Current Trends On Deep Learning Models Based On Brain Tumors

Abstract

Out of many in Medical Science, a meticulous disease is BRAIN TUMOR. A Brain Tumor is an abnormal growth of cells inside the brain. As a result, tumors started growing from the brain tissue itself. As the analysis goes on with the data provided by health, it has been recognized that DEEP LEARNING has a major application in the fields of brain tumors. Deep Learning has a significant role in the area of computer vision; its application helps in the recognition of brain tumors to get a highly accurate vision as minute errors may cause disaster to our lives. So, for these reasons, BRAIN TUMOR SEGMENTATION has become a vital challenge for medical purposes. Earlier, several methods have been tried for this illness, but they all lack accuracy. But now, Deep Learning presents a solution to this problem through the concept of Brain Tumor Segmentation. Upon this work, studies were undertaken in different areas of brain MR images, and applied different networks for segmentation. The Results in using these networks of segmentation for MR images can be analyzed by comparing these results with a single network. At the first stage, nonlocal devices and methods have been applied to avoid noises. On the second stage, MR images (cranial) were classified. And at last, the tumors will be segmented. As per studies, the accuracy of cranial MR images is around 97%. Experiments prove that this method is efficient enough to use in computer aided brain tumor detection.