Computer analysis of upper gastrointestinal endoscope images

Abstract

Gastro-oesophageal reflux disease (GORD) occurs in up to 40% of adults in the West. Oesophagitis is a major determinant in the treatment of GORD but its current classification systems are subjective. In order to help to provide objective interpretation of upper gastro-intestinal (GI) endoscope examination and reduce inter-observer variability, we developed a computer image analysis system. Digital video recordings were made on patients with clinical evidence of reflux oesophagitis. Cross-sectional profiles of hue and saturation data were analysed on images from seven patients with grade B or C oesophagitis (LA grading). This analysis showed clear changes in hue (p = 0.01) and saturation (p = 0.001). These results suggest that quantification of upper GI endoscopic images is feasible and may help in objective assessment.