At the American Academy of Ophthalmology (AAO) 2026 meeting in New Orleans, a panel of ophthalmologists explored real-world applications of artificial intelligence (AI), including the use of conversational technology to conduct preoperative assessments, educate patients, and evaluate postoperative symptoms. The course, “From Consult to Postop: How AI Elevates Patient Care and Practice Efficiency in Cataract Surgery,” featured Aisling Higham, MBBS; Scott E LaBorwit, MD; Lauren S Blieden, MD; Eric L Crowell, MD; and Eileen Choudhury Bowden, MD.

The transcript below was edited for clarity.

Scott LaBorwit: Hi, I'm Dr. Scott LaBorwit, and we're excited as a team to be invited to the American Academy of Ophthalmology to present about what we're finding from consult to post-op—how we're leveraging AI technology to elevate patient care for cataract consult, surgery, and post-op care. So we have a great quick summary of what we're presenting, as far as topics are concerned, that really are going to have a huge impact and change health care. So I want to start off with Aisling to kick off the topic.

Aisling Higham: Amazing. Thanks so much, Scott. I'm Dr. Aisling Haim, and I'm based in the UK. I think really the super exciting thing about using AI in health care is how we can use it to transform and really deliver care in a way that never previously was possible. I'm sure many of you have already used some of the technologies coming through, and every week there'll be a new AI model with ever-increasing capabilities. I'm impressed on a weekly basis by what this type of tech can now do. And conversations that maybe previously were a bit stilted using AI are now fluent, and you could easily believe that it's another human you are having the conversation with. It has knowledge in a way that is not possible for any human, can speak multiple languages, and you can use all of this to really think about how we deliver care in a completely different way.

I think what is, however, very important for us as clinicians and physicians is that, whilst we have all this super exciting technology and we know it can transform care, we actually have to be able to do it in a safe way. And we have to, as clinicians, understand how to use these types of technology, how we know that it's safe, what kind of things to look out for—and assure ourselves that, for our patients, it's the right way to go. So Eric will cover a little bit more about that later, but I think really the key thing for me is it's just such an exciting field, and there's so much potential to do things differently. So I'll just pass it on to Eileen to show how actually we've been able to already use some of the technology in a real-world space, not just in the kind of AI-hype space. It's already happening.

Eileen Choudhury Bowden: Thanks, Aisling. I'm Dr. Eileen Bowden, and I work here in Austin, Texas, at the University of Texas at Austin Dell Medical School. We are at an academic center here, and we've been using the Dora (Ufonia) system to help us with our cataract patients and reduce the burden of cataract post-op visits. Traditionally here in the United States, patients who receive cataract surgery are often seen post-op day 1, often post-op week 1, and then post-op month 1. And for many places with a growing number of people who need cataract surgery, but a shrinking population of ophthalmologists, it is difficult to provide the multiple post-op visits that we've traditionally been doing in a very efficient, sustainable way. And also, it's a burden for patients these days who live very far away from their ophthalmologists and need to travel over an hour or 2 hours to receive care.

It's just not reasonable to ask them to come for frequent, multiple post-op visits, especially if, as we know, after cataract surgery, there's a relatively low risk of complications, especially after that initial day 1 period. So we have been using the Dora system to reach out to our patients at week 1, have a conversation with them, evaluate their symptoms, and then determine whether or not they need that visit. Right now, we're in the trial stage, so we're still seeing the patient at week 1, but then comparing our visit in clinic to the responses that they give to Dora. And it's been a really positive experience for everyone involved. Our patients are loving being involved. They get excited about participating in the trial because they're really excited about AI and what the technology can offer the future of medicine. And the results we talk about in our presentation show that there's a really good agreement between what we're seeing in our Dora phone calls vs what we're seeing in clinic, along with a pretty low rate of missing any significant adverse symptoms or exam findings.

So it's been a really wonderful experience. I'm looking forward to continuing the trial, taking it to the next stage, and learning more about what AI can help us do in our field. And now I think I'll pass it on to Scott to discuss a little bit more.

Scott LaBorwit: Well, thanks for passing off the digital baton there, Eileen. Great summary. I bring a perspective from private practice, so I only do cataract surgery in a mostly premium intraocular lens (IOL) cataract practice, maybe doing 25 or 30 cases each week. And so my focus over time has become more about the patient experience: How can I even come from the front end where they're overwhelmed with information at the consult, and by the time they get to me, they can be exhausted? So having an opportunity for a large language model like Dora to contact patients and talk to them whenever they want—Dora calls them. They could call her back at 2 in the morning after a night shift. They can say, “Wait till my wife is home so she can help me with the answers.”

I helped write those files, so Dora never forgets to ask, “Did you have trauma? Have you had steroid exposure?” She asks all the necessary questions. I have great staff, but staffing has been much more challenging over the last couple of years since COVID-19, so even my best technicians don't always ask those questions—they have 2 more patients waiting for them, they're a little tired because lunch is coming up. Dora doesn't get hungry, she doesn't sleep, and she doesn't have another patient to see. So using this AI technology gets that component done. When they come to my office, I'm telling my techs, “Hey, vision pressure dilate; I'll handle the rest.” Well, really, Dora or the language model handled that. What surprised me the most is that patients, when I ask them about their experience, give Dora 5 out of 5, and so they're happy to have that part done.

