A hybrid approach to segmentation of two channels cerebral MR images

Abstract

A hybrid approach is presented in this paper to segmenting the brain matter, as assessed by magnetic resonance (MR) imaging, into three major tissue classes of gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF). First, a fuzzy clustering algorithm is used to divide the original T1 and T2 weighted MR images into groups with similar intensity distributions. Then a multiple level reasoning method is adopted to label the pixels of the cerebral MR image into one of the three of the tissue classes. Finally, the symmetric index is calculated for these tissue classes to show the possible abnormalities in the brain tissues.