PT  - JOURNAL ARTICLE
AU  - Fu, Huazhu
AU  - Li, Fei
AU  - Xu, Yanwu
AU  - Liao, Jingan
AU  - Xiong, Jian
AU  - Shen, Jianbing
AU  - Liu, Jiang
AU  - Zhang, Xiulan
AU - for iChallenge-GON study group
TI  - A retrospective comparison of deep learning to manual annotations for optic disc and optic cup segmentation in fundus photos
AID  - 10.1101/2020.05.05.20091660
DP  - 2020 Jan 01
TA  - medRxiv
PG  - 2020.05.05.20091660
4099  - http://medrxiv.org/content/early/2020/05/09/2020.05.05.20091660.short
4100  - http://medrxiv.org/content/early/2020/05/09/2020.05.05.20091660.full
AB  - Purpose Optic disc (OD) and cup (OC) segmentation are fundamental for fundus image analysis. Manual annotation is time consuming, expensive, and highly subjective, while an automated system is invaluable to the medical community. The aim of this study is to develop a deep learning system to segment OD and OC in fundus photos, and evaluate how the algorithm compares against manual annotations.Methods A total of 1200 fundus photos with 120 glaucoma cases were collected. The OD and OC annotations were labeled by seven licensed ophthalmologists, and glaucoma diagnoses were based on comprehensive evaluations of the subject medical records. A deep learning system for OD and OC segmentation was developed. The performances of segmentation and glaucoma discriminating based on the cup-to-disc ratio (CDR) of automated model were compared against the manual annotations.Results The algorithm achieved an OD dice of 0.938 (95% confidence interval (CI), 0.934-0.941), OC dice of 0.801 (95% CI, 0.793-0.809), and CDR mean absolute error (MAE) of 0.077 (95% CI, 0.073-0.082). For glaucoma discriminating based on CDR calculations, the algorithm obtained an area under receiver operator characteristic curve (AUC) of 0.948 (95% CI, 0.920-0.973), with a sensitivity of 0.850 (95% CI, 0.794-0.923) and specificity of 0.853 (95% CI, 0.798-0.918).Conclusions We demonstrated the potential of the deep learning system to assist ophthalmologists in analyzing OD and OC segmentation and discriminating glaucoma from non-glaucoma subjects based on CDR calculations.Translational Relevance We investigate the segmentation of OD and OC by deep learning system compared against the manual annotations.Competing Interest StatementThe authors have declared no competing interest.Funding StatementNo external funding was receivedAuthor DeclarationsAll relevant ethical guidelines have been followed; any necessary IRB and/or ethics committee approvals have been obtained and details of the IRB/oversight body are included in the manuscript.YesAll necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived.YesI understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).YesI have followed all appropriate research reporting guidelines and uploaded the relevant EQUATOR Network research reporting checklist(s) and other pertinent material as supplementary files, if applicable.Yeshttps://refuge.grand-challenge.org https://refuge.grand-challenge.org ODoptic discOCoptic cupCDRcup-to-disc ratioMAEmean absolute errorCIconfidence intervalAUCarea under the receiver operating characteristic curveONHoptic nerve headRNFLretinal nerve fiber layerIOPintraocular pressureVFvisual fieldsSDstandard deviationOCToptical coherence tomographyAS-OCTanterior segment OCT