Two-level Combined Classification Technique using Ranklet Transformation for the Detection of MRI Brain Tumor

These days, medical sector plays significant role as people have become more aware towards their health issues. However, it is observed that by and large the medical analyses towards diagnosis of disease are accomplished by medical experts, which is not only a very time-consuming process but also involves subjectivity. Thus, a methodology has been proposed for the detection of the anomalies to overcome the above constraints. In this paper, we have mainly focused on brain tumor diagnosis using MRI modality. Initially, Expectation Maximization Algorithm is used for segmentation. Thereafter, for feature extraction we have implemented Ranklet Transformation. Finally, combined classification technique has been implemented in such a way that Auto-Encoder classifier is followed by Binary SVM classifier. In this paper, we have also compared our results with traditional SVM, and with the accuracy rate of 94.6% it is concluded that the performance of our proposed model is effective and robust.