Evaluating the Effectiveness of the Convolution Neural Network in Detecting Brain Tumors

Abstract

Brain tumors are considered one of the most fatal diseases. Due to the uncontrolled replications of cells in the brain, a tumor forms and adversely affects the host. Hence, early detection of a brain tumor is critical for the planning of treatment and for the patient's survival. And, considering the fact that brain tumors come in a variety of shapes, sizes, and types, manual identification of brain tumors is difficult, time-consuming, and error-prone. The field of artificial intelligence, especially its subfield of machine learning and deep learning, has played a very positive role in transforming many sectors of business, many areas of study, and different areas of technology. One of the fields where deep leaning shows great impact is the field of medical imagery. Medical imagery involves the analysis of medical images. Visual tools such as X-rays, sonography, microscopes, and cameras all aid in the accurate diagnosis of illnesses. Instead of manual analysis of these medical images, these visual aids have also been used as a source of images for technologies like machine leaning and deep learning, which can and have been applied successfully for early detection of abnormalities. Convolution neural network is one of the important deep learning techniques that has been successfully used for analysis of images. Convolution neural network is being actively explored for its effectiveness and use in the field of medical imagery in general and brain tumor detection in particular. This chapter proposes to evaluate and analyse the effectiveness and accuracy of convolution neural network in detecting brain tumors.