Find Abstract

Custom Search

HUMAN BRAIN SEGMENTATION FROM MRI

This thesis presents 2 methods for human brain segmentation. 1) Multithreshold, determining gray level ranges of each tissue by finding many threshold points . 2) Fuzzy C-Mean Algorithm, using membership of each gray level that depends on cluster center for segmenting classes of gray level. In addition, later algorithm is developed to take a small running time. Results of this thesis show that human brain images can be segmented to several distinct tissues such as gray matter, white matter, cerebrospinal-fluid (CSF) and skull.