<  Back to the Polytechnique Montréal portal

A benchmark for endoluminal scene segmentation of colonoscopy images

David Vazquez, Jorge Bernal, F. Javier Sánchez, Gloria Fernández-Esparrach, Antonio M. López, Adriana Romero, Michał Drożdżal and Aaron Courville

Article (2017)

Published Version
Terms of Use: Creative Commons Attribution.
Download (2MB)
Cite this document: Vazquez, D., Bernal, J., Sánchez, F. J., Fernández-Esparrach, G., López, A. M., Romero, A., ... Courville, A. (2017). A benchmark for endoluminal scene segmentation of colonoscopy images. Journal of Healthcare Engineering, 2017. doi:10.1155/2017/4037190
Show abstract Hide abstract


Colorectal cancer (CRC) is the third cause of cancer death worldwide. Currently, the standard approach to reduce CRC-related mortality is to perform regular screening in search for polyps and colonoscopy is the screening tool of choice. The main limitations of this screening procedure are polyp miss rate and the inability to perform visual assessment of polyp malignancy. These drawbacks can be reduced by designing decision support systems (DSS) aiming to help clinicians in the different stages of the procedure by providing endoluminal scene segmentation. Thus, in this paper, we introduce an extended benchmark of colonoscopy image segmentation, with the hope of establishing a new strong benchmark for colonoscopy image analysis research. The proposed dataset consists of 4 relevant classes to inspect the endoluminal scene, targeting different clinical needs. Together with the dataset and taking advantage of advances in semantic segmentation literature, we provide new baselines by training standard fully convolutional networks (FCNs). We perform a comparative study to show that FCNs significantly outperform, without any further postprocessing, prior results in endoluminal scene segmentation, especially with respect to polyp segmentation and localization.

Open Access document in PolyPublie
Subjects: 1900 Génie biomédical > 1900 Génie biomédical
2600 Robotique > 2603 Vision artificielle
2700 Technologie de l'information > 2700 Technologie de l'information
2800 Intelligence artificielle > 2800 Intelligence artificielle (Vision artificielle, voir 2603)
Department: Département de génie informatique et génie logiciel
Research Center: Non applicable
Funders: Imagia Inc., Spanish government, iVENDIS, SGR Projects, CERCA Programme/ Generalitat de Catalunya, TECNIOspring-FP7-ACCI grant, FSEED, NVIDIA Corporation
Grant number: AC/DC TRA2014-57088-C2-1-R, DPI2015-65286-R, 2014-SGR-1506, 2014-SGR-1470, 2014-SGR-135
Date Deposited: 09 Mar 2020 15:57
Last Modified: 10 Mar 2020 01:20
PolyPublie URL: https://publications.polymtl.ca/3596/
Document issued by the official publisher
Journal Title: Journal of Healthcare Engineering (vol. 2017)
Publisher: Hindawi Publishing Corporation
Official URL: https://doi.org/10.1155/2017/4037190


Total downloads

Downloads per month in the last year

Origin of downloads


Repository Staff Only