An Overview of Segmentation and Classification Techniques: A Survey of Brain Tumour-Related Research

Abstract

A brain tumour is a condition in which brain cells develop abnormally. Early treatment can prevent complications, which can be fatal if no early diagnosis or treatment is obtained. The brain tumour has multiple stages and varies in shape and size, necessitating a different type of treatment. It is typically identified through general symptoms such as headache and nausea. To confirm the presence of a potential brain tumour, medical imaging modalities such as Magnetic Resonance Imaging (MRI), Computed Axial Tomography (CAT), and Positron Emission Tomography (PET) are used. Despite the use of high-precision machines to capture brain images, radiologists and specialists are still susceptible to misreading or misinterpretation of medical images due to the large volume of data. Furthermore, manual delineation of brain tumours is tedious and time-consuming. To address these issues, brain tumour segmentation and classification image processing have been extensively researched in recent years due to the increased demand for medical image diagnosis. To address these issues, several new and improved techniques were introduced. Therefore, this paper proposed to overview the algorithms and methods of image enhancement, skull stripping, texture feature extraction, segmentation, and classification involved in an automated MRI brain tumour segmentation and classification. It is expected that this paper would provide new research directions for MRI brain tumour segmentation and classification.