AI outperforms human experts in detecting glaucoma

Artificial intelligence (AI) might be able to help doctors make glaucoma screening widely available, a new study says.

A trained AI program correctly identified patients with glaucoma 88% to 90% of the time, compared with 79% to 81% for human graders, researchers reported Saturday at a meeting of the American Academy of Ophthalmology (AAO) in Orlando, Florida.

“Glaucoma remains one of the most common causes of vision loss that can’t be repaired globally,” said lead researcher Dr. Anthony Khawaja, a professor at the University College London Institute of Ophthalmology.

“To date, screening is too expensive for glaucoma, but I hope that artificial intelligence solutions, in combination with other approaches such as targeting by genetic risk, will be the solution,” Khawaja added in a news release.

For the study, researchers had both an AI program and human experts evaluate more than 6,300 participants, of whom nearly 700 had glaucoma in at least one eye.

Glaucoma occurs when fluid pressure builds up in the eye. The pressure does damage to the optic nerve, creating blind spots in a person’s vision and eventually leading to total blindness.

The AI and human experts graded people’s glaucoma risk based on a key measure of the eye disorder vertical cup-disk ratio, which tracks changes in eye structure caused by fluid pressure buildup.

Researchers said their results are encouraging because only 11% of the eyes in the study were suspected of glaucoma, which mirrors the proportion that would be found during routine screening.

They said the accuracy might be further improved by including other indicators of glaucoma risk, such as pressure within the eye.

Glaucoma is typically treated using eyedrops to lower eye pressure. In some cases, surgery might be required, according to the AAO.

Copyright © 2025 HealthDay. All rights reserved.