Grey wolf optimization technique with U-shaped and capsule networks-A novel framework for glaucoma diagnosis

Area of Science:

• Ophthalmology
• Artificial Intelligence
• Medical Imaging

Background:

• Glaucoma is a leading cause of irreversible blindness globally.
• Current diagnostic methods are often time-consuming and require specialized expertise.
• Early detection is crucial to prevent significant vision loss.

Purpose of the Study:

• To develop an automated glaucoma screening platform.
• To enhance the accuracy and efficiency of glaucoma diagnosis.
• To overcome limitations of manual screening processes.

Main Methods:

• A hybrid segmentation technique using Grey Wolf Optimization Algorithm (GWOA) and U-Net++ for optic disc extraction.
• A deep learning classification model, CapsNet, for robust glaucoma detection.
• Integration of segmentation and classification for a comprehensive diagnostic system.

Main Results:

• Achieved 96.01% accuracy in combined segmentation and classification.
• Demonstrated superior performance compared to traditional glaucoma screening methods.
• Highlighted the system's capability for early-stage glaucoma detection.

Conclusions:

• The automated system significantly improves clinical accuracy in glaucoma diagnosis.
• This AI-driven approach offers a more efficient and accessible screening solution.
• The developed framework shows strong potential for reducing global glaucoma-induced blindness.