Segmentation and classification of brain tumour

Malignant and benign types of tumour infiltrated in human brain are diagnosed with the help of MRI images. The proposed methodology uses fuzzy entropy clustering algorithm with the comparison of SVM, K- Nearest Neighbour and Naive Bayesian classification. The model aims at classifying the tumour region and segment the tumour from MRI images. Fuzzy entropy is the statistical measure of randomness in an image and pixel values that occurs in the probability. The MRI image database used is Whole Brain Atlas. The performance of SVM, K-NN and Naive Bayesian classifiers are compared and finally SVM classifier achieves better accuracy.