AGE challenge: Angle Closure Glaucoma Evaluation in Anterior Segment Optical Coherence Tomography

Area of Science:

• Ophthalmology
• Medical Imaging
• Artificial Intelligence

Background:

• Angle closure glaucoma (ACG) is a severe form of glaucoma, distinct from open-angle glaucoma, characterized by anatomical abnormalities in the anterior chamber angle (ACA).
• Elevated intraocular pressure due to ACA abnormalities can lead to glaucomatous optic neuropathy, visual impairment, and blindness.
• Anterior Segment Optical Coherence Tomography (AS-OCT) is a key imaging modality for differentiating angle closure from open angles, but a lack of standardized datasets hinders AI development.

Purpose of the Study:

• To address the scarcity of public AS-OCT datasets for glaucoma research.
• To establish an evaluation framework for automated angle closure detection and assessment techniques.
• To benchmark and compare artificial intelligence (AI) methods for scleral spur localization and angle closure classification using AS-OCT images.

Main Methods:

• The Angle closure Glaucoma Evaluation (AGE) challenge was organized, releasing a dataset of 4800 annotated AS-OCT images from 199 patients.
• The challenge included two tasks: scleral spur localization and angle closure classification.
• Eight participating teams submitted their AI models for evaluation on the provided dataset and framework.

Main Results:

• The top-performing method for scleral spur localization achieved an average Euclidean Distance of 10 pixels (10 µm).
• For angle closure classification, several AI algorithms demonstrated satisfactory performance, with two achieving 100% accuracy.
• The challenge facilitated a comprehensive comparison of various AI approaches for AS-OCT image analysis.

Conclusions:

• The AGE challenge successfully provided a benchmark dataset and evaluation framework for AS-OCT-based glaucoma assessment.
• AI techniques show significant potential for advancing automated detection and classification of angle closure glaucoma.
• These advancements can lead to improved diagnostic tools and better patient outcomes in managing ACG.