Segmentation of Magnetic Resonance image

the paper presents a k-means algorithm based on image grey level,and test the brain's MR image segmentation by this algorithm.The gray value of the pixels in the image gives the consistent property with the cluster-center pixels.The multi-regions in a image can be merged together according to the gray value.The application of this algorithm on MR image segmentation is successful.The results showed white matter,grey matter,the cortex,the ventricles of the brain and the background.