ABSTRACT

Magnetic resonance imaging (MRI) is a low-risk, non-invasive imaging technique without ionizing radiation hazard, providing high quality/high contrast images and functional images of anatomical structures and organs. In this study, a method to classify the MRI images of the brain related to stroke is presented. Stroke is a sudden development of neurological damage. In this paper, the MRI images for stroke classification use Gabor filters and Histograms to extract features from the images. In the proposed feature extraction method, the features from Gabor filter and histogram features are fused. The extracted features are classified using Support Vector Machine (SVM) with various kernels. Experimental results are shown that the presented method achieves satisfactory classification accuracy