Research Of 3D Vascular Diagnostic System

Health is one of the most popular topic,people focus on how to improve the environment and quality of their living,particularly,the cardiovascular has been the first threat,in research,with a large number of the medical auxiliary means coming,such as CT,MRI,this situation has turned well,but the diagnosis means only in 2D slices displayed human organs,which often rely on abundant experience to achieve accurate diagnosis,the diagnosis is not intuitive,and not enough experience for the doctor,this is a problem,while as computer technology developed,the visualization brought the dawn to this problem,our target is on angiography tomography To achieve three-dimensional reconstruction and extraction of three-dimensional model of the axis,for doctors to provide more convenient and accurate information for the diagnosis of vascular diseases.In this paper,according to the needs of the three dimensional model of vascular diagnosis,the experimental results are compared with the mature pretreatment methods,which provides high quality images for the next step.In depth analysis of the surface reconstruction and voxel reconstruction algorithm,with the aid of the open source library VTK to achieve the slices reconstruction.All kinds of algorithm on image segmentation,the paper proposes a segmentation method of improved 2D maximum cross entropy,the method by Gauss distribution histogram to slice set random sampling anti approximation probability histogram,instead of the traditional direct use of Gauss distribution calculation of target and background,the threshold vector is more reasonable.From the two-dimensional distance transform skeleton extraction methods,this paper proposes a method of calculating distance transform based on dynamic programming and regional search,analyzes the advantages and disadvantages of the algorithm.And the method is applied to 3D distance transform to extract the axis of three-dimensional blood vessel.In the acquisition of "coarse" axis,the axis of pixel accumulation or saw such deficiencies,through the fairing of removing redundant pixels on axis is optimized by interpolation algorithm.Finally,from the points of view of software engineering,this paper discusses the processing of a simple 3D visualization aided diagnosis system based on the research content.Mentioned in this article and the achievement of the content cannot meet the real demand of the practice.