| Cardiovascular disease which often arises from coronary artery stenosis is an important cause of poor health. Currently, visual estimation based on angiograms is the main clinic diagnosis method. But when evaluating its reliability through the contrast of crown arteries radiography results demonstrated by previous studies, 16-row CT is found with high accuracy in examining in-stent restenosis. Besides, it is a promising non-wound method in this aspect. Combining the computer and medical image processing and analysis techniques, this thesis makes a systematic research on the computer-aided analysis method for vessel extraction and mesurement based on CT images. In view of coronary stenosis, this thesis gives a brief introduction to image smoothing and image enhancement techonology adopted in the process of image pre-processing, and provides several processing examples using CT images. And the methods used include median filtering, multipication enhancing and so on. The adoption of image pre-processing technology can protect image edge information as well as increase the contrast between the object and the background in the image, thus making preparations for further extration of vessels.A study of imaging principles and gray level characteristic of CT images is made in this thesis, and an approach to extracting vessel centerline is provided according to gray level structure charactistic of vessels. This method takes the interactive way, with the starting point and ending point given first and the centerline between these two points worked out automatically using the algorithm.The thesis first deals with the extraction of vessel edges and centerlines, then measure the diametral function of the vessel segment along the centerline, predict its stenosis conditions innitially by means of linear regression, and work out the simulating diameter according to the stenosis condition and finally obtain data on stenosis, thus realizing the quantitative analysis of vessels. |
PDF Full Text Request