Brain Tumor MRI Image Segmentation using Deep Learning

Abstract

Throughout the human history, Brain tumor is a well-known deadly disease. Curing of Brain tumor is entirely dependent on early detection of the disease. If a disease can be detected at its earlier stage, it’s mortality rate may come south. Due to advancement of various scientific methods, early detection becomes possible.  Even though various sophisticated methods are available, early detection suffers a setback due to their structural similarity and sizes. In this article we proposed an automatic early brain tumor detection method. Our proposed approach depends on Extreme learning Machine (ELM). For our experiments, we used MRI image data set. In this work, three models are used to enhance the classifier performance. These models are ELM, CNN-PCA and CLACHE. The backbone of our proposed method is CNN. The CNN architecture leverages both 2D and 3D feature to enhance the exactness of the segmentation. The experimental outcomes of our projected model produce high accuracy in terms of segmentation when compared with its counterparts.