Brain Tumor Segmentation Using Convolutional Neural Network in MRI Images Based on Deep Learning

Abstract

The brain is the master organ of the human body, responsible for every function of the organism. Brain tumors are one of several common types of disorders that can affect the brain. Tumors refer to the unchecked expansion of malignant cells in any region of a human body. To provide the best care possible, a prompt and correct diagnosis is essential. To distinguish between various types of brain tumors, MRIs are the most often utilised medical imaging modality. Compared to other types of medical imaging, MRIs offer more appealing visual analysis and greater flexibility. Computer aided diagnosis systems that are powered by artificial intelligence can accurately and thoroughly read and analyse even the most intricate medical data. One of AI's most exciting new subfields, deep learning, is developing answers for problems in many different areas. Convolutional Neural Networks (CNNs) are a type of deep neural network used in deep learning for the purpose of analyzing and understanding visual data. To classify brain abnormalities, the primary goal of this study is to create a convolution neural network and train it from scratch using brain tumors as a case study. The suggested CNN model for brain abnormality classification is constructed with a stack of convolution and pooling layers, an activation function, and batch normalization to carry out feature learning and vi classification. In order to extract these traits, the proposed model.