The Endoscopic Images Visualization in Clinical Decision Support Systems

摘要

Virtual chromoendoscopy technology allows to increase the diagnostic efficiency by using digital processing methods. The article proposes various algorithms for the virtual chromoendoscopy implementation based on a tone correction as the first step, and with non-linear methods for correcting brightness and contrast as the second step. An open base of Kvasir endoscopic images was used as an experimental base for evaluating the effectiveness of the proposed algorithms. The visual effect provided by the proposed algorithm corresponds to or, in a significant number of cases, exceeds the visual effect of closed virtual endoscopy technologies, such as i-scan and FICE.