Research On Edge Extraction Of The Intravascular Ultrasound Sequential Image Based On Active Contour Model

In recent years the cardiovascular morbidity is increasing, so the research and treatment of cardiovascular disease becomes one of hotspots in medical. X-ray imaging and vascular endoscopic are traditional methods of diagnosis of the cardiovascular disease, but these two methods can't get vascular tissue layers of images. With the development of the Intravascular Ultrasound (IVUS) technology and the wide application of the ultrasonic imaging diagnostic equipment, the ultrasonic imaging technology has become one of important methods of diagnose and auxiliary therapy for doctors.This article mainly analyzes and studies IVUS images systematically in the following aspects, based on standard IVUS image data sets provided by Xi'An BitStrong Corporation of Science and Technology. First, images are preprocessed and a large amount of experiments are done in spatial domain. Then a suitable method of preprocessing IVUS images is found out as a result of analyzing preprocessing results of IVUS images. Second, algorithm of threshold segmentation, algorithm of image segmented with Ostu, algorithm of edge detection, algorithm of region growing are implemented, and limitations of these basic algorithms are analyzed. Third, the algorithm based on Active Contour Model is studied. Then analyzing the properties of IVUS sequential image and the calculation process of active contour model, an improved algorithm of Active Contour Model is proposed according to the principle of information measure. IVUS sequential images are processed using the improved algorithm and results show that the improved algorithm can not only improve segmentation precision but also be partly applied to clinical diagnosis. Finally, a prototype system of IVUS sequential image is developed, which can calculate diameter and area of the blood vessel more precisely and provides data support for research and the clinical.