Swarm based Optimization Technique for Detection of Brain Tumor in T2-Weighted MRI Images

Tumor detection is one of the most critical tasks from the brain MRI images. Commonly magnetic resonance scanner produces brain images with burst tissues where distinctive and combined sights of the tissues are required. The manual view of such tissues on image is impossible and leads to generate errors. Hence with the help of soft computing techniques, the detection of tumor region can be effectively done which will assist the radiologist extensively without errors. Several soft computing techniques have been proposed to improve the accuracy and reduce the false contour detection in medical images. In this work automatic brain tumor detection from MRI images using nature inspired meta-heuristic optimization technique is proposed. The proposed methodology consists of four stages such as preprocessing, segmentation, feature extraction and classification. In preprocessing, the quality of the image is enhanced with median filter by removing the noise. The particle swarm optimization (PSO) algorithm segments the pre-processed image and several textural and shape features are extracted through gray level co-occurrence matrix (GLCM) technique. Finally, the support vector machine (SVM) classifies the extracted tumor from the brain MRI images. The performance of the proposed automated detection method is evaluated on publically available dataset and real images and the obtained results are compared with existing methods. This method yields good, robust and fast segmentation results.