Artificial intelligence in corneal diseases: A narrative review

Area of Science:

• Ophthalmology
• Medical Artificial Intelligence
• Biomedical Engineering

Background:

• Corneal diseases pose a significant global health challenge, particularly in areas with limited access to specialized eye care.
• Artificial intelligence (AI) presents a transformative approach to automate the diagnosis and management of various corneal conditions.

Purpose of the Study:

• To conduct a narrative review on the application of AI in diagnosing and managing key corneal diseases.
• To explore emerging trends and challenges in AI for corneal disease management.

Main Methods:

• Review of AI applications across diverse corneal conditions including keratoconus, infectious keratitis, pterygium, dry eye disease, Fuchs endothelial corneal dystrophy, and corneal transplantation.
• Analysis of AI models integrating multimodal imaging (corneal topography, slit-lamp, OCT) and clinical data.

Main Results:

• AI models demonstrate high diagnostic accuracy, frequently surpassing human expert performance.
• Emerging AI trends include using biomechanical data for keratoconus detection and in vivo confocal microscopy for infectious keratitis diagnosis.
• AI shows promise in predicting disease progression, treatment success, and post-transplant complications.

Conclusions:

• AI offers a powerful tool for enhancing the diagnosis and management of corneal diseases, improving accessibility to eye care.
• Addressing challenges like data heterogeneity, validation, and model interpretability is crucial for widespread AI adoption in ophthalmology.