Segmentation and analysis of underlying structures in an image is of paramount importance in medical image processing. The study of many brain disorders involves accurate tissue segmentation of brain magnetic resonance images. Manual segmentation of brain tissues, namely white matter (WM), grey matter (GM) and cerebrospinal fluid (CSF) in MRI by an human expert is tedious for studies involving larger database. The segmentation is complicated by the overlap of MR intensities of different tissue classes and by the presence of a spatially and smoothly varying intensity in-homogeneity due to presence of RF field. The prime objective of this work is to develope strategies and methodologies for an automated brain MRI segmentation scheme. The brain magnetic resonance (MR) image segmentation problem is addressed in supervised and unsupervised framework and is formulated as pixel labeling problem. Stochastic models, i.e. Markov Random Field (MRF) model and Hidden Markov Random Field (HMRF) model are employed to estimate image labels. The notion of Tabu search is hybridised with other optimization algorithms to reduce the computational burden.