Artificial intelligence—where does it fit in eyecare? For example, is AI going to replace an eye doctor? The answer is no. I don't think that's even a practical thought in this process as such. But artificial intelligence is starting to show up in a number of studies through instrumentation evaluation of images.

So, the idea that we could take an image of the eye and teach a computer to recognize things that are different or off from a normal person's eye, in that same age group, and then give us a percentage of risk that, okay, maybe this patient could develop cardiovascular disease or Alzheimer's. I think that that is a huge possibility. I think we'll definitely get there.

We do currently use AI actually, OCT scan, and it breaks down and shows us the layers of your retina. And there's a lot of research that a lot of those layers can be affected by all these different diseases. So, there's AI that's coming out eventually that will read those layers and say, okay, maybe this person is at risk of Parkinson's or Alzheimer's or whatever. But right now, we already use those layers and AI to determine if you're at risk for diseases that affect your eyes. So, glaucoma, for example, one of those layers gets really thin in glaucoma. It's called your nerve fiber layer. And so right now we use an AI that takes your nerve fiber layer thickness, compares it to other people's normal nerve fiber layers that are the same age as you, and then it gives us a percentage of risk of progression. So that is AI, and we use that every day. I use that every day when I see patients. Right now,

When we think about AI and tying it to this retinal image, a number of studies have been done on patients who they had retinal images for, coupled with their health history. So, if we have a picture of the inner eye, a retinal image, and we have a health history, but it's de-identified, meaning it's not tied to any person, we can use artificial intelligence and actually teach it. To do any of this, you need volumes and volumes of information, some normal, some abnormal, and some barely abnormal, if you will.

So, when we take an image of your eye, we take an image of your retina, it's called your fundus. If you've seen, it's like an orange disc picture. All kinds of diseases can affect this, right? If you have diabetes, there's little spots of blood, like blood dots. If you have high blood pressure, there can be pinching or narrowing of the blood vessels. And right now, I go in and look at that image. Eventually, if a computer could scan it and highlight those areas and say, here is a blood splat. It doesn't require me then to look. I am a human. I can make mistakes. I can miss things. A computer would be much more accurate. That computer could then grade that disease and say, this person is diabetic and this is their risk for developing worse problems. So, then we could take that number and then say, “Hey, you are at high risk of developing this, so maybe we should make some changes with your primary care doctor or to your medications.”

They did a two year to three year look back in some of these studies to try to validate, can a retinal image provide us a risk factor for a patient currently or in the future? And the answer was it was 70 to 80% accurate on the first pass. And we're seeing these validated by different sites, using different databases around the world today. I think what's going to happen is that now that they've figured out that the OCT is this even bigger wealth of information than the retinal image, because it has so many applications because of the neuro factor—it's an extension of the brain—they're going to find more and more and more conditions through the AI programs assessing the OCT. And I don't think that people are going to be able to get enough of it. It's like, whoa, I truly see that this may be the opportunity for eyecare to be at the forefront of health screenings.

About EyeMed

EyeMed Vision Care is a leader in vision benefits.1 We deliver stand-out vision benefits, and an outstanding member experience across America’s largest vision network.2 EyeMed is based in Cincinnati, Ohio. Learn more at eyemed.com.

  1. EyeMed internal book of business, 2023.
  2. Based on the EyeMed Insight network and analysis of competitors’ largest networks via competitor data, 2024