3D Reconstruction Of Coronary Arteries Using MSCT Images

CAD is one of the most severe diseases that threaten the human beings, furthermore, population suffers from it is increasing and the sufferer trends to younger. Therefore, it is important to detect CAD earlier, so that doctors can make early treatment for the patient. Multi-slice Spiral Computed Tomography (MSCT), especially 64-slice spiral CT, provides a non-invasive way for the clinical diagnosis of CAD, for its high spatial and temporal resolution and thinner slice thickness.It is difficult to reconstruct the 3D image of the coronary arteries due to their complex physiological modality and interference of the heart movement. Thus, it is just on the threshold of the coronary arteries research till now.The aim of the thesis is to do some investigation on 3D segmentation of human coronary arteries. We focus on an approach of 3D segmentation of coronary arteries based on the latest framework of 3D segmentation method, expand the theory and make some improvements.The contents are listed below:Analyze the history and current situation of the reconstruction of coronary using MSCT imagesIntroduce relative physiological, pathological knowledge, CT knowledge and two toolkits for image processing and visualizationResearch common methods for the pretreatment, segmentation and reconstruction of MSCT imagesDiscuss the latest framework for 3D segmentation of coronary arteries using MSCT images, supplement and modify some of its theoriesUtilizing ITK, implement and validate the modified method for 3D segmentation of coronary arteries using MSCT imagesAnalyze and evaluate the modified methodIt has been proved that left and right coronary arteries can be extracted from MSCT images with the 3D segmentation arithmetic based probability decision and active contours.