Brain Tumor Segmentation from MRI Images using Deep Learning-based CNN with SVM Classifier
Brain Cancer is one of the most threatening disease today. It caused due to the uncontrolledgrowth of unhealthy cells in the brain that could be either cancerous or non cancerous. In today’s worlda brain tumor are not only a life threatening disease but isalso the prominent reason behind numerous deaths. Magnetic Resonance Imaging (MRI) is mostlyused in brain tumor analysis. In this work, a newdeep learning algorithm that is based on CNN with SVM is presented for efficient and automatic segmentation of brain tumor. The steps involved in the processing include preprocessing of the input image, extracting the important features, performing image classification, and finally segmenting the tumor in the brain. The MRI images are smoothenedand the segmentedusing Watershed segmentation. The importantfeatures are then extracted from it based on the shape of the tumor, its position, and feature surface in the brain. Experimental results of the proposed method showthat 92.59% accuracy in evaluation when compared tothe existingalgorithms.