Gabor Filter Based Automated Enhancement of Brain Tumors

Present work introduces 2D Gabor filtering technique for enhancement of brain tumors in presence of obscuring tissue in the surrounding and low illumination. Since Gabor filter can capture the edge information from different orientations, filter response is strong in all directions of edges. Enhanced tumor is completely segmented using fuzzy c-means clustering (FCM) algorithm and subjected to shape-based feature extraction employing the concept of extrema. Feature values obtained are used to train a classifier to determine the class (benign/malignant) of the tumor.