An Efficient Approach for Brain Tumor Detection Using Deep Learning Techniques

Abstract

In today’s biomedical scenario, medical automation is improvising and is in demand on an extensive scale. Detecting brain tumor also needs automation for further treatment. There are many brain imaging methods like X-Ray, CT scans, MRI, EEG, and PET. The main aim of these medical imaging techniques is to extract all meaningful and precise information from images with the least error possible. MRI is a non-invasive technique that is currently used for brain analysis, where an experienced doctor has to manually and precisely identify brain tumors in the MRI scans. This paper presents computer-aided deep learning approaches to detect and segment the brain tumor from the MRI images taken from the standard dataset. The proposed work focuses on three main steps like preprocessing of the images, classifying data into the tumor and non-tumorous images, and segmenting the brain tumor region. Convolutional Neural Network (CNN) is a deep learning technique that is mainly used in image classification. In this work, a VGG16 CNN architecture is utilized to classify brain images. Further, computer vision is used to semantically segment the tumorous areas in the MRI scans. The weights of the trained CNN VGG-16 model are used to build bounding boxes and to segment the tumor regions of the MRI scans using the Mask R-CNN model. Mask R-CNN model is a widely used computer vision algorithm that segments and highlights the tumor region in the MRI scan. Thus, using CNN and computer vision techniques, the proposed approach can able to achieve a highly precise medical automation system that can detect and identify brain tumors with improved precision. KeywordsBrain tumorMRI scansConvolution neural networkMask R-CNN