Magnetic Resonance Imaging Data From The Automatic Segmentation Of Brain Tissue

In the field of medical image processing and analysis, the medical image segmentation has always been one of the classic problems. And brain magnetic resonance imaging study in the field of medical image is a hot now. The brain is a vital organ of the human body, and it is a complex structure of organization, so brain image segmentation become an important research in the processing of medical image. There are many segmentation algorithm at present, but all have some limitations. The research in view of the brain MRI medical image segmentation is proceeding in the paper.Based on the principle of the nuclear magnetic resonance imaging technology, with summarizing the characteristics of the nuclear magnetic resonance imaging and discussing brain magnetic resonance image segmentation algorithm, the main work of the paper are as follows.First, based on the morphological algorithm for stripping skull, the algorithm is through seeking out the largest connection areas (the human brain part) of images to realize the skull detachment. The first step, use the opposite sex filter operator do image filter to get filter I_filter image. Second, with using the LOG (Laplace) boundary detection operators find out the brain of the boundary to get I_edge boundary image. Then using morphological methods look for the biggest the connected area (brain area) to get a template I_template. Finally, using the template function to get the final image filter to remove the skull of I_skulled after image.Second, after stripping the image of the skull, this paper apply the c-means clustering (Fuzzy C-Means clustering, FCM) algorithm for image segmentation of brain MRI (Magnetic Resonance Imaging, Magnetic Resonance Imaging), realizing the research of segmentation algorithms of brain in the brain of white matter (WM), brain gray matter (GM) and cerebrospinal fluid (CSF). But the medical image usually exist a lot of noise, which can seriously affect the accuracy of the segmentation results, however this algorithm dealing with the noise of the image don’t have a good effect. Therefore, this paper presents an improved fuzzy c-means algorithm based on clustering algorithm, this method comprehensive utilization neighborhood and the pixel neighborhood of information, which is useful for inhibition the noise. Through a novel method of distance calculation replace Europe type distance measure in conventional fuzzy c-means algorithm method, the division can achieve denoising in the process, eventually, spliting the brain organization. Through a large number of experimental and comparison of different algorithm, this algorithm is proved to be effective and feasible.