Automatic Segmentation Approach Based Possibility Theory For The Classification Of Brain Tissues

The paper presents a study and an evaluation of a novel unsupervised segmentation technique based aggregation approach and some possibility theory concepts. Information provided by different sources of MR images is extracted and modeled separately in each one using MPFCM (Modified Possibilistic Fuzzy C-Means) algorithm, extracted data obtained are combined with an operator which can managing the uncertainty and ambiguity in the images and the final segmented image is constructed in decision step. The efficiency of the proposed method is demonstrated by segmentation experiments using simulated MR Images.