AI Resilience Report forOphthalmologists, Except Pediatric

Ophthalmology is one of the most AI-active fields in medicine, but the technology is mostly augmenting doctors rather than replacing them. Because eye care depends heavily on imaging, computers have gotten really good at reading retinal scans. In a recent survey, ophthalmologists identified AI as by far the "most transformative trend in ophthalmology," with 78% of respondents citing it — a distant second was AMD/geographic atrophy treatments at 11%, according to Ophthalmology Times ^[1].

Three FDA-cleared systems — LumineticsCore (Digital Diagnostics), EyeArt (Eyenuk), and Aeye-DS (Aeye Health) — now autonomously screen for diabetic retinopathy in primary care offices, as Retinal Physician ^[2] explains, and Scotland's deep-learning triage system cut manual grading workload to about 44% while keeping 96.6% sensitivity. A German expert told Ophthalmology Times Europe ^[3] that the greatest impact is coming from AI applied to retinal imaging for diabetic retinopathy and other retinal diseases, where algorithms are approaching or matching expert performance, while ambient scribes and automated documentation tools are beginning to reduce administrative burden. A narrative review in Annals of Eye Science ^[4] similarly finds that AI systems have demonstrated expert-level performance in detecting diabetic retinopathy, AMD, and glaucoma, with expanding applications in ROP, cataract, and refractive error.

Surgery itself — the highest-skill work — remains firmly human, with AI providing decision support and precision guidance, not replacement.

Adoption is speeding up but unevenly. On the "fast" side, a looming workforce shortage is pushing practices to embrace automation: Review of Ophthalmology ^[5] reports that advanced imaging, AI-based screening, and remote monitoring could offset some workforce shortages, though they require infrastructure, training, and oversight. Documentation tools are spreading quickly too — a commentary in Nature's Eye journal ^[6] frames AI scribes as a "new era of efficiency" for ophthalmology.

Major professional bodies are also lending legitimacy: Eyes On Eyecare ^[7] reports that the American Academy of Ophthalmology's Preferred Practice Pattern and clinical guidelines will integrate directly into an AI search engine, with global access for MDs for free. On the "slow" side, real barriers remain — translating strong algorithmic performance into real clinical value without adding cognitive or time burden, plus issues around generalisability, explainability, and trust when models are trained on curated datasets that may not reflect real-world populations, along with dataset bias, workflow integration barriers, reimbursement constraints, regulatory uncertainty, and risks of technology-driven health inequities. The takeaway: if you're drawn to eye care, AI is a powerful teammate — not a career-ender.

Surgery, empathy, and complex judgment still belong to humans.