Home / Health / AI Predicts Diseases From Eye Scans
AI Predicts Diseases From Eye Scans
30 Jun
Summary
- AI system Reti-Pioneer analyzes eye photos for disease prediction.
- Technology trained on UK and China data, tested across ethnicities.
- Eye scan analysis takes 31 seconds, showing potential for early detection.
New research reveals an artificial intelligence system, Reti-Pioneer, capable of predicting six common metabolic and hormonal diseases from photographs of the back of the eye. These include type 2 diabetes, high blood pressure, high cholesterol, gout, osteoporosis, and thyroid disease.
The AI was trained on over 100,000 eye images from participants in the UK and China. Subsequent testing in China and Singapore assessed its efficacy across diverse ethnicities and healthcare settings, demonstrating its broad applicability.
This innovative technology leverages the fact that retinal blood vessels and tissue can exhibit early signs of these conditions. The system's ability to analyze even blurry images is a significant advancement, potentially reducing reliance on traditional blood tests.
Reti-Pioneer completes an analysis in just 31 seconds, with over 80% of patients reporting high satisfaction. While accuracy is highest for diabetes and gout, researchers are optimistic about future improvements for other conditions, though further trials and regulatory approval are necessary.
