Image Mining Methodology for Detection of Brain Tumor: A Review

Abstract

A human brain contains number of tissues that relate to achieving proper functioning of brain. Meanwhile, any abnormal growth in these tissues may change the functioning and this is generally referred as brain tumor. Brain tumor is mainly of two types low grade or benign (Grade 1 and Grade 2) and high grade or malignant (Grade 3 and Grade 4). Brain tumor can be detected with MRI images by applying image processing steps and some machine learning algorithms. Brain MRI images undergo processing by using different techniques such as image enhancement, clustering and classification for detecting the level of brain tumor. The study shows that the filtering operations, edge detection algorithms, morphological operations and clustering are some of the important steps employed for detecting the various levels of brain tumor. This paper mainly focuses on preparing the comparison review on the basis of the referenced proposed methodology, feature extraction and classification methods with its results, future scope along with the advantages and disadvantages of the research done by different professionals and compiling it into one paper. This helps to provide scope for future research directions in brain tumor classification.