Pune study: smartphone AI detects diseases from retinal images offline
A new study shows that a smartphone-based AI can accurately detect major eye diseases by analyzing retinal images of your eyes.
Researchers in Pune tested the Medios AI multi-disease software on 193 adults and found it could pick up conditions like glaucoma, age-related macular degeneration (AMD), and diabetic retinopathy, all while functioning entirely offline.
AI 98.2% glaucoma sensitivity 99.0% specificity
The AI was almost as sharp as trained eye doctors, with glaucoma sensitivity of 98.2% and specificity of 99.0%, and strong results for AMD and diabetic retinopathy too.
Because it works offline, this tech could make eye checkups way more accessible, especially in places where seeing a specialist isn't easy.
It's a promising step toward catching vision problems early in resource-limited settings.