system.Firstly, this paper reviews the current development of three-dimensional reconstruction technology in terms of data sources, algorithms, software platforms. Then the Marching Cubes Algorithm especially its ambiguous issues have been researched, and an improved method based on the asymptotic Discriminance has been put forward to eliminate ambiguity.Secondly, the principles and process of the Ray Casting Algorithm is elaborated,designs corresponding two-dimensional transfer functions for the skin, muscle and other different parts based on the voxel scalar value and gradient magnitude.Finally, a three-dimensional medical image reconstruction system is designed.Through the experiment for multiple sets of medical data, the system can achieve functions which include extraction of patient information, adjust and display of single DICOM image,surface rendering, volume rendering and virtual cutting. |
PDF Full Text Request