Identification and Classification of Brain Tumor from MRI with Feature Extraction by Support Vector Machine
Alok Sarkar, Md. Maniruzzaman, Md. Shamim Ahsan, Mohiuddin Ahmad, Mohammad Ismat Kadir, Saiful Islam
Abstract
A brain tumor is one kind of illness or clutter of the human brain that emerges within the brain of humans for the unusual development of cells that have increment quickly in numbers in a liberal way. Medical image handling is the foremost vital and inventive range, significantly magnetic resonance imaging modalities. The complete work methods presented in this paper are pre-processing, segmentation, feature extraction as well as classification, which are applied to distinguish the brain tumor from magnetic resonance images. By using a support vector machine classifier, we identify at which specific region in the human brain is growing or hold a tumor, and by observing their features we find out which one is benign and which one is malignant. The rates of accuracy, sensitivity, and specificity obtained in this work are 98.30%, 98%, and 100%, respectively.