Related Experiment Video
Updated: Jan 19, 2026
09:17
Using Retinal Imaging to Study Dementia
Published on: November 6, 2017
22.2K
Fully automated method for glaucoma screening using robust optic nerve head detection and unsupervised segmentation
Amed Mvoulana1, Rostom Kachouri1, Mohamed Akil1
1Gaspard-Monge Computer Science Laboratory, A3SI, ESIEE Paris, Université Paris-Est, Paris, France.
Summary
This study presents an automated method for glaucoma screening using retinal images and the Cup-to-Disc Ratio (CDR). The efficient algorithm achieves 98% accuracy, aiding early diagnosis and accessibility in remote areas.
Area of Science:
- Ophthalmology
- Medical Imaging
- Computer Science
Background:
- Glaucoma is a leading cause of irreversible vision loss worldwide.
- Early detection through screening is crucial to slow disease progression.
- Automated analysis of retinal images offers a scalable solution for glaucoma screening.
Purpose of the Study:
- To develop a computationally efficient, fully automated method for glaucoma screening and diagnosis from retinal fundus images.
- To enable glaucoma screening in remote locations with limited clinical resources via mobile device implementation.
- To accurately assess optic nerve head (ONH) damage using the Cup-to-Disc Ratio (CDR).
Main Methods:
- A robust optic disc (OD) detection using brightness criteria and template matching.
- Efficient optic cup (OC) and optic disc (OD) segmentation via texture-based and model-based approaches.
- Automated Cup-to-Disc Ratio (CDR) computation for glaucoma classification.
Main Results:
- The method achieved 98% accuracy in classifying healthy versus glaucomatous patients on the DRISHTI-GS1 dataset.
- Outperformed existing state-of-the-art CDR feature-based approaches.
- Demonstrated excellent performance metrics for glaucoma screening and diagnosis.
Conclusions:
- A fully automated, low-computational method for glaucoma screening from retinal images was successfully developed.
- The approach facilitates early glaucoma detection and diagnosis, improving specialist assessments.
- Potential for integration into mobile health systems to enhance widespread visual health programs.
