Diagnostic accuracy of artificial intelligence-based automated diabetic retinopathy screening in real-world settings: a systematic review and meta-analysis.

Artificial intelligence (AI) has been shown to have similar diagnostic accuracy to humans in mammography and other imaging studies. This review summarizes the accuracy of AI-based diabetic retinopathy (DR) screening from retinal images. Pooled analysis from 34 studies shows AI-based screening has acceptable performance for DR screening.