The Study On Segmentation Of Liver Tumor In CT Image

Liver cancer is one of the most common digestive tract cancers in our country, having a serious threat to human life and health. How to effectively improve the diagnosis and treatment of liver cancer and to maximize the reduction of mortality rate has become an urgent problem in clinical application of medicine. Medical image processing for tumor identification, diagnosis and treatment provides an efficient and accurate technical means, and accurate segmentation of tumor is one of the key steps, not only provides an important basis for the early radiation therapy of tumor, but also directly affects the treatment effect of the patients. Therefore, how to improve the CT image of liver tumor segmentation accuracy rate has got the attention of many researchers at home and abroad.The research content of this subject is the segmentation method of liver tumor in CT images. Firstly, transforming image’s mathematical morphological gradient to enhance its contrast. Then further setting up function relationship between structural elements radius and gradient within specific neighborhood and modifying image gradient, therefore, to enhance the polymerization degree of the target edge, removed the noise and local minimal caused by irregular detail and reduced the migration of object’s contour position at the same time. Finally, using improved level set method to segment single or multiple targets, on consideration of the image gradient information.The main research contents of this paper include the following aspects:(1) By morphological gradient image as the foundation, established the function between the sum of mean variance and the gradient value, to determine the gray level difference of different pixel and its neighborhood; at the same time, determining structural element size by the establishment of the mapping function between the sum of mean variance and the map of structural elements.(2) Different from the traditional closing operation, using multi-scale structure elements of different grayscale neighborhood of gradient image provides the corresponding viscous morphological closing operation to correct point by point, smoothing the gradient image, maintain accurate definition of high gradient edge contour and position, and eliminate local minima exist in the value of the gradient image.(3) According to the characteristics of the image gradient corrected, the target segmentation and background gradient form a good contrast, using improved level set method for object segmentation can effectively inhibit limited factors in traditional segmentation process, and topological structure changes in the process of evolution through the curve, can realize the multi object contour detection in image.Through simulation experiments in clinical liver CT image on the platform of MATLAB7.0, the method proposed in this paper effectively solves the problem of pseudo segmentation caused by the target edge blur in level set algorithm(the highest accuracy can reach 98.9%), and to reduce the computation complexity of traditional level set algorithm(Complete split in 3.24 seconds), applied to the medical image segmentation with low contrast and blurred edge have good results.