Automatic Detection and Classification of Brain Tumor using Image Processing Techniques

Abstract

Brain tumor cases are uncommon and sometimes can be fatal. There are four grades of brain tumor. According to the grades and the treatment delivery, the brain tumor can be curable or treatable. The diagnosis of the brain tumor in the early stages leads to the effective treatment for the affected patients. This paper presents the efficient method for automatic detection of brain tumor. Some Matlab Techniques play major role in detection of tumor. Firstly, three types of edge detection methods (Canny, Robert and Prewitt) are used for finding the points and edges of the tumor in MRI Images. Some preprocessing steps like Morphological Reconstruction and Segmentation takes place for constructing the images from small components. Then, ROI Process is applied for highlighting the tumor area. The grey level co occurrence matrix (GLCM) and the PNN neural networks are utilized to classify whether the tumor is benign or malignant or metastasis tumor. It is observed that the experimental results of the proposed method gives better automatic detection of brain tumor with high precision in examination with different methods.