Automatic ulcer detection scheme using gray scale histogram from wireless capsule endoscopy

Abstract

Wireless capsule endoscopy (WCE) is one of the most recent and effective video technologies to diagnose gastrointestinal (GI) diseases, such as ulcer in GI tract. Because of long duration of WCE videos, it is a burden for the physicians to analyze large number of image frames. This time consuming process often leads to human error. Hence, an automatic scheme for ulcer image detection in WCE video has great demand. In this paper, an automatic scheme based on image histogram is proposed to detect ulcer frames in WCE video. In order to compute the histograms from pixel values, instead of using conventional RGB color plane, gray scale is used, which offers the scope of utilizing brightness of color. Based on extensive experimentation, it is found that histogram patterns, obtained from ulcer and non-ulcer images in gray scale, exhibit significant differences. Cumulative pixel number of the pixel values in gray scale over an optimum threshold is chosen as feature through histogram analysis. Moreover, this 1-D feature offers computational advantage and ease of implementation. For the purpose of classification, the support vector machine (SVM) supervised classifier is used. Also, performance of the proposed method obtained from several other classifiers is compared. The performance of the proposed method is tested on several WCE images taken from publicly available WCE video database and it is found that, the proposed method offers superior classification performance, in comparison to that obtained by some existing methods, in terms of accuracy, specificity, and sensitivity.