Support Vector Machine and Particle Swarm Optimization Based Classification of Ovarian Tumour

Ovarian tumour is the most widely recognized reason for death among gynecological malignancies. There are various sorts of clinical and nonclinical highlights that are utilized to examine and break down the contrasts among kindhearted and dangerous ovarian tumors. Computer Aided Diagnosis (CAD) frameworks of high precision are being created as an underlying test for ovarian tumor order rather than biopsy, which is the present highest quality level indicative test. The system uses the K-means clustering for segmentation and the classification methodology is done by the KSVM (Kernel Support Vector Machine) and PSO (Particle Swarm Optimization) classification approach. This automatic framework consists of four steps: preprocessing, segmentation, feature extraction and feature selection, classification, finally, the parameter values of the KSVM (Kernel Support Vector Machine) classifier are dynamically optimized using the PSO (Particle Swarm Optimization) optimization algorithm. It is a bio-inspired optimization algorithm, and PSO optimizers to get the best out of the classification accuracy. An efficient ovarian tumour segmentation, feature extraction and selection by using PSO and classification is offered in this work by combining different Kernel SVM Classifier which provides accurately identify and classify the ovarian tumor in MR image.