Related Experiment Video
Updated: Jun 12, 2026
09:17
Using Retinal Imaging to Study Dementia
Published on: November 6, 2017
Detecting the optic disc boundary in digital fundus images using morphological, edge detection, and feature
Arturo Aquino1, Manuel Emilio Gegundez-Arias, Diego Marin
1Department of Electronic, Computer Science and Automatic Engineering, ”La Rábida” Polytechnic School, University of Huelva, 21071 Huelva, Spain. arturo.aquino@diesia.uhu.es
IEEE Transactions on Medical Imaging
|June 22, 2010
Summary
This study introduces a new method for segmenting the optic disc (OD) in retinal images, crucial for automated ophthalmic disease diagnosis. The approach accurately locates and segments the OD, achieving high overlap with true regions.
Area of Science:
- Ophthalmology
- Medical Imaging
- Computer Vision
Background:
- Automated diagnosis of ophthalmic pathologies relies on accurate optic disc (OD) detection.
- Existing OD segmentation methods require evaluation for clinical applicability.
Purpose of the Study:
- To present a novel template-based methodology for segmenting the optic disc (OD) from digital retinal images.
- To propose a voting-type algorithm for initial OD pixel localization.
Main Methods:
- A template-based segmentation approach utilizing morphological and edge detection techniques.
- Circular Hough Transform for approximating the OD boundary.
- A voting-type algorithm for initial pixel location within the OD.
Main Results:
- The OD location algorithm achieved 99% success rate with an average computation time of 1.67 seconds.
- The OD segmentation algorithm yielded an average common area overlap of 86% with true OD regions.
- Average segmentation computation time was 5.69 seconds.
Conclusions:
- The proposed methodology offers an effective and efficient solution for automated optic disc segmentation in retinal images.
- The study provides a valuable tool for developing automated diagnostic systems for ophthalmic diseases.
- A comparative discussion of OD segmentation models is included.
