Abstract
Automatic tumor segmentation can only be as successful as the feature extraction techniques it relies on. While many such techniques have been employed, it is still not quite clear which of feature extraction methods should be preferred. To help improve the situation, we present here the results of a study in which we compare the efficiency of using Gabor-wavelet features and statistical features, which are two main groups of competent and successful texture-based features in tumor segmentation. To be more specific, we experiment with three different segmentation techniques that employ Support Vector Machines (SVM), K-Nearest Neighbor classifiers (KNN), and the K-Means classifiers. The system that serves as our test-bed includes tumor slice detection, feature extraction, feature selection, and finally feature classification and comparison. The method implementation and the results are discussed. [PUBLICATION ABSTRACT]