Ensemble based brain tumor classification technique from MRI based on K fold validation approach

Abstract

In this research paper, the brain MRI images were classified by considering the excellence of pretrained models of CNN on a public dataset to classify two classes of tumors. An image processing-based computer-aided diagnosis system is crucial for the precise differentiation of malignancies. This paper presents brain tumor classification methods utilizing deep learning and ensemble deep learning algorithms based on magnetic resonance imaging (MRI). BTM-ET categorizes tumors into the tumor class as well as a no tumor class. Base models utilized to make an ensemble are ResNet50 and Squeezenet. Deep MRI features were obtained with a convolutional neural network. Deep learning classifiers categorize these tumor types based on extracted deep characteristics. The system dataset comprises 2800 MRI images from the BRATS dataset and 30 MRIs from public hospitals. BTM-ET attained a performance enhancement of 97.9%. In today’s world, brain cancer is one of the most dangerous diseases with the highest death rate; detection and classification of brain tumors due to abnormal growth of cells, shapes, orientation, and location is a challengeable task in medical imaging. Magnetic resonance imaging (MRI) is a typical method of medical imaging for brain tumor analysis. Automatic brain tumor classification is a huge step forward in the quest for early disease detection and treatment. Artificial intelligence helps us get things done quicker and better. We proposed employing a collection of deep learning models to automate the process of brain tumor categorization.