Open AccessAutomatic classification of MR brain tumor images using KNN, ANN, SVM and CNN
Author(s) -
N. Hema Rajini
Publication year - 2017
Publication title -
international research journal of engineering, it and scientific research
Language(s) - English
Resource type - Journals
ISSN - 2454-2261
DOI - 10.21744/irjeis.v3n1.895
Subject(s) - artificial intelligence , support vector machine , convolutional neural network , computer science , pattern recognition (psychology) , magnetic resonance imaging , artificial neural network , brain tumor , contextual image classification , feature vector , feature (linguistics) , feature extraction , image (mathematics) , medicine , radiology , pathology , linguistics , philosophy
A brain tumor classification system has been designed and developed. This work presents a new approach to the automated classification of astrocytoma, medulloblastoma, glioma, glioblastoma multiforme and craniopharyngioma type of brain tumors based on first order statistics and gray level co-occurrence matrix, in magnetic resonance images. The magnetic resonance feature image used for the tumor detection consists of T2-weighted magnetic resonance images for each axial slice through the head. To remove the unwanted noises in the magnetic resonance image, median filtering is used. First order statistics and gray level co-occurrence matrix-based features are extracted. Finally, k-nearest neighbor, artificial neural network, support vector machine and convolutional neural networks are used to classify the brain tumor images. The application of the proposed method for tracking tumor is demonstrated to help pathologists distinguish its type of tumor. A classification with an accuracy of 89%, 90%, 91% and 95% has been obtained by, k-nearest neighbor, artificial neural network, support vector machine and convolutional neural networks.