Optimal texture feature based analysis of mr brain images using wavelet transform

Abstract

In this thesis a novel algorithm for fuzzy segmentation of magnetic newlineresonance imaging MRI data is proposed Intensity inhomogeneities are newlineestimated using fuzzy logic The proposed algorithm is formulated by newlinemodifying the objective function of the standard fuzzy cmeans FCM newlinealgorithm to compensate for intensity inhomogeneities The proposed newlinealgorithm uses Genetic Algorithm GA to optimize the modified fuzzy newlinecmeans function Experimental results on both simulated and real MR newlineimages are given to demonstrate the effectiveness and efficiency of the newlineproposed algorithm newlineThe second major work of this thesis is the identification of optimal newlinetexture features for classification of brain tumors newline newline