Journal of Innovation in Computer Science and Engineering

Brain tumor has been the focus of recent research, most of which is dealing with glioma. Gliomas are the most frequent primary brain tumor which originate from glial cells and grow by infiltrating the surrounding tissue. Tumor structures show a significant amount of variation in size, shape and localization precluding the use of related mathematical priors. Some tumor structures such as necrotic or cystic region or the solid tumor core cannot easily be associated with local channel-specific intensity modifications, but are rather identified based on the wider spatial context and their relation with other tumor compartments. This project proposes a joint generative and discriminative probabilistic model for segmentation of brain lesions. Generative models consider prior information about the structure of the observed data and exploit such information to estimate latent structure from new data. Moreover, it assumes that all voxels of a tissue class have about the same image intensity which is modeled through a Gaussian distribution. The segmentation method, whose parameters can be estimated very efficiently through the Expectation Maximization (EM) procedure, treats image intensities as nuisance parameters which makes it robust in the presence of the characteristic variability of the intensity distribution of MR images the discriminative model extensions maps the output of the generative model to arbitrary label with semantic and biological meaning, such as "tumor core" or "fluid-filled structure", but without a one-to-one correspondence to the hypo-or hyper-intense lesion area identified by the generative model. Experimental results shows that the generative model that has been designed for tumor lesions generative well to stroke images, and the extended discriminativediscriminative model to be one of the top ranking methods in glioma patient scans set.