A Hybrid Method for Brain Tumor Detection using Advanced Textural Feature Extraction
Pratima Purushottam Gumaste*
and Vinayak K. BairagiDepartment of Electronics and Telecommunication, AISSM’s Institute of Information Technology, Savitribai Phule Pune University, Pune, India.
Corresponding Author E-mail : pratimagumaste@gmail.com
Abstract: Brain tumors vary in their position, mass, nature, and consistency of these lesions. Due to the similarities found between brain lesions and normal tissues, many challenges are faced by the researcher in developing algorithms for tumor segmentation. Brain tumor abstraction is thought-provoking job in medical image handing out because brain image and its structure is complicated. Segmentation plays a vibrant role in processing of medical images. Magnetic Resonance Imaging has become a particularly useful medical diagnostic tool for diagnosis of brain and other medical images.The objective of this paper is to develop an algorithm that facilitates the study of feature extraction from the brain right and left hemispheres. The proposed study, also highlight a completely different advanced higher order statistical features extracted from the chosen region of brain slice. The tumor area is extracted from statistical features using Support Vector Machine. The proposed methodology can be used to locate tumor tissues based on a single-spectral structural Magnetic resonance image.
Keywords: Benign Tumor; Gray Level Cooccurrence Matrix; Magnetic Resonance Imaging (MRI); Statistical Features; Textural Features; T2 Weighted Images. Back to TOC