Colorectal polyp detection from local features leveraging deep CNN

Abstract

A colorectal polyp is a small cell cluster that develops on the lining of the colon. The majority of colon polyps are harmless, but some can develop into tumors later on and can cause colorectal cancer. Though the exact causes of polyp formation are not known, some factors contributing to polyp formation are smoking, excess alcohol consumption, obesity, lack of exercise, consumption of processed food, or having a family history of colorectal polyps. However, if the polyp is detected at an early stage, then it is possible to prevent the tumor or colorectal cancer. In this study, our aim will be to train a CNN model and evaluate its performance. Analysis of medical images using deep learning has been confirmed to be quite successful. CNN showed notably higher accuracy in the domain of medical image classification than traditional machine learning algorithms. CNN is a cutting-edge method of deep learning that is widely used in image or object recognition and classification.