Polyp Segmentation in Colonoscopy Images Using Fully Convolutional\n Network

Abstract

Colorectal cancer is a one of the highest causes of cancer-related death,\nespecially in men. Polyps are one of the main causes of colorectal cancer and\nearly diagnosis of polyps by colonoscopy could result in successful treatment.\nDiagnosis of polyps in colonoscopy videos is a challenging task due to\nvariations in the size and shape of polyps. In this paper we proposed a polyp\nsegmentation method based on convolutional neural network. Performance of the\nmethod is enhanced by two strategies. First, we perform a novel image patch\nselection method in the training phase of the network. Second, in the test\nphase, we perform an effective post processing on the probability map that is\nproduced by the network. Evaluation of the proposed method using the\nCVC-ColonDB database shows that our proposed method achieves more accurate\nresults in comparison with previous colonoscopy video-segmentation methods.\n