ORIGINAL PAPER Colonic Polyp Detection in CT Colonography with Fuzzy Rule Based 3D Template Matching Niyazi Kilic & Osman N. Ucan & Onur Osman Received: 8 January 2008 / Accepted: 11 April 2008 / Published online: 22 June 2008 # Springer Science + Business Media, LLC 2008 Abstract In this paper, we introduced a computer aided detection (CAD) system to facilitate colonic polyp detection in computer tomography (CT) data using cellular neural network, genetic algorithm and three dimensional (3D) template matching with fuzzy rule based tresholding. The CAD system extracts colon region from CT images using cellular neural network (CNN) having A, B and I templates that are optimized by genetic algorithm in order to improve the segmentation performance. Then, the system performs a 3D template matching within four layers with three different cell of 8×8, 12×12 and 20×20 to detect polyps. The CAD system is evaluated with 1043 CT colonography images from 16 patients containing 15 marked polyps. All colon regions are segmented properly. The overall sensitivity of proposed CAD system is 100% with the level of 0.53 false positives (FPs) per slice and 11.75 FPs per patient for the 8×8 cell template. For the 12×12 cell templates, detection sensitivity is 100% at 0.494 FPs per slice and 8.75 FPs per patient and for the 20×20 cell templates, detection sensitivity is 86.66% with the level of 0.452 FPs per slice and 6.25 FPs per patient. Keywords CT colonography . Colon segmentation . Genetic algorithm . Cellular neural networks . Fuzzy logic . Colonic polyp detection Introduction Colon cancer is still an important health problem that causes serious morbidity and mortality. It is the second leading cause of cancer death in the developed nations [1, 2]. Today, available detection techniques include fecal occult blood test, barium enema, sigmoidoscopy and colono- scopy. Virtual colonoscopy (VC) or Computed tomography colonography (CTC) is an emerging method for polyp detection through the entire colon. This method is gaining increasing attention as a potential diagnostic test for colorectal cancer [3, 4]. As a screening modality, virtual colonoscopy has another advantage of making use of computer-aided detection (CAD) techniques to examine the internal tissue image textures beyond the inner surface of the colon. Computer-aided detection is attractive because it has the potential to reduce radiologists’ interpretation time, as well as increasing the diagnostic accuracy in the detection of the polyps [5–7]. In the last decade, research has been focused on developing automated computer-aided detection (CAD) methods and several approaches have been proposed. In this regard, Yao et al. [8] proposed segmentation method which is based on the combination of knowledge-guided intensity adjustment, fuzzy c-clustering and deformable models. Yoshida and Nappi’ s[6] CAD scheme extracts the colon in which the polyps are detected. Summers et al. [9] and Vining et al. [10] have proposed CAD schemes that extract the surface of the colonic wall and evaluate the curvature of the surface to detect polyps. Kiss et al. [11] combined the surface normal distribution and sphere fitting. Paik et al. [12] developed a new method based on surface normal overlap. Jerebko et al. designed classification scheme for colonic polyps using neural networks and support vector machine committee [7]. Chowdhury et al. J Med Syst (2009) 33:9–18 DOI 10.1007/s10916-008-9159-3 N. Kilic (*) : O. N. Ucan Engineering Faculty, Electrical and Electronics Engineering Department, Istanbul University, Avcilar, Istanbul, Turkey e-mail: niyazik@istanbul.edu.tr O. Osman Ragip Gumuspala Cad. No:84, Istanbul Commerce University, 34378 Eminonu, Istanbul, Turkey