Artificial intelligence is the current buzzword in all walks of life, but few know that it was first coined in 1956! It has taken time for technology to catch up to its promise, but it’s improving every day and eye care is one of the major fields that can benefit. The good news is that it can’t succeed without our expertise—to interpret, to lead the process and to carry out the decisions. But in combination with ODs, AI has the potential to shape optometry’s future.

Areas where I expect AI will help include enhancing diagnostic accuracy, identifying diseases—even those outside of eye care—increasing practice efficiency, providing better treatments and expanding access to remote regions. The recent “Healthcare from the Eye” partnership between Topcon and Microsoft will use AI to identify disease early and create integrated systems to allow for collaborative care. 

Current AI Tools

ChatGPT and other chatbots are already changing how we work. While it can write a paper, it can also amass data and science from everywhere on the web into a single document. My cousin even used it to write a love letter to his wife and then, after reviewing the first draft, asked the chatbot to make it less sappy— and it complied. Maybe not the most romantic approach, but the point is that you can modify what’s provided, and the need for a human is essential. 

Disease diagnosis is another obvious fit. Deep-learning algorithms need large quantities of labeled data to be trained, and eye care has that in abundance. 

A 2023 study showed that an AI system had a higher sensitivity for detecting mild DR than either general ophthalmologists or even retina specialists.1 Another involving more advanced DR also showed improvement over human assessment.2 This May, we witnessed the FDA approval for an AI-based camera to obtain and analyze DR with 88% sensitivity and 94% specificity.

A colleague and I worked on an AI-driven OSD algorithm that amazed me in terms of providing insights that may have taken months or years to observe. However, the program would have failed without humans guiding the process.

Emerging Capabilities

Ambient listening and transcription technology—allowing hands-free charting for optometrists—may be one of the most exciting new AI areas. The AI directly captures, structures and summarizes key information in real-time during patient consultations, filtering for relevant details to create concise documentation of each patient’s record. 

A study conducted at a major ophthalmic center with over 300,000 consultations showed doctors gained two hours per day each, over 96% of text was deemed accurate and the charts were 2.5 times more detailed than the previous manual entry method. The time saved has been used to increase patient visits by up to 30% in some clinics. An optometry-only EHR software (Barti) incorporates this exact technology into its system. To date, not a single optometrist using Barti has returned to their previous set-up. 

A technology for advanced AMD called Eyedaptic also uses AI to better serve patients. It already improves reading and functional vision capabilities by over 50% by using augmented reality glasses to move an image off the macula to healthier retinal tissue, but its most recent release taps into generative AI and large-language models to visualize and interpret real-time data. For example, it can read text, describe a room, locate objects and help users with other daily tasks and activities that otherwise may not have been possible due to their vision loss. 

A fascinating new spectacle lens called ColorBoost (Hue Lens) is designed using generative AI to enhance color vision. Hue’s AI-driven process uses data about ocular biology and lens chemistry to create spectacles designed for various activities, including pickleball and a multitude of special environments. These lenses are used by popular eyewear brands and also used in ballistic-protection eyewear for military and government applications.

While there are many opportunities for AI in optometry, ultimately it’s our ability to monitor these advances—through our filter of helping patients, improving practice efficiency and the enjoyment of clinical practice—that will help determine precisely how AI shapes the future of eyecare.

Dr. Karpecki is medical director of the Dry Eye Institutes of Kentucky and Indiana. He is the Chief Clinical Editor for Review of Optometry and chair of the New Technologies & Treatments conferences. A fixture in optometric clinical education, he consults for a wide array of ophthalmic clients, including ones discussed in this article. Dr. Karpecki's full list of disclosures can be found here.

1. Lim JI, Regillo CD, Sadda SR et al. Subgroup comparison of the EyeArt system with ophthalmologists' dilated examinations. Ophthalmol Sci. 2022;3(1):100228. 

2. Ruamviboonsuk P, Krause J, Chotcomwongse P, et al., Deep learning vs. human graders for classifying diabetic retinopathy severity in a nationwide screening program. NPJ Digit Med. 2019;2:25.