A Modified Artificial Neural Fuzzy Inference System Model for Brain Tumor Segmentation

Processing of medical images is quite critical in many applications as they find numerous applications including detection and hence early treatment of many fatal diseases. An efficient and fast brain tumor segmentation algorithm based on an ANFIS model powered by a multi resolution approximation platform is proposed in this paper. Brain tumor detection involves two important stages namely segmentation and classification. The former has been focused in this paper with a modified ANFIS (Artificial neural fuzzy inference system) model implemented for effective segmentation. Image segmentation is used to extract the abnormal portion necessary for volumetric analysis. Segmentation is an integral part of tumor detection as the output of segmentation determines the effect of the treatment on the patient which can be judged from the extracted size and shape of the abnormal portion. The proposed algorithm has been tested with MR brain images and an effective segmentation with fast convergence rate has been observed and reported in this paper. Prior to segmentation, the input image has been pre processed in the frequency domain and features extracted from the low frequency and high frequency sub bands generated from the contourlet decomposition. The algorithms have been developed on MATLAB platform. The simulation proves that compared with the other latest methodologies, the propose method outperforms.