A Novel Shape Feature Descriptor for the Classification of Polyps in HD Colonoscopy

This work proposes a new method analyzing the shape of connected components (blobs) from segmented images for the classification of colonic polyps. The segmentation algorithm is a novel variation of the fast level lines transform and the resultant blobs are ideal to model the pit pattern structure of the mucosa. The shape of the blobs is described by a mixture of new features (convex hull, skeletonization and perimeter) as well as already proven features (contrast feature). We show that shape features of blobs extracted by segmenting an image are particularly suitable for mucosal texture classification and outperforming commonly used feature extraction methods. Additionally this work compares and analyzes the influences of image enhancement technologies to the automated classification of the colonic mucosa. In particular, we compare the conventional chromoendoscopy with the computed virtual chromoendoscopy (the i-Scan technology of Pentax). Results imply that computed virtual chromoendoscopy facilitates the discrimination between healthy and abnormal mucosa, whereas conventional chromoendoscopy rather complicates the discrimination.