Research Of Contour Extracting And Application Using Snake Model For Medical Images Sequence

With the rapid development of the computer technology and the successful application of the medical imaging in clinical diagnosis, computer medical images have played an increasingly important role in clinical diagnosis and treatment. Boundary extraction and segmentation is an indispensable tool of special organizations quantitative information extraction in medical image and it is the basis of analyzing, understanding and handling the images, and it is an important prerequisite to medical visualization. Because of the ambiguity, complexity and diversity of medical image, the conventional edge detection operators extracted the boundaries which are not continuous, and the traditional image segmentation technology brings the disadvantage of imprecise division in the weak edge easily. Therefore, the research on how to extract the boundary of ROI and segment the ROI of medical images accurately is the starting point of this paper.The goal of this project is acquiring more accurate and rapid methods of contour extraction and segmentation for medical image, which is help to increase the accuracy of 3D medical image reconstruction, measurement and identification of ROI, and better to support medical research and clinical diagnosis.This dissertation made a widely review and survey of imaging principle of medical images, the current main methods and techniques of image segmentation, both fully developed and under exploration, and analyzed the respective application spheres, advantages and disadvantages. This paper studied the principle of the classical active contour model in detail, and summed up the different characteristics of the different active contour models. Combining these bases and the inherent characteristics of medical images, this paper proposed a new Snake model from the view of mathematics geometry and artificial intelligence. In this model, a new function of the internal energy was proposed based on the barycenter point of control points and the center of the inscribed circle of triangle of three consecutive control points, which brought better convergence on concave boundary; a novel external energy function was designed using the principle of discard false point, which guided the Snake model curve to deform to real edge more effectively; and a dynamic selection algorithm was raised based on the greatest gradient value which improved the computational efficiency. Qualitative and quantitative analysis of the pros and cons of the novel Snake model was given lastly.At the same time, this paper proposed a new initialization method of the novel Snake model for CT, MRI and other medical images sequence segmentation. We applied this method and the new Snake model algorithm to the medical visualization system of CT, MRI and so on. The experimental results showed that the novel algorithm enabled the medical visualization system to depict ROI accurately and reduce the time complexity. The reconstruction precision and the operational efficiency of the system are improved.The last part of this paper generalized the work of my graduation topic, and put forward the further research ideas.