Features for Classification of Polyps in Colonoscopy

Colonoscopy is the gold standard for detection of colorectal polyps that can progress to cancer. In such an examination physicians search for polyps in endoscopic images. Thereby polyps can be removed. To support experts with a computer-aided diagnosis system, we compare different methods for automatic detection. Comparable to traditional pattern recognition systems, features are initially extracted and a classifier is trained on such data. Afterwards, unknown endoscopic images can be classified with the previously trained classifier. In this contribution we concentrate on the extension of the feature extraction module in the existing system. New detection methods are compared to existing techniques. Several features are tested, such as Graylevel Co-Occurrence Matrices (GLCM), Local Binary Patterns (LBP), and Discrete Wavelet Transform features. Different modifications on those features are applied and evaluated. We extend feature detectors to use color in different color spaces. We also compare different classifiers such as Support Vector Machines (SVM) and k -Nearest Neighbor classifier.