The Research Of3D Segmentation Technology On Coronary CTA Image

Development of medical image segmentation technology has been an importantway of medical image processing. The cardiovascular disease has become the mainthreat to the health of human. So the coronary artery image segmentation has a greatclinical application value in the medical image processing. The accurate segmentationfor coronary artery image can help doctors to diagnose cardiovascular diseaseeffectively, so as to determine solutions for subsequent treatment.CT angiography or CTA is a kind of non-invasive angiographic method. It ispoint to the contrast agent in intravenous iodine by the image processing of computerfor imaging which can display the contrast parts intuitively. With the development ofinterventional radiology, angiography has become an important means in clinicalapplication and plays a very important role in interventional therapy. Angiographyplays an important role in head and neck and diseases of the central nervous system,heart, great vessels diseases, and tumors, and in the diagnosis and treatment ofperipheral vascular disease.The traditional coronary dimensional image segmentation algorithm mainlyaimed at two-dimensional image, however, with the development of science andtechnology, the segmentation of2-D image isn’t accord with doctors and researchersneed, a three-dimensional image contains the pixel’s space information oftwo-dimensional, which can improve the accuracy of segmentation. In view of thecurrent situation of the development of coronary artery segmentation, from theperspective of coronary CTA images3-D segmentation, this paper analyzes thecurrent main medical image segmentation algorithm and compares the advantages anddisadvantages of all kinds of algorithm and application scope in detail. Expounds theprinciple of graph theory algorithm in detail, according to the characteristics of thecoronary artery itself, a coronary CTA images of3-D segmentation algorithm is put forward which realizes the coronary artery image segmentation finally, the researchcontent has completed the following aspects:Firstly, due to coronary artery angiography image has the redundant organizationnear the coronary artery, this paper use the threshold segmentation to deal with thethree-dimensional data and remove the excess tissue and organs, and prepare for thesubsequent segmentation so that can improve the efficiency of subsequentsegmentation effectively.Secondly, use region growing algorithm to process the image after thresholdsegmentation, region growing algorithm can divide the image into several regionseffectively and can make the structure simplify, interested in images of the area ismore prominent.Thirdly, combined with gray and spatial information to construct networkdiagram for the simplified image after threshold and region growing process, andrealize network diagram segmentation in graph cut algorithm for coronary arterysegmentation image.Fourthly, use the visualization toolkit VTK to realize the processing3-Dreconstruction in the result of segmentation and each stepThe experimental results show that, this article combines the advantages of bothregional growth and graph cut algorithm, makes full use of the search of localinformation by the region growing method, and at the same time use the graph cutmethod to control the global information, then the combination of the region growingand graph cut algorithm is applied into the three-dimensional coronary artery imagesegmentation, under the premise that guarantees the accuracy of the segmentation, toachieve the goal of the efficient segmentation, achieved the goal of efficientsegmentation.