Accurate Brain Tumor Detection Using Optimized Feature Selection Based Hybrid Algorithm

Over the past 20 years, advances in image processing technologies have made it easier for medical professionals to diagnose cancers from photographs. Medical imaging systems and technologies have included computer-aided diagnosis systems (CADs), which have demonstrated a 10% increase in the rate at which positive cases are detected by doctors. An increasingly popular field of study is brain tumors, specifically glioblastoma multiforme (GBM). An essential stage in the process of analyzing this pathology is tumor segmentation. Because manual segmentation differs amongst observers, it is frequently inconsistent. A solution to this problem has been suggested by automated segmentation Techniques like (CNNs) convolutional neural networks which are machine learning pipelines based on the way synapses (connections) and neurons (called nodes) function biologically, have attracted attention. In order to learn more about CNNs and find an example pipeline for segmentation, we first conduct an educational analysis of CNNs before examining their potential for brain tumor segmentation. Next, we examine a new discipline called radiomics to learn more about how CNNs might be used in the future. To assist physicians in identifying brain cancers from magnetic resonance imaging (MRI) pictures, a hybrid algorithm was developed as part of this study.