Part 2: Ask the Experts: What problem in ophthalmology do you wish there was a solution for?

This series features experts in ophthalmology sharing their thoughts on the one unsolved challenge they wish there was a solution for.

Video Transcript:

Editor's note: The below transcript has been lightly edited for clarity.

Joaquin DeRojas, MD:

I wish that there was a solution for presbyopia in a way that exceeds what we currently have. We have excellent technology, and it's getting better and better. So this is kind of a dream that will continue throughout time, but it will be nice to, one day, be able to actually create a device that can accommodate in the way that our natural eyes do.

Valerie Biousse, MD:

There are two aspects of implementing a camera in an emergency department that will need a lot of work in the next couple years. One is figuring out a streamlined way to implement the camera. You know, it's one the cameras exist. People who can interpret the cameras are ready. It's easy to plug in a camera and integrate the camera in any electronic health record nowadays. What's difficult is to educate the personnel in an emergency department and connect the camera to the specific flow of the emergency department, and then change the flow in the emergency department.

So one of my big wish is to have a streamline process, a little bit like a manual of procedure, which could even come with the camera. And people would say, I'm buying a camera for a primary care office, or an emergency department, or a neurology office, and here is the manual of procedure that you can use in order to make it happen easily. That's one aspect. And I wish that this will happen soon, because it is the only way for us to really disseminate those cameras everywhere, they will be useful.

The second aspect is increasing the access to artificial intelligence using those cameras. So we know that it is completely possible to have any deep learning algorithm automatically interpret the fundus photos. It is feasible. It can be done by anyone. Multiple research studies have shown that. What is very, very difficult is to make sure that these AI deep learning algorithms are actually available, either in the cameras, or ideally on the cloud, so that each time a picture is taken, the picture is uploaded to the cloud and an answer comes back. Not just for diabetic retinopathy screening, which is great in primary care offices, but not really useful in emergency departments. In emergency departments, we need an automatic response that will say the optic nerve is normal, the retina is normal. It's a central retinol artery occlusion. I suspect giant cell arteritis, or the nerve is swollen, or it could be papilledema intracranial pressure, and we have created those algorithms. The big, big challenge right now is complete the real life studies that are necessary so that the FDA agrees to clear those deep learning algorithms, so that we can use them with the cameras.