AI Integration in Eyecare: Enhancing Patient Outcomes and Practice Efficiency

Maria Sampalis, OD:

Hi everyone. I’m Dr. Maria Sampalis. I practice in Cranston, Rhode Island.

Question:

How do you see artificial intelligence currently being integrated into eye care practices and what specific tasks or areas of diagnosis do you find AI most useful for?

Maria Sampalis, OD:

Yeah, I mean, I think we started with AI. I’m seeing it a lot with early signs of diabetic retinopathy. I think a lot of technology has integrated this to be able to notify a patient if there is diabetic retinopathy on a retinal image. I think there’s a lot of potential for the future, and I see artificial intelligence will continue to improve as long as deep learning evolves as well.

Offices are also using ChatGPT for their offices, for doctor patient letters, for social media, for job hiring, things like that. So it’s the early stages, but a lot of potential and a lot of tasks that I think we can use in our practice. Even early diagnosis now, I just saw some things from Vision Expo East with glaucoma and OCT measurements.

Question:

Can you share any experiences or examples of how AI technology has improved patient outcomes or efficiency within your own practice or in collaboration with other eye care professionals?

Maria Sampalis, OD:

Yeah, so I use AI in my practice. I’ve used Altris AI. It’s FDA cleared AI software that analyzes OCT scans. So it’s been great to help narrow down pathological detection of diseases in the eye. If I was questioning a certain diagnosis, kind of helped me pinpoint what it was before if I need to send to a retina specialist.

I think AI over the years is going to change a lot in the practice, but also we’ll see what a progressive disease might look like or the chances of disease from a healthy retina. The stuff that we can’t see as a practitioner. But I think AI scribe will be introduced fairly easily into optometry practices as well for EMR and again, ChatGPT. There’s a lot of other software out there now. There’s a new software, SCI Dry Eye software, which helps accuracy and diagnosis and patient outcomes for dry eye. There’s a lot more going to come out soon too.

So I think it’s a great way to help get better patient outcomes, help the practitioner be more efficient and help grow your practice.

Question:

What are some of the challenges or limitations you’ve encountered in implementing AI solutions in eye care, and how do you address these obstacles to ensure successful integration and utilization?

Maria Sampalis, OD:

Yeah, so these are still in the works. It still needs process, still need the practitioner, but there’s cyber risks. Of course, there’s ethical concerns, how the deep learning process and it still needs diverse data. It can be biased in the AI. So this is still in the works. It still needs to be worked on, but I mean there’s a lot of potential for it in our industry and really to help us grow our business and have better patient outcomes. Maybe help reduce the loss of vision with glaucoma, macular degeneration and other things. Things that we might not be able to catch right away in just a routine eye exam that maybe something like this can check for. Maybe there’s stuff out there right now with diabetic retinopathy, but also for checking a scan of the retina for heart disease.

Question:

With the rapid advancement of AI technology, what potential do you envision for its future role in eyecare diagnosis, treatment planning and patient management?

Maria Sampalis, OD:

Yeah, it can be a great screening tool. Again, AI can detect glaucoma or diabetic retinopathy. Diagnostics. It can help professionals with combination of signs, symptoms, testing. Maybe we put in the signs and symptoms that we’re seeing and maybe help us give us differential diagnosis. Help with patient education to be able to have a script for patients when we’re talking about cataracts or something that we’re not comfortable with. Let’s say like a new disease that we’re not comfortable, that maybe we don’t see often being able to help with that. Maybe it will also help with compliance and help decrease the need for extensive patient education in the office as well.

But also, it could be a way to teach optometrists in the future with AI models on things that we’re maybe not comfortable with. Maybe optometrists that’ve been out of school for a while. Be able to simulate certain diseases in patients and help with techniques and education on what it is and how to diagnose and treat it.

Question:

How do you ensure that the use of AI in eyecare compliments rather than replaces the expertise and judgment of the eyecare professional, and what steps do you take to maintain patient trust and confidence in the care provided?

Maria Sampalis, OD:

Yeah, so I think we need to be at the table. This is new technology out and I think if we’re not at the table to make these decisions, others will. I think it is important to know that it is here to stay. It will help us as clinicians, but it won’t replace us as clinicians. I think that patient-doctor relationship is important. I think patients need to know that they still need a doctor to take the accurate measurements, to still have a clinical protocol to follow and a treatment and plan decision that is ultimately up to the doctor to make a decision. But I think it could be a great assistance to the practice and a great assistant to maybe a population of patients that might not have access to doctors to be able to have a final say when the doctor signs off.