Automated Brain Tumor Diagnosis and Severity Analysis from Brain MRI

摘要

Analysis of brain MRI is of utmost importance as it reveals the underlying details of the main controlling portion of human body. In this paper, we have proposed a fully automated approach to differentiate abnormal brain images from healthy MRI. The proposed method segments out the tumor region from the abnormal MRI by analyzing the energy profile of the image pixels. After tumor segmentation, the tumor features are analyzed to classify the degree of malignancy. This approach can be applied to segment both high grade and low grade tumors.