Main Article Content
One of the categorization tasks in the medical field is brain tumor segmentation which consists on the de-blocking method of tumor area from images. Almost, this job is finished manually by medical specialist who is not always evident due to the equivalence between tumor and normal tissues and the huge variety in tumors. Consequently, medical image segmentation remainders an actual challenge which has attracted the attention of some analyst in last few years. Our thought is to refer this problem as a classification problem where the goal is to eminent in the middle of normal and abnormal pixels on the basis of some characteristics, namely intensities and texture. We using Multilayer perceptron (MLP) and conditional random field (CRF) for correct spotting tumor.