Description of the WLPolyp-UCdb
· WLPolyp-UCdb is a database containing images that display colonic
polyps, as well as, normal colonic mucosa. It was generated from 42 different
colonoscopy short videos, of 42 different patients, recorded with a White Light
(WL) video colonoscope, by medical doctors from the Faculty of Medicine of the
University of Coimbra, Portugal.
· Polyp frames: there is a total of 1680 instances of 42
different polyps (40 images per polyp, obtained by sampling each video every 10
frames).
· Normal frames:
there is a total of 1360 images displaying normal colonic mucosa.
· The images are
RGB color images with 576x720 pixels (per color channel).
Description of the folders
· There are 2 folders
named Polyp and Normal.
· The Polyp
folder contains 42 subfolders corresponding to images extracted from the 42
videos: inside each subfolder there are 40 polyp images (40x42=1680 polyp
images).
· The Normal
folder contains the 1360 normal colonic mucosa frames.
Copyright
· Images are
property of the University of Coimbra, Portugal.
· The use of
this database is completely restricted for research and educational purposes.
The use of this database is forbidden for commercial purposes.
Citation and Access to the database
If you use this database for your experiments please include the
following references:
·
Pedro N. Figueiredo, Isabel N. Figueiredo, Luis Pinto, Sunil Kumar, Richard Tsai,
Alexander Mamonov: Polyp
Detection with Computer Aided Diagnosis in White Light Colonoscopy: Comparison
of Three Different Methods, Endoscopy
International Open, 2019; 07: E209-215. (DOI: https://doi.org/10.1055/a-0808-4456 )
· Isabel N. Figueiredo, Mahdi Dodangeh, Luis
Pinto, Pedro N. Figueiredo, Richard Tsai : Fast Colonic
Polyp Detection using a Hamilton-Jacobi Approach to Non-Dominated Sorting, Biomedical
Signal Processing and Control, 61 (2020) 102035 (DOI: https://doi.org/10.1016/j.bspc.2020.102035)
A form with personal information needs to be completed to download the WLPolyp-UCdb database.
Download Form & WLPolyp-UCdb database