MRI brain images segmentation

This paper presents an automatic segmentation algorithm for MRI (magnetic resonance imaging) brain image. The algorithm integrates histogram-based fuzzy C-means (FCM) method and connectivity-restrained multi-scale segmentation algorithm to segment the brain into three major classes of gray matter (GM), white matter (WM) and cerebrospinal fluid (CSF). First, histogram-based FCM algorithm is used to segment the images. Then segmented result was refined by connectivity restrained clustering. The algorithm was implemented in MATLAB code. Satisfactory results were obtained.