Overview
- Researchers in Portugal tested an AI image‑analysis system on eye photos from 671 adults aged 55–65 in a peer‑reviewed Lancet Primary Care study.
- The AI detected 78% of glaucoma cases compared with 75% identified by clinicians in the trial cohort.
- The tool ruled out disease in 95% of non‑cases versus 91% for doctors, indicating fewer false positives.
- AI triage referred 66 people for specialist review versus 118 by clinicians, yet both pathways resulted in 40 confirmed diagnoses.
- The results point to fewer unnecessary referrals and potential relief for specialist clinics, pending broader validation and regulatory clearance for routine use.