Google's ARDA retinal AI tops 1M patients in three countries

Over a million patients have been screened for diabetic retinopathy by Google's deep learning tool across three countries, according to a Comment in Nature Medicine published on 23 September 2026.

The paper's corresponding author is Richa Tiwari of Google, with clinician co-authors from each site, and the piece is framed as a deployment-lessons essay rather than a new accuracy result. "Scaling a successful deep learning tool across three highly distinct settings – in India, Thailand and Australia," they write, "offers cross-cutting insights that may inform the expansion of healthcare artificial intelligence globally."

The three anchor sites are the Aravind Eye Care System in Madurai, Rajavithi Hospital in Bangkok, and Lions Outback Vision at the Lions Eye Institute in Western Australia. Google has licensed the underlying tool, the Automated Retinal Disease Assessment or ARDA, to three commercial operators: Forus Health, AuroLab and Perceptra. The stated target is six million AI-assisted screenings across India and Thailand within ten years.

The public abstract publishes no per-site accuracy figures, no referral counts, and no breakdown of how the million screenings split between the three countries. The full text sits behind the Nature Medicine paywall. Two of the people we follow in our Who's Who directory posted the link.