2020/05/28 by Bajwa, Muhammad Naseer, Wolfgang Neumeier, Singh, Gur Amrit Pal +6 · 7 citations
Computer Science · Medicine · #Computer Vision and Pattern Recognition (cs.CV) #Digital Imaging for Blood Diseases #FOS: Computer and information sciences #FOS: Electrical engineering #Glaucoma and retinal disorders #Image and Video Processing (eess.IV) #Machine Learning (cs.LG) #Retinal Imaging and Analysis #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2006.09158
openalex publication_date 2020/05/28 · openalex created_date 2021/03/15 · openalex updated_date 2026/07/28
Scarcity of large publicly available retinal fundus image datasets for\nautomated glaucoma detection has been the bottleneck for successful application\nof artificial intelligence towards practical Computer-Aided Diagnosis (CAD). A\nfew small datasets that are available for research community usually suffer\nfrom impractical image capturing conditions and stringent inclusion criteria.\nThese shortcomings in already limited choice of existing datasets make it\nchallenging to mature a CAD system so that it can perform in real-world\nenvironment. In this paper we present a large publicly available retinal fundus\nimage dataset for glaucoma classification called G1020. The dataset is curated\nby conforming to standard practices in routine ophthalmology and it is expected\nto serve as standard benchmark dataset for glaucoma detection. This database\nconsists of 1020 high resolution colour fundus images and provides ground truth\nannotations for glaucoma diagnosis, optic disc and optic cup segmentation,\nvertical cup-to-disc ratio, size of neuroretinal rim in inferior, superior,\nnasal and temporal quadrants, and bounding box location for optic disc. We also\nreport baseline results by conducting extensive experiments for automated\nglaucoma diagnosis and segmentation of optic disc and optic cup.\n