Brain Tumor Detection based on Magnetic Resonance Imaging Analysis Using Segmentation, Thresholding and Morphological Operations

Abstract

The brain tumor is one of the most dangerous and deadly diseases in the world. Almost 95% of the patient die due to this disease. The main two classes of brain tumors are benign and malignant tumors. The first failure towards their cure is the late detection and diagnosis. Researchers are on their way to find and discover an accessible and easy method for its earlier detection so that its uncontrollable spread can be controlled before a life-threatening situation. MRI (Magnetic Resonance Imaging) is the best technique for image processing and makes it easier to detect a tumor area in the brain. In this paper, we used preprocessing, segmentation, thresholding, feature extraction, and classification to detect brain tumors in MRI images. A GUI (Graphical User Interface) is also designed to load the images, plot the results, and tell the user/doctor whether the MRI is of a healthy subject or a Tumor patient. The total algorithm takes less than 5 seconds to give complete classification results and accurately classifies 98.99% times.