Ophthalmology at Crossroads for Information Technology, AI

Ophthalmology has already moved towards telemedicine and artificial intelligence (AI) before the COVID-19 pandemic, but these changes are now accelerating, making it crucial for ophthalmologists to adapt.

Ophthalmology, like many medical specialties, is confronted with accelerated changes as a result of the COVID-19 pandemic. Prior to the pandemic, telemedicine was only used for retinal screening, but it has evolved into all aspects of ophthalmology, explained Aaron Lee, MD, MSCI, associate professor at the University of Washington, during a session on information technology at the American Academy of Annual meeting of ophthalmology 2021.

Artificial intelligence (AI) has taken ophthalmology by storm, he said. There are diagnostic models where a person isn’t even up to date, which only happens in ophthalmology – no other area has advanced so far with AI, Lee said.

With this massive change, ophthalmologists must decide how to respond to these changes: deny and ignore, resist the change, or shape the change.

When a new technology is introduced, people experience the Gartner Hype Cycle. After the new technology first emerges, there is a “peak of inflated expectations” – the idea that this technology can do anything and replace humans. This is followed by the “valley of disillusionment” – the realization that the technology is not yet fulfilling all expectations at this point in time. From there is the ascending “slope of enlightenment” – understanding the use of this technology based on early adopters. Finally, there is the “productivity plateau” – the beginning of mainstream adoption.

AI has gone through this cycle. The innovation trigger happened around 2015 when the first deep learning models were developed, explained Lee. We are currently on the downward path of inflated expectations. He expects us to be on the path of enlightenment by 2026.

There are 2 possible future perspectives he sees for AI in ophthalmology. He calls the first the “Skynet” future, in which he envisions a conveyor belt of patients taking pictures with a robot and then an AI model decides what they need: If everything is in order, they can go home or they are directed to a room for a procedure. If they are led into a room, a robot in combination with AI carries out the procedure or operation.

There is no ophthalmologist in this scenario.

In the second scenario, the future consists of “the same stuff as the present”. One reason for this is that there are significant limitations in AI models.

  • There’s a real problem with the fragility of AI models, Lee said. AI “can make a completely catastrophic mistake … and there is no way to find out why.”
  • It can be difficult to get AI to do what we want it to do. For humans, we can see an example of a tree and then 6 different types of trees and understand that they are all trees. AI models can’t do that yet.
  • It’s impossible to really know how and why AI models work. So far we have not been able to understand why or how a model can do a miraculous job.

As a result, there could be a future where not much is going to change because “we can’t overcome some of these things” and ophthalmology may not change that much, Lee said.

In its ideal future, AI would be used as a clinical tool to increase efficiency, improve outcomes, and save eyesight. This would mean that ophthalmologists spend less time typing in electronic patient records and more time talking to the patient. AI would allow clinicians to incorporate information and override the AI model if necessary.

Lee reminded people of what happened to GPS devices. There were people who followed the instructions so blindly that they actually drove into a lake or off a bridge.

“When we work with these tools, we still have to be the responsible person to integrate the information,” he said. Making clinical decisions is still an art form that involves knowing about the patient and their life.

Lee succeeded April Maa, MD, associate professor at Emory University School of Medicine, who discussed how telemedicine will remain relevant after the pandemic.

“Telemedicine is part of the good that came out of the COVID disorder,” Maa said. Telemedicine is no longer the “red-haired stepchild” of medicine. Now we can use telemedicine platforms to iterate and improve because so many doctors used them during the pandemic. “Now we can transfer the knowledge we have gained to the future.”

However, it’s important to note that telemedicine isn’t just video chat. For example, it is not possible to use video chat to say whether a patient’s glaucoma has worsened. Instead, telemedicine will be relevant for the future due to the possibility of hybrid models. Maa described a scenario where a patient comes to the office for testing or has a test done at home, the clinician then reviews the results and video them with the patient.

Telemedicine is not going away either because patients love it. Patient satisfaction with telemedicine is high and more convenient for them. Maa pointed out that this can lead to better compliance and eyesight for her patients. In addition, after the pandemic, patients now expect a telemedicine option that cannot be offered and can be detrimental to a practice.

“I think telemedicine in the future of our ophthalmology will be beneficial for us financially and for our lifestyle.” [doctors]”Said Maa.” It will increase patient satisfaction and be a practice builder. And most importantly, it will promote equity in healthcare and ensure that everyone can see well for as long as possible. “