Machine Learning for Detection and Classification of Human Brain Tumor: A Survey

• Published in: 2023 International Conference on Information Technology, Applied Mathematics and Statistics (ICITAMS) pp. 122 - 128
• Main Authors: Al Hussen, Sara Ali, Alsaadi, Elham Mohammed Thabit A.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 20-03-2023
• Subjects: brain tumor(BT); classification; Clustering algorithms; Image segmentation; machine learning(ML); Magnetic resonance imaging; magnetic resonance imaging(MRI); Organizations; segmentation; Shape; Solids; Surveys
• DOI: 10.1109/ICITAMS57610.2023.10525497

Intracranial solid tumors arise due to uncontrolled and abnormal cell division. Classification of brain tumors depends on factors such as the tumor's location, the type of tissue it consists of, its malignant or benign nature, and other factors. Magnetic resonance imaging (MRI) is one of the techniques frequently used to detect tumors in the brain Magnetic resonance imaging (MRI). It provides important detail that is used in accurately scanning the human body's internal organization. problem with brain tumors is the complexity and variations in tumor location, shape, size, and intensity from patient to patient. Also, tumor boundaries are usually unclear and irregular. Some studies have appeared in the detection and classification of human brain tumors. Therefore, some of these studies will be discussed in this paper. Humans have developed many algorithms to get better results in brain tumor detection. They used the latest machine-learning techniques to solve this problem. Through these studies, we conclude that the best accuracy was achieved using deep learning techniques like Deep Auto-Encoder (DAE) based on Jaya Optimization Algorithm (JOA) for classification and Bayesian Fuzzy clustering (BFC) to segment up to 98.5%.