Robust optic disc location via combination of weak detectors

Area of Science:

• Ophthalmology
• Medical Imaging
• Computer Vision

Background:

• Accurate localization of ocular structures is crucial for diagnosing and monitoring various eye diseases.
• Existing automated methods for optic disc detection face challenges with image quality and anatomical variations.

Purpose of the Study:

• To develop and validate a robust, automated approach for precise optic disc localization in retinal images.
• To improve the reliability of optic disc detection for clinical applications.

Main Methods:

• A novel algorithm detects independent candidates for the optic disc, macula, and major blood vessels (arcades).
• Candidates are ranked by reliability, and anatomical knowledge is used to select the optimal triplet (disc-macula-arcades).
• The method was evaluated on the STARE dataset and wide-field SLO images from an OPTOS instrument.

Main Results:

• The automated approach achieved a 91.4% detection rate on the STARE dataset.
• A 100% detection rate was achieved on the 20 OPTOS-acquired wide-field SLO images.
• The method demonstrated robustness across healthy and pathological retinal images.

Conclusions:

• The proposed robust approach offers reliable automatic optic disc localization.
• This technique has the potential to aid in the early detection and management of retinal diseases.
• High accuracy on diverse datasets validates the clinical utility of this method.