Artificial intelligence for geographic atrophy: pearls and pitfalls

Area of Science:

• Ophthalmology
• Medical Imaging
• Artificial Intelligence

Background:

• Geographic atrophy (GA) is a vision-impairing, late-stage manifestation of age-related macular degeneration (AMD).
• Accurate monitoring of GA progression is crucial for evaluating new therapies and managing patient care.

Purpose of the Study:

• To review recent advancements in artificial intelligence (AI) for the clinical management of geographic atrophy (GA).
• To highlight AI's role in screening, early diagnosis, and progression measurement of GA.

Main Methods:

• Review of recent literature on AI applications in GA management.
• Analysis of AI systems for segmenting GA lesions on multimodal retinal images (CFP, FAF, OCT).
• Evaluation of AI's utility in measuring GA progression for newly approved therapies.

Main Results:

• AI systems show substantial advancements in segmenting GA lesions using multimodal retinal imaging.
• Optical coherence tomography (OCT) provides high-resolution 3D data optimal for AI algorithm development.
• AI methods are becoming indispensable for patient management and measuring progression in GA therapy.

Conclusions:

• AI offers innovative solutions for GA screening, early diagnosis, and patient management.
• Real-world implementation of AI in GA clinical practice is beginning.
• The review discusses advances and challenges in AI methods for GA clinical implementation.