Other Open Access
Huazhu Fu; Yanwu Xu; José Ignacio Orlando; Hrvoje Bogunovi; Fei Li; Xiulan Zhang
This is the challenge design document for the "2nd Retinal Fundus Glaucoma Challenge 2020", accepted for MICCAI 2020.
Glaucoma is currently the leading reason of irreversible blindness in the world. This challenge brings together the medical image analysis community to develop novel automated classification and segmentation methods and to evaluate and compare them for glaucoma detection and optic disc/cup segmentation on a large dataset of retinal fundus images. 1st REFUGE was organized successfully in MICCAI 2018, where 28 teams registered out of which 11 submitted the test set results and participated in. In this year, our REFUGE-2 challenge will utilize a larger dataset (2000 images with glaucoma labels and disc/cup annotations) with more camera modalities (four fundus cameras), which are collected from multiple medical centers. The additional 800 images are captured using two new camera modalities. The REFUGE-2 challenge will consist of three tasks: 1) Classification of clinical Glaucoma; 2) Segmentation of Optic Disc and Cup; 3) Localization of Fovea, which gives attention to the clinical application and technical research.