Segmentation of Tissues from Brain MRI Using Fuzzy Local Gaussian Mixture Model

Abstract

Magnetic Resonance Imaging (MRI) is the heart of medical imaging technology, providing high resolution three dimensional images of soft tissues in human brain. Segmentation of brain tissue in MRI is a crucial preprocessing step in several medical research and clinical applications. Many methods were developed to classify the tissues from the MRI images. MRI medical imaging uncertainty is widely presented in data because of the noise and blurs in acquisition and the partial volume effects originating from the low resolution of the sensors. In particular, borders between tissues are not clearly defined and memberships in the boundary regions are intrinsically fuzzy. The conventional (hard) clustering methods restrict each point of the data set to exactly one cluster. Fuzzy sets give the idea of uncertainty of belonging described by a membership function. Therefore, fuzzy clustering methods turn out to be particularly suitable for the segmentation of MRI medical images. In this paper we focus the attention on the GMM and FLGMM methods for tissue segmentation from MRI Brain images.