An efficient detection of brain tumor using fused feature adaptive firefly backpropagation neural network

Abstract

Early diagnosis of tumor will increase the survival probability from a deadly disease called a brain tumor. Classification of the tumor from the tumor affected MRI using the concept of medical image processing assist better treatment and surgical planning. This paper proposes a fused feature adaptive firefly backpropagation neural network for classification which comprises preprocessing, feature extraction, selection, and fusion to achieve high classification accuracy. The preprocessing step uses the average filter for reducing the intensity variation of the images. The Gabor wavelet feature extraction extracts the locality, orientation, and frequency of the tumor image which provides texture information for classification. The kernel principal component analysis (KPCA) feature selection selects the small subset of features to reduce the redundancy and increase the relevancy of the feature. The Gaussian radial basis function (GRBF) for feature fusion provides the distinguished information from the multiple sets of features. Finally, the proposed approach accurately classify the tumor with high accuracy after applying the fusion results. The results are simulated in MATLAB and it proves the improved accuracy, sensitivity, specificity of the classified tumor of the proposed approach.