A systematic review of ulcer detection methods in wireless capsule endoscopy

Abstract

Background: Ulcers are one of the most prevalent disorders in the gastrointestinal (GI) tract, affecting many people worldwide. Wireless capsule endoscopy (WCE) emerges as the most non-invasive way to diagnose ulcers in the GI tract. However, manually reviewing images captured by WCE is a tedious and time-consuming process. Implementing a computer-aided ulcer detection system can facilitate the automatic evaluation of these images. Methods: Many researchers have proposed various models to develop automatic ulcer detection methods. This research aims to conduct a systematic review by scouring four repositories (Scopus, PubMed, IEEE Xplore, and ScienceDirect) for all original publications on computer-aided ulcer detection published between 2011 and 2024. The review follows the the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guidelines. Results: The full texts of 89 scientific articles were reviewed. The contributions of this paper are two-fold: I) it reports and summarizes the current state-of-the-art ulcer detection algorithms; and II) it finds the most appropriate and preferable method in terms of color space, region of interest selection, feature extraction, and classifier. Conclusion: The paper concludes with a discussion of the challenges and futuredirections for ulcer detection.