Detection and Classification of Tumour in Brain MRI

Abstract

Brain Tumour is an abnormal cell formation inside the brain. They are mainly classified as benign and malignant tumours. Magnetic Resonance Imaging (MRI) is used for effective diagnosis of brain tumour. An automated method for detection and classification of brain tumour is preferred as analysis of MRI manually is a difficult task for medical experts. The proposed method uses Adaptive Regularized Kernel based Fuzzy C-Means Clustering (ARKFCM) for segmentation. A combination of Support Vector Machine (SVM) and Artificial Neural Network (ANN) is proposed for detection and classification of brain tumour based on the extracted features. A dataset of 94 images is considered for validation of the proposed method which resulted in an accuracy of 91.4%, Sensitivity of 98%, Specificity of 78% and Bit Error Rate (BER) of 0.12. Comparison of the proposed method is carried out with other conventional methods and the results are tabulated.