Polyp Segmentation with Fully Convolutional Deep Dilation Neural Network Evaluation Study

Guo, Yun Bo and Matuszewski, Bogdan orcid iconORCID: 0000-0001-7195-2509 (2020) Polyp Segmentation with Fully Convolutional Deep Dilation Neural Network Evaluation Study. In: 23rd Conference on Medical Image Understanding and Analysis (MIUA), 24-26 July 2019, Liverpool, UK..

[thumbnail of Author Submitted Manuscript] PDF (Author Submitted Manuscript) - Submitted Version
Restricted to Repository staff only
Available under License Creative Commons Attribution Non-commercial No Derivatives.

896kB

Official URL: https://doi.org/10.1007/978-3-030-39343-4_32

Abstract

Analyses of polyp images play an important role in an early detection of colorectal cancer. An automated polyp segmentation is seen as one of the methods that could improve the accuracy of the colonoscopic examination. The paper describes evaluation study of a segmentation method developed for the En-doscopic Vision Gastrointestinal Image ANAlysis – (GIANA) polyp segmenta-tion sub-challenges. The proposed polyp segmentation algorithm is based on a fully convolutional network (FCN) model. The paper describes cross-validation results on the training GIANA dataset. Various tests have been evaluated, includ-ing network configuration, effects of data augmentation, and performance of the method as a function of polyp characteristics. The proposed method delivers state-of-the-art results. It secured the first place for the image segmentation tasks at the 2017 GIANA challenge and the second place for the SD images at the 2018 GIANA challenge.


Repository Staff Only: item control page