Artificial Intelligence (AI) and Retina
Caroline R. Baumal, MD: Just like OCT and anti-VEGFs were developed hand-in-hand, I think that artificial intelligence (AI) is going to find more of a place in retina. AI might help us decipher if some of these therapies that we have developing could be right for a specific type of patient. I mean, overall, I think we probably start to treat our AMD patients just a little late. What if we could treat them earlier to prevent them from developing geographic atrophy or to prevent choroidal neovascularization from leaking? It would be great if we could look at huge data sets to decipher this information, something bigger than we're used to doing.
I think we probably need AI and machine learning to help us decipher that. I envision AI as something that we use like a Harrison's book that we carry around in our pocket, something that helps us determine, looking at various patient factors, what's the best treatment for this patient that we have in our chair within 5 seconds, tells us what the best treatment is. We'll still have a really important role. We'll still have to assess the patient's response to the treatment, but it can help us better direct our patient to one of the appropriate therapies.
David Eichenbaum, MD: AI is super exciting because it doesn't just have a promise in MAT which is a huge benefit that we could derive from AI, but looking at AI from an entirely different perspective, if we could deploy AI into at-risk populations like diabetics, even with commercially available platforms like the iDX camera, if we could deploy that outside of the primary care doctor's office, put it into commercial settings, put it into retail settings, put it into settings where we think we have diabetics, and use AI to capture disease without the burden of the human, we could potentially take care of a lot of patients earlier.
We can use AI to both perhaps find patients, and then we can use AI to treat them in the most appropriate fashion. It has this huge spectrum in the ways it can be applied, and it's going to be a revolutionary part of patient care in a different way than drugs are, I think, once we figure out exactly how to use it and how to deploy it. It's going to be a lot of fun to see that develop as well.
