Comparative Study of Various Unsupervised learning algorithms for Brain Tumor Detection

Abstract

The Brain Tumor is frightening numerous individuals worldwide. A tumor could be a mass of abnormal cells developed within the brain. Any growth inside skill which is very restricted area can cause problems. Brain tumors of two types cancerous (malignant) or noncancerous (benign). Magnetic Resonance Imaging (MRI) is the technique used in hospitals to examine patients and fix the severity of certain damages. It produces high-quality images. The brain tumor detection in early stage is important factor from patient treatment point of view. The developments in machine learning (ML) and image processing (IP) algorithms have made it possible. The brain tumor identification using image processing is one kind of exciting chore. It is also a critical clinical diagnosis problem. It is a difficult task because of complexity and variation in tumor tissue characteristics like its shape, size, gray level intensities, and location. It needs to be precise, strong, and efficiently irrespective of impacts caused by various parameters. Considering all these real-time challenges, this research is intensive in the direction of highlighting the strong points and boundaries of unsupervised learning algorithms discussed in the contemporary works. Also, the paper also provides a critical assessment of the surveyed literature.