Automatic classification and detection of brain tumor using hybrid k-means radial basis function neural network and fast fuzzy C-Means algorithm

This research work presents an automatic detection and classification of brain tumor using a K-Means based Radial Basis Function Neural Network (RBFNN) from the MR images. In the first step the MR images has been segmented by the K-means algorithm and the features are extracted from the images using GLCM (Gray Level Co-occurrence Matrix) feature extraction technique. Further in the second phase the extracted features have been aligned as input to the proposed Hybrid K-Means based Radial Basis Function Neural Network for the classification of brain tumors. The weights of the Radial Basis Function Neural Network are updated by the PSO (Particle Swarm optimization) algorithm and also the centers of the Radial Basis Function Neural Network are chosen by K-means algorithm, so we name the proposed model as Hybrid K-Means based RBFNN model. The malignant and benign tumor has been clustered by the Fast Fuzzy C-Means for visual localization and the performance of the proposed model has been compared with the Fast Fuzzy C-Means, KNN algorithm, and Fuzzy C Means algorithm. The simulation results provide the significance in terms of quality parameters and accuracy.