Image Classification using MRI Images in Brain Tumor Dr. Nandhagopal N.*, Jaichander C.**, Ponniwalavan R.** *Assoicate Professor, Department of ECE, SKP Engineering College, Tiruvannamalai, India **Associate Professor, Department of ECE, SKP Engineering College, Tiruvannamalai, India ***Assistant Professor, Aringer anna college of Engineering & Technology, Palani, India Online published on 15 September, 2016. Abstract This article presents the classification of brain images to sense the stages using unverified classifier and abnormal detection through a spatial Fuzzy clustering algorithm. The Support Vector Machine with radial basis function is used for Stage classification. Due to the structure of the tumor cells, the detection of the brain tumor is a very challenging job. This scheme presents a segmentation technique, Spatial Fuzzy C-Mean clustering algorithm, for segmenting computed tomography images to sense the brain tumor in its early stages. The stages of the brain tumor benign, malignant or normal are classified using these classifiers. The manual analysis of the brain samples requires the intensive trained person in order to prevent diagnostic errors and it is also time consuming and inaccurate. The results of the segmentation are used as a source for the Computer Aided Diagnosis system which is used for the early detection of the brain tumor and it improves the chances of the survival rate for the patient. In this the non sub sampled contourlet transform is used to decompose the image for representing contour edges. The simulated result shows that the Fuzzy based segmentation results are more accurate and reliable than thresholding and clustering methods in all cases. Support Vector Machine with image and data processing techniques was employed to implement an automated Brain Tumor classification. Decision making was performed in two stages: feature extraction using the 16 level wavelet decomposition followed by Dual Tree Complex Wavelet Transform features and the classification using Support Vector Machine (SVM). The presentation of the SVM classifier was evaluated in terms of training, performance and classification accuracies. Support Vector Machine (SVM) gives a rapid and correct classification than other neural networks and it is a gifted tool for classification of the tumors. Top Keywords MRI Brain Tumor, Segmentation, Classification, Spatial Fuzzy c-means Clustering, Dual Tree Complex Wavelet Transform, Support Vector Machine Classifier. Top |