| With the rapid development of 3D printing technology, it has been put into a large amount of application areas, especially in the field of computer medicine. Disease inspection has been a major problem in the field of medicine for a long time. Compared to the diagnosis of skin diseases, the diagnosis of human organ disease is much more difficult for the complexity and diversity of medical images for human organs. However,with the wide application of 3D printing technology in the medical field, the problem is gradually being solved. The 3D modeling technology applied into the medical industry can produce accurate models of various organs, which could radically solve the problem that doctors only being able to diagnose diseases by their own experience. However, there is no universal and efficient method to build 3D surface model from CT image set by now.So far, the modeling methods of 3D modeling technology can be divided into three categories, among which, image based modeling method is used most widely and also the key content of the paper. From the perspective of data processing, the core technology of the 3D modeling based on CT image set includes: obtaining boundary point set of multiple target regions by data processing on CT image set, realizing the construction of 2D contour lines of multiple faults based on that boundary point set, and conducting surface modeling of 3D objects based on multilayer 2D contour lines. After introducing the application of 3D model in medical field and research status of related technologies,the four main parts of this research are as follows:The main objective of the first part of the work is to identify regions of the target organ tissue from medical the CT image set. The specific research steps are given. Firstly,we preprocess the CT image set which is necessary to construct the 3D surface model for realizing the process of reading and displaying the CT image set. Then, we complete the binaryzation of grey value of medical images on the method of dual threshold. Lastly, for the existence of adhesive and fracture phenomenon on these binary images, the goal of separating the target organ and repairing of the fracture boundary is realized by morphological operations, which is proved by the experiment.The main goal of the second part is to identify the boundary point set of its corresponding regions from the target area in the image. The specific research steps are given. Firstly, for the problem that the possibility of the existence of holes in the target connected region,this paper proposes a method,we get the set of interior points in the target area by searching it based on the breadth first search algorithm. Then, the set of boundary points for the target region is obtained by selecting boundary points from internal point set of the target region.The main goal of the third part is to construct 2D contour lines of target areas based on the boundary point set achieved above. The specific research steps are given. Firstly,we create the sequence of polygon endpoints corresponding to the set of boundary points,namely the 2D contour lines of target areas,by a boundary tracing algorithm. Then,this paper presents an improved polygon optimization algorithm, we finish the polygon optimization of those contour lines by using the large amount of redundant endpoints and cusp phenomenon in the polygons found after image recognition, including vertex simplification and smooth curve, easy to build 3D model.The main goal of the last part is to generate a 3D surface model based on polygonal contour profiles of multiple faults in target organ tissues. The specific research steps are given. Firstly, according to the multilayer 2D contour lines, we achieve the generation of triangular facets between layers by the method of shortest diagonal. Then, we complete the triangulation of polygons for some particular top and bottom surfaces of the target organ. Lastly, the triangular patches are saved for the future printing work. |
PDF Full Text Request