Related Experiment Video
Updated: Sep 13, 2025
Using Retinal Imaging to Study Dementia
Published on: November 6, 2017
Diagnostic Accuracy of AI Models in Detecting Different Inherited Retinal Diseases: A Systematic Review and
Sadra Ashrafi1, Hamid Ahmadieh1, Kia Bayat1
1Ophthalmic Research Center, Research Institute for Ophthalmology and Vision Science, Shahid Beheshti University of Medical Sciences, Tehran, Iran.
Purpose:
The purpose of this study was to evaluate the diagnostic accuracy of artificial intelligence (AI) models in detecting different types of inherited retinal diseases (IRDs), including retinitis pigmentosa (RP), Stargardt disease, and familial exudative vitreoretinopathy (FEVR).
Methods:
Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, a systematic search of databases, such as ScienceDirect, PubMed, Scopus, WoS, and IEEE, was conducted. Studies were included in meta-analysis if data on true positive, false positive, true negative, and false negative could be extracted for AI models, diagnosing IRDs. Data extraction included study characteristics, diagnostic accuracy metrics, and imaging modalities, with QUADAS-2 used for risk of bias assessment. Pooled sensitivity and specificity were calculated using random-effects models, and a summary receiver operating characteristic curve was generated.
Results:
Overall, 5412 articles were identified through the database search. After screening, 22 studies were included, with 21 analyzed quantitatively. The meta-analysis showed a pooled sensitivity of 94% and specificity of 99% for RP, 96% sensitivity and 99% specificity for Stargardt disease, and 85% sensitivity and 99% specificity for FEVR. The diagnostic odds ratio was 2486 for RP and 2236 for Stargardt disease. Subgroup analyses indicated that wide-field imaging had higher specificity for RP compared to standard fundus photography.
Conclusions:
AI models demonstrate high diagnostic accuracy for IRDs, particularly for RP and Stargardt disease. These findings underscore the potential of AI as an adjunct diagnostic tool, although real-world validation is needed for broader applicability across diverse clinical settings.
Translational Relevance:
These findings support the integration of AI into ophthalmic care to improve early detection and patient management.
