AI retinal imaging tool capable of screening up to 100 diseases

Researchers at Canada’s École de technologie supérieure (ÉTS) and the firm Diagnos have created an artificial‑intelligence system that analyses routine retinal photographs to identify roughly one hundred diseases. The range includes ocular conditions as well as cardiovascular, systemic and neuro‑degenerative disorders. The AI was trained on about 500,000 patient images and works with equipment already used by optometrists and ophthalmologists.

The system is designed to assist clinicians rather than replace them. It highlights image regions that support its conclusions, provides a confidence level for each finding, and explains the reasoning behind its diagnoses. Professor B. Ayed, chief scientist at Diagnos, said, "We want to help clinicians do rapid, population‑scale screening of ocular and systemic diseases to prevent vision loss or other dangerous conditions." Operations chief Yves‑Stéphane Couture added, "We want the AI to be explainable so clinicians can understand and explain its decisions to patients." The technology follows recent FDA approvals for AI‑based diabetic retinopathy screening and aims to expand AI‑assisted detection to a broader set of health issues.