Haustral fold detection for CT colonography images using Gabor filter

Haustral folds are anatomically meaningful structures in the human colon. In this paper we present a method for haustral fold detection for CT colonography images. The 3D colon surface is first preprocessed into a 2D flattened colon. Then the image processing is done on 2D images. The proposed method makes used of a 2D Gabor filter to extract the feature of haustral folds. Then Sobel operator is performed on the filtering results to acquire the edge of the folds. Finally the folds are identified using thresholding method. Experimental results show that our method has a satisfactory detection rate on haustral folds.