Markovian Segmentation of Brain Tumor MRI Images

Abstract

Image segmentation is a fundamental operation in image processing, which consists to di-vide an image in the homogeneous region for helping a human to analyse image, to diagnose a disease and take the decision. In this work, we present a comparative study between two iterative estimator algorithms such as EM (Expectation-Maximization) and ICE (Iterative Conditional Estimation) according to the complexity, the PSNR index, the SSIM index, the error rate and the convergence. These algorithms are used to segment brain tumor Magnetic Resonance Imaging (MRI) images, under Hidden Markov Chain with Indepedant Noise (HMC-IN). We apply a final Bayesian decision criteria MPM (Marginal Posteriori Mode) to estimate a final configuration of the resulted image X. The experimental results show that ICE and EM give the same results in term of the quality PSNR index, SSIM index and error rate, but ICE converges to a solution faster than EM. Then, ICE is more complex than EM.