Detection of ulcerative colitis severity in colonoscopy video frames

Abstract

Ulcerative colitis (UC) is a chronic inflammatory disease characterized by periods of relapses and remissions affecting more than 500,000 people in the United States. The therapeutic goals of UC are to first induce and then maintain disease remission. However, it is very difficult to evaluate the severity of UC objectively because of non-uniform nature of symptoms associated with UC, and large variations in their patterns. To address this, we objectively measure and classify the severity of UC presented in optical colonoscopy video frames based on the image textures. To extract distinct textures, we are using a hybrid approach in which a new proposed feature based on the accumulation of pixel value differences is combined with an existing feature such as LBP (Local Binary Pattern). The experimental results show the hybrid method can achieve more than 90% overall accuracy.