Brain Tumor Detection using Convolutional Neural Network (CNN)

Abstract

Begin with image processing for technology to detect brain tumors. I.e. (The identification of tumor/cancer cells from brain images is primarily based on image recognition methods, since these images are complex and human eyes are not ideal for interpreting the transformed cells with several degrees of changes). There are different types of instruments to help diagnose brain tumors, such as MRI scans, CT scans, etc. The device that can detect any organ and brain problem is MRI (Magnetic Resonance Imaging). Segmentation cell multiplication is an important strategy for processing brain tumor images. The segmentation or multiplication of cells will recognize the tumor along with its neighboring compartments and nearby tissues, but it is difficult enough to repair and shape the morphological changes caused by the tumor. Even though there are a number of current works on the subject. Many methods, such as template-based K means algorithm, fuzzy logic algorithms, threshold segmentation, etc., have been used to establish image processing, but the precision of the performance rate is still not up to the mark. In our proposed methodology our main purpose is to get a clearer image form MRI. We would try to use CNN algorithm which is more flexible and convenient. That will detect the position of the tumor automatically. This proposed methodology will be more efficient and faster to identify the tumor region and also it will be more effective and accurate for brain tumor detection and segmentation. Our main focus is on the methods used to identify brain tumors through image segmentation.