Optic Disc and Cup Segmentation Challenge
Overview
The goal of this challenge is to evaluate new and existing algorithms for automated segmentation of Optic Disc and Optic Cup in monocular retinal Color fundus images.
CAD algorithms for segmenting the optic disc (OD) and cup (OC) have been proposed. Benchmarking these algorithms on a variety of datasets is essential to establish their robustness as well as effectiveness. A successful solution would hold great promise to reduce the workload of the opthalmologist while at the same time reduce the subjectivity in diagnosis
The challenge will be held under the Medimage workshop accompanying ICVGIP 2016.
About the Challenge
Drishti-GS is a dataset meant for validation of segmenting OD, cup and detecting notching. The images in the Drishti-GS dataset have been collected and annotated by Aravind Eye hospital, Madurai, India. This dataset is of a single population as all subjects whose eye images are part of this dataset are Indians. The dataset is divided into two: a training set and a testing set of images. Training images (50) are provided with groundtruths for OD and Cup segmentation and notching information. The testing set has 51 images for which ground truth is not provided currently. For further details of the dataset please refer to the paper and visit the website
How to participate
- Please read the challenge terms and conditions.
- Create an account in the portal to download the training and test images along with the Groundtruth for the training set. The same portal should be used to submit the results on the test set.
- Please register for the challenge by filling in the online registration form . The information must be identical to the one used to create account in the portal in step 2. This registration is a mandatory step for submitting results to the challenge.
- The submissions will be evaluated on the test set by the organizers. The results will be declared during the workshop and subsequently listed in the DRISHTI-GS website.
- Participating teams must submit a one page abstract describing their method and present their work during the workshop.
- Please e-mail any queries/doubts/issues at: medimage@iiit.ac.in with "Challenge" as the subject of the e-mail.
SUbmission Format
The submission must comprise of a zip file containing binary masks of the segmentation results of the 51 binary test images saved in 'png' format in two separate folders named "od" and "cup" with file names identical to the corresponding fundus image. Additionally a one page abstract describing the method in pdf format is mandatory.
A maximum of three uploads for the solution is allowed in which case only the latest version will be evaluated. To submit the code login to the portal (step 2 of "how to participate"), provide submission name again (same as the zip files name explained below without '.zip' extension) and upload the zip file. The zip file must be named as $teamname$_$submission n$ . For eg., if team name is "TeamABC" then second version of the submission will be TeamABC_2.zip
Submissions not adhering to the proper naming policy will not be evaluated.
Publication
This challenge is organized in conjunction with and with the support of ICVGIP 2016.
The participating teams who present their work during the workshop will be invited to contribute to a joint journal paper describing and summarizing the challenge outcome, which will be submitted to a high impact SCI journal in the field.
IMPORTANT DATES
- Registration is opened : 3rd October 2016
- Submission deadline for results : 3rd December 2016
- The challenge workshop takes place : 19th December 2016