ESCRS 2026: Why AI is already reshaping glaucoma care and patient conversations

Artificial intelligence (AI) is outperforming even expert clinicians in glaucoma detection, and ophthalmologists should embrace it rather than fear it, according to Ana Miguel, MD, PhD, FEBO, FEBO-G, GCSRT, speaking with Ophthalmology Times Europe at the 44th Congress of the ESCRS in London.

AI outperforms experts, but doesn't replace them

Miguel's group published a study in The Lancet comparing AI with a group of European glaucoma experts; AI, as expected, outperformed the clinicians. “The message is: don't be afraid of artificial intelligence. It's here to help us, but you still need to use it cleverly,” she said. In the same study, she noted, AI's advantage over experts helped the wider patient population by decreasing the amount of time experts needed to spend on screening—allowing more patients to be screened overall. Ophthalmologists, she said, need to become very familiar with AI while still relying on their own clinical judgement, or “natural intelligence.”

Expanding access through telemedicine

Miguel also uses AI extensively in telemedicine. In parts of France, where she practises, healthcare access is limited; teams of secretaries and orthoptists see patients, taking retinal photos and measuring refraction and intraocular pressure with an air-puff tonometer. An algorithm then assesses whether a patient needs a full consultation. “I still review every single retinal photo myself, but it saves me some time,” she said.

Barriers to wider adoption

Despite its promise, Miguel identified real obstacles to AI adoption. Fear of the technology remains common, particularly among older, more traditional clinicians, she said. AI tools also require rigorous validation and regulatory oversight, because poorly trained systems will produce wrong conclusions. “It can be very helpful, but in my opinion we still need human validation, as a second check,” she said.

Patients are already asking AI

Patients—especially younger ones—are increasingly turning to tools such as ChatGPT with questions about their symptoms, Miguel said. She noted that ChatGPT has adjusted its guidance to recommend seeing a physician rather than offering a diagnosis outright, a shift she called “both good and bad”: patients become more engaged in their own care, but the same trend can also affect their confidence in their ophthalmologist. That dynamic, she said, underscores the importance of the doctor-patient relationship, which shapes how open patients are about their symptoms and their adherence to treatments such as eye drops. “AI is absolutely here to stay, and you have to adapt to it,” Miguel said.