Authors : R. Meenakshi and P. Anandhakumar
Abstract: Tumor as the definition goes is an abnormal growth of cells which can occur in any part of the human body. Such growth when occurs in brain is called brain tumor. Only symptoms and no causes have been found so far. Brain tumors give very normal symptoms like nausea or headache which may occur due to other reasons also. Therefore, early identification of such tumor is very much necessary. Its threat level depends on a combination of various factors like the type of tumor, its location, size and developmental stage. Many techniques exist for scanning the brain like Computed Tomography (CT) scan, Magnetic Resonance Imaging (MRI), etc. to test the tumor existence. MRI is most common one used for used for analyzing the brain as the images produced are of high precision and applicability. The main objective of this study is to classify the brain MRI dataset for the existence or non existence of tumors. The proposed method uses convolution of orthogonal operators with edge detection operators which is applied on the image. Classification of the image is done using ARTMAP of mirror neurons. The classification accuracy is 90% for the proposed method which is better when compared to BPN based classification.
R. Meenakshi and P. Anandhakumar, 2014. Classification of Brain Tumor Images using Orthogonal Based Composite Operators and Artmap of Mirror Neurons. International Journal of Soft Computing, 9: 189-193.