Early detection of precancerous polyps that impersonate normal colon cells is very important to enhance the mortality of individuals with high risk of colorectal cancer. Hence, additional technologies for early colorectal cancer detection are required to improve the current detection methods.
Virtual colonoscopy is a new detection method that doesn’t require any surgical preparation as opposed to optical colonoscopy. The aim of this paper is to utilize virtual colonoscopy in order to aid in the early detection of colorectal cancer. This will potentially increase the protective effect of early colonoscopy screenings.
In this paper, a fully automated colorectal cancer detection system is proposed. The system utilizes advanced image processing and machine learning techniques to extract polyps from 3D DICOM images. After which, the polyps are analyzed and classified as being either benign or malignant.