Fuzzy rulebase parameter determination for stabilized KH interpolation based detection of colorectal polyps on colonoscopy images

In the case of computer aided diagnosis it is advantageous to apply such computational intelligence methods, that can be related to direct measured data by means easily understandable to medical experts. Fuzzy reasoning, if the rulebase is generated from plausible statistical parameters of the image to be analysed, is easy to understand thus can be easily accepted by the society. In the case of colorectal polyps, which might develop into colorectal cancer, thus the population-wide screening would be advisable, more methods are available, but none of them is accepted as standard and effective method. A method based on simple statistical parameters and entropies of image segments is presented, and the effect of determining the rulebase parameters on the efficiency of finding the polyp segment is studied for stabilized Koczy-Hirota rule interpolation.