Artificial intelligence applications in inherited retinal dystrophies

Area of Science:

• Ophthalmology
• Genetics
• Medical Artificial Intelligence

Background:

• Inherited retinal dystrophies (IRDs) are a leading cause of blindness with significant genetic and phenotypic variability.
• Current molecular diagnostic pathways for IRDs are often lengthy, costly, and resource-limited.
• Artificial intelligence (AI) presents a transformative potential for improving IRD diagnosis and management.

Purpose of the Study:

• To review current literature on AI-driven approaches for diagnosing inherited retinal dystrophies.
• To highlight the potential applications and existing challenges of AI in the field of IRDs.

Main Methods:

• A comprehensive literature review of AI applications in inherited retinal dystrophies was conducted.
• Key AI-driven approaches and their reported outcomes were analyzed.

Main Results:

• AI models demonstrate efficacy in predicting disease-causing variants, differentiating phenotypically similar IRDs, and segmenting retinal layers.
• Potential future applications include genetic counseling, progression prediction, and personalized treatment outcome prediction.
• Significant barriers to clinical adoption include lack of standardization, data variability, imaging inconsistencies, and AI's "black-box" nature.

Conclusions:

• AI holds substantial promise for revolutionizing IRD diagnosis and management.
• Addressing challenges related to data, standardization, and AI transparency is critical for widespread clinical adoption.
• Continued research and interdisciplinary collaboration are essential to harness AI's full potential in IRDs.