Research On Filter And Segmentation Of Ultrasound Vessel Image

Medical Ultrasound Imaging, Computer Tomography, Magnetic Resonance Imaging and Positron Emission Tomography are four modern medical imaging techniques. Among them, Medical Ultrasound Imaging become the first choice of disease census and an important way of disease diagnosis due to its unique characteristic of no hurt, no pain, noninvasive, non-ionizing radiation, cheapness and real-time. Meanwhile, because of the rise of the living level, cardiovascular diseases turn into the main disease which impacts on health and leads to death. Ultrasound used in the medical diagnosis of cardiovascular diseases will be of great significance. The task of this thesis is to segment vessels from ultrasound images. Therefore, this study not only has its theoretical value, but also has a very broad prospect in application.Because the impact of inherent speckle noise of ultrasound images, direct segmentation of vessel is difficult. Based on the investigation and reviewing of previous works, this thesis adopts the method of segmentation after filtering.This thesis first gives an overview of the ultrasonic image filtering methods then emphasizes on the analysis of the method based on anisotropic diffusion. By improvement of the CLMC model and the J.Weickert model, a new comprehensive filtering method is introduced. We used the experiment on simulated data and actual data, and evaluated the result by subjective and objective evaluation index. The results showed that the new method has much better efficiency compared to other method referred.Secondly, according to the characteristic of ultrasound vascular images, which can described into some class by grayscale, a variety multi-threshold and clustering of segmentation methods are used to segment the images before and after filtering. The results demonstrated that just simple segmentation method can get satisfying effect and the segmented effect after filtering is obvious better than before doing.Finally, the entire thesis is summed up, and the prospect of the future study is given.