TDACVO: EXPONENTIAL TASMANIAN DEVIL ANTI-CORONAVIRUS OPTIMIZATION-ASSISTED DEEP MODEL FOR MULTI-LEVEL BRAIN TUMOR CATEGORIZATION

Abstract

Tumors in the brain represent a general and serious disease that leads to small life expectancy. Hence, effective diagnosis preparation is indispensable to enhance the life quality of patients. This paper utilizes MRI for diagnosing the tumor in the brain. However, the MRI thwarts manual classification and poses drawbacks regarding precise quantitative measurements. Thus, automated cataloging of brain tumors is a crucial task. The main aim of this research is to provide an optimized deep model for multi-level classification with the help of MRI. The adaptive Wiener filter and Region Of Interest (ROI) mining are used to perform denoising to make an accurate diagnosis. The first-level classification is accomplished with Shepard Convolutional Neural Network (ShCNN) and its training is done with Tasmanian Devil Anti-Coronavirus Optimization (TDACVO), which categorizes suitably. If an abnormality is found, Deep Q-Network (DQN) and Exponential Tasmanian Devil Anti-Coronavirus Optimization (ETDACVO) are used to execute the second-level classification. The proposed ETDACVO-DQN provided better competence with an utmost accuracy of 93%, precision of 91.8%, and recall of 92.7%.