Potential early-stage cancer identification technique during endoscopic examination

Abstract

In its early stages, cancer does not appear very different from the surrounding areas. Inexperienced physicians do not always accurately locate all cancer areas suitable since there are similarly colored and outlined organ surfaces. However, potential early-stage cancer has certain surface symptoms in terms of the color and the distribution of blood tubes. The main purpose of this work is to provide information concerning the identification of early-stage cancer during endoscopic examination. The proposed algorithm has been developed thorough image processing techniques, thereby incorporating and adjusting the statistical value to determine correct lesion recognition. There is 80% in accuracy to allocate the suspect lesion location from the three video experiments.