When I sit down, I walk in the room much more intelligent, and I'm ready to get down to business about their care, their particular case, and the education. And then Dora educates patients based on how they answer their questions—she will give different levels of education. Also, the language model can ask questions that I never would. They tell Dora many things, and I'm always surprised by how much they appreciate that and how much they do that. So it allows me to have a summary before I walk in the room. I walk in much more intelligent, and it helps me with my premium line of patience.

It's been a game changer for me. One of the things that bothers me the most about AI is—I'm within a big group, and the business-decision leaders are finding AI very useful for revenue cycle management, for setting appointments, for watching productivity, but nobody's bringing AI into the exam room. So the critical piece is the doctor and the patient. How can we fight to preserve trust and accelerate our clinical decision-making by having that information? This is the first technology that I'm aware of in my platform that allows me to do that. We already learned from electronic health records (EHRs) that it didn't help us with patient care. It just helped allow everybody else to see what's happening in the exam rooms. So I get audited more. I have more training sessions to be more compliant. And then I kind of joke that productivity used to be quarterly, but with AI, now I might take a coffee break, and I get a text asking why productivity was down 15% in the last 2 hours.

So that's where AI has been. And then this technology gets into the HIPAA space, makes doctors able to [turn] away from the EHR and face patients again, talk about what matters the most, and build that trust and do the clinical decision-making. So anyway, that's been my experience with this technology, and it's the tip of the iceberg, but that's what we were discussing broadly.

Eric Crowell: Scott brings up a great point about how technology has transformed our practices—sometimes for the better, sometimes maybe not for the better. But I think that's something that we need to be really careful about with AI: How is this going to transform our practice, but how is it going to do so in a safe manner? I think we've heard a lot of things in the news, just all these large language models and the different algorithms that are out there that exist for AI sometimes can cause problems. So we've had chatbots that have encouraged people to actually commit suicide, which is not good. There are other AI models recently that you may have heard of, such as the Hugging Face hacking incident when they actually broke out of their sandbox and started attacking other things in a coordinated fashion. We need to make sure that that does not happen with our health care helpers.

And so I think that's part of the reason why we're doing a clinical trial is we need to prove that this is going to do what it says it's going to do—no more and no less. And so that is already within a regulatory framework, but we might need to broaden that regulatory framework just to make sure that, as we're moving forward with more and more AI helpers, that we're doing it in a way that is actually helpful and not going to potentially cause harm.

Aisling Higham: Amazing. I mean, super useful insights for everybody. I was just going to ask everyone: Do you have any piles of wisdom or tips for the ophthalmologist who's thinking, okay, this could potentially work in my practice? How would you go about getting involved?

Scott LaBorwit: I would just encourage all doctors to say, “OK, what is real AI, and how can I be a part of that process?” I think we learn, going back to the EHRs, we learned that sometimes if we're not in the loop, if we're not helping to build it, it can then become built on top of us. So we're on this podcast because we are riding that wave right now, and thank you at the Dell Medical Center for actually doing the research to prove it. Don't just say, “I'm not into AI,” or if you just wait, then yeah, things are going to continue building around you, as I mentioned, whereas if you're part of the process—how can we get it to fight for the doctor-patient relationship?

And this is just one example, but there are many. So find ways to seek that out, find ways to be a part of the process. And I'll pull you in on this over at the Dell Medical School. Eileen, what kind of surprised you when you started using AI in the pathway? Was it as scary as you thought? Were patients as arm-crossed as it seems, or is it actually a positive experience? What are your thoughts just off the cuff?

Eileen Choudhury Bowden: Yeah, thanks, Scott. As I touched on a little bit ago, I have been surprised by how excited patients have been to enroll in our trial, to be involved with the study, to use AI. Things are different than they were probably even 10 or 15 years ago—people are very used to technology. They know how technology can benefit their lives. They've lived with the internet for most of their lives at this point, so they're very technology-forward. And even the patients who are well into their 70s and 80s are really excited about what AI can do for the future. So there's been pretty low resistance, so I would say to ophthalmologists who are interested in trying this out, don't be afraid of your patients maybe being scared of this. Many of them are very excited, have a positive experience, and are happy to cut down the number of visits. They don't like going into the clinic, so anything that can help keep them at home and doing the things they'd rather be doing rather than trudging into your office and sitting around for 30 minutes or an hour, whatever it may be, they are all for it.

Scott LaBorwit: Yeah, it's kind of funny. I got interested in this technology because my wife called back after she ordered a pizza to correct something, and they said, "Oh, that was our digital assistant you were talking to." And if they could fool my wife, then I said, "That's technology that's evolved beyond." So that's why I said it might be ready for prime time. As I meet with patients, everyone's giving Dora 5 out of 5. There's only been 3 that gave her 4s and one that gave her 3s. But then by the end of the consult, I felt like I wanted to give the patients a 2 or 1 because they were just difficult personalities, not really open to things.

And the funniest thing for me was that my first patient with this technology was an ophthalmologist who’s retired, and I was scared to go in the room. I was like, "What'd you think?" And he's like, "That was amazing. I wish I had that. I could have been much more efficient with my time and my patient's time." And it's funny because this technology is not there to make my interaction shorter. It's not to replace the visit. It's to make it more meaningful. I'm not cutting my time down with patients, but I'm telling them when I walk in the room, because I'm smarter, we can really focus on them and their concerns and plan. So it's not a substitute; it's an enhancement. And we're in control of the data we use. It's not a medical device. You're still the only one in the exam room who can sign the chart, so this makes you better, but doesn't replace you.

I hope you find this helpful. And of course, we're happy to be contacted if you have further questions in any way we can help.