A Survey on Identification of Glioblastoma Multiforme and Low-Grade Glioma Brain Tumor Type

Abstract

Brain tumor type identification is a challenging task for radiologists and physicians. The aim of this paper is a survey on the latest techniques of brain tumor type identification using magnetic resonance imaging i.e. MRI. MR image of the brain tumor might show clear tumor but identification of tumor type is more crucial. By identifying the exact type of tumor, doctors can proceed with further treatment. The invasive biopsy medical procedure can identify the exact type of tumor, but it usually leaves a post-surgical impact on patients because it requires the insertion of a medical device into the human brain as well as it is a time-consuming process. This is the motivation for this study in which digital image processing methods and machine learning help further treatment between surgeon and computer. These methods offer other opinions to the doctors and support for the understanding of MR images, thus increasing diagnostic accuracy. This survey aims to analyze the different techniques of brain tumor type identification which includes different phase like pre-processing, segmentation, feature extraction, and classification. Finally, a vital evaluation of the latest technology, as well as the possible future scope of different techniques, are discussed.