Research Of Brain Magnetic Resonance Image Segmentation Technology

Today’s society,with the development of science and technology,the growth of the economy,people more and more attention to their health problems,brain as the body’s central nervous system,detection and treatment of diseases of the brain is Paramount.Accurate craniocerebral three-dimensional model is established to brain surgery,the implementation of remote medical treatment such as brain medical diagnosis,auxiliary and imparting medical knowledge of the brain.Craniocerebral organizations of nuclear magnetic resonance image(MRI)segmentation precision is nuclear magnetic resonance image(MRI)brain images 3D model reconstruction,the basis of quantitative analysis,and visualization,and brain MRI brain image segmentation can be extracted in a special group of quantitative information,can solve the craniocerebral medical image of the sample data in the subsequent analysis work intensity is high,this has become a research hotspot nowadays.But Mr Brain images is a complex medical images,because the brain tissue mixed overlapping each other together there is no clear boundary between,the differences between different individual is bigger,plus the inhomogeneity of the magnetic field in the process of imaging,the partial volume effect and the influence of noise image caused by the inherent uncertainty,and cerebrospinal fluid,a skull and the head cavity in MRI images,gray and very approximate,make the segmentation problem is more complex and difficult.In nondestructive to acquire accurate human body 3D model head,need from brain MRI sequence image segmentation through the scalp,skull,gray matter and white matter and cerebrospinal fluid and other five major groups.In order to solve this problem,this paper proposes a whole segmentation method based on combining FCM&VBM.First by BET method,this method extracts from the original MRI image cerebral area,and then by FCM algorithm to get the brain area is subdivided by grey,white matter and cerebrospinal fluid,then through the VBM segmentation algorithm segmented,scalp,skull and brain cavity and brain area,finally,the segmentation of various organizations smoothly and morphological processing,eventually split out the required five organizational success.With k-means clustering algorithm and morphological segmentation method contrast found in this paper,the segmentationalgorithm in edge gradient under the condition of higher organization has a lower form distortion.