Identifying Grades of Glioma using Support Vector Machine Recursive Feature Elimination

The objective of this research is to classify glioma-a type of brain tumor, according to their grades by combining various classificationmethods and conventional magnetic resonance imaging. Determining Gliomas grades falls under the category medical image analysis. Imageanalysis includes the following: ROI definition (extraction), feature selection and classification. Feature selection till date is done using SVMRFEalgorithm. SVM-RFE stands for Support Vector Machine- Recursive Feature Elimination. But this algorithm can only classify glioma gradeII, IV. The extracted feature of grade III is similar to the features of grade II or grade IV. Hence, they are either classified as grade II or grade IV. This paper aims at improving the existing classification method so as to identify grade III as well. Also in the existing systems the ROIextraction is done manually. Hence, the existing systems are semi-automatic. This work also aims at designing a fully automated system