Medical Image segmentation and classification plays an important role in the characterization of tumor

Medical Image segmentation and classification plays an important role in the characterization of tumor. The accuracy in recognizing it is highly essential since the treatment planning is based on its identification. A new hybrid technique based on modified Fuzzy C-Means segmentation algorithm and classification based on support vector machine is proposed. The proposed technique consists of four stages mainly segmentation, feature extraction, feature selection and classification. Modified Fuzzy C-means (FCM) algorithm which has an improved computation rate, modified cluster center and updating membership value criterion is used. Several statistical features are extracted to yield a better performance for classification techniques. Feature extraction stage extracts a set of 14 features using GLCM (Gray Level Co-occurrence Matrix). To select the discriminative features among them, Sequential Forward Selection (SFS) algorithm is used. Support Vector Machine (SVM) classifier is used to classify the MR brain images into benign or malignant. Receiver Operating Characteristic (ROC) curve analysis is done for calculating the misclassification rate. The experiment result of proposed system achieves high classification accuracy whose effectiveness is measured in terms of sensitivity and specificity.