Related Experiment Videos
Artificial Intelligence in Inherited Retinal Diseases: Imaging-Based Applications and Emerging Trends
Eirini Maliagkani1, Anna Gkoritsa1, Konstantinos Tyrlis1
11st Department of Ophthalmology, General Hospital of Athens "G. Gennimatas", National and Kapodistrian University of Athens, Athens, Greece.
Seminars in Ophthalmology
|July 22, 2026
Summary
Artificial intelligence (AI) shows promise in analyzing retinal images for inherited retinal diseases (IRDs). While AI excels at tasks like disease detection and classification, further research is needed to overcome limitations for clinical use.
Area of Science:
- Ophthalmology
- Medical Imaging
- Artificial Intelligence
Background:
- Inherited retinal diseases (IRDs) present complex diagnostic and management challenges due to their genetic and phenotypic heterogeneity.
- Current retinal imaging advances improve assessment, but early diagnosis, monitoring, and prognosis remain difficult.
- Artificial intelligence (AI), particularly deep learning (DL), offers potential solutions for these challenges in ophthalmology.
Purpose of the Study:
- To review recent studies on AI applications in retinal imaging for IRDs.
- To summarize AI's role in disease detection, segmentation, classification, and prognostic assessment.
- To identify current limitations and future directions for AI in IRD management.
Main Methods:
- Narrative review of peer-reviewed studies from 2019-2025.
- Focused on AI applications in optical coherence tomography (OCT), fundus autofluorescence (FAF), and color fundus photography for IRDs.
- Examined AI for automated image analysis tasks including detection, segmentation, classification, and prognosis.
Main Results:
- AI models demonstrate high performance in automated image analysis for IRDs like retinitis pigmentosa and Stargardt disease.
- Deep learning methods show strong capabilities in retinal layer/lesion segmentation, classification, and structure-function analysis.
- Emerging AI applications in prognostic modeling and identifying functional retinal regions are less explored, with limitations including small datasets and lack of validation.
Conclusions:
- AI holds significant potential for improving IRD diagnosis and management through automated image analysis and biomarker assessment.
- Current AI applications are limited by small datasets, insufficient external validation, and lack of longitudinal studies.
- Future research requires collaborative, multicenter efforts, multimodal data integration, and interpretable AI for broader clinical adoption.