International Journal of Scientific Research and Engineering Development

Glaucoma is a progressive ocular disease and a major cause of irreversible blindness worldwide. Early detection is essential, as vision loss from glaucoma cannot be restored. Clinical diagnosis often relies on analyzing the optic nerve head, particularly the optic disc (OD) and optic cup (OC), and calculating the cup-to-disc ratio (CDR), a key biomarker for disease progression. Manual assessment of retinal fundus images is time-intensive, subjective, and dependent on expert availability, limiting its scalability for largescale screening. This study proposes a deep learning-based automatic glaucoma detection framework using EfficientDet-D0 with an EfficientNet-B0 backbone. The model is designed to localize OD and OC boundaries with high precision, enabling accurate CDR estimation and subsequent glaucoma classification. EfficientDet’s compound scaling ensures a balance between accuracy and computational efficiency, making the system suitable for real-world deployment, including telemedicine and portable diagnostic devices. Evaluations on publicly available fundus datasets demonstrate strong localization accuracy and competitive classification performance, with high sensitivity and specificity. The framework reduces reliance on manual interpretation, offering a scalable, objective, and efficient solution for glaucoma screening. By automating OD and OC detection and CDR computation, this approach contributes to timely diagnosis, improved accessibility, and the global effort to prevent glaucoma-related blindness.

P.Revathy, Dr.R.Jayaprakash,"An Advanced Deep Neural Network Model for Early Glaucoma Screening through Optic Nerve Head Segmentation and Quantitative CDR Analysis" International Journal of Scientific Research and Engineering Development, V9(1): Page(1304-1309) Jan-Feb 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.