Louis Pasquale, MD: Evolution of Artificial Intelligence in Ophthalmology


Ophthalmology is an amazing field
that’s always been on the cutting edge of new technologies. I mean if you
take a quick snapshot through time, one of the things that got me
interested in ophthalmology in the first place is they were among the first
people to use an operating microscope. In ophthalmology, was the discovery of the
first tumor suppressor gene in retinal blastoma. With one of the more
recent revolutions in genomics, one of the first genes discovered in genomics
was complement factor H from macular degeneration and now we’re seeing a new
revolution in AI and ophthalmology is really on the forefront of that. It’s
interesting. Dr. Eric Topol, who is a expert in big data, he’s a cardiologist
here in California has, himself, stated that ophthalmology is really on the
cutting edge of artificial intelligence. What can artificial intelligence do for
ophthalmology? Many things. I’m a glaucoma specialist and in our field
there are many gaps in glaucoma care. If you put five experts in a room and
you showed them images of an optic nerve, there would be considerable
disagreement about whether or not that optic nerve constitutes glaucoma—this
damage or not. It’s kind of surprising that that’s the case, but it’s true. In
the realm of visual field testing, we would love to have an algorithm that
could objectively detect visual field progression and what we’re seeing
outside of glaucoma is explosions in the area of artificial intelligence. Recently
the FDA approved IDX which is an autonomous algorithm for
detecting referral diabetic retinopathy and now you’ve seen an explosion around
the world using that method or related AI algorithms to detect diabetic
retinopathy, which is really critical because we have an epidemic, so to
speak, of diabetes and we have a relative lack of manpower to identify
patients who might have diabetic retinopathy and need treatment and we
all know that diabetic retinopathy is a treatable disease. So, the sky
is the limit about what AI could do next in my field it would be wonderful
if AI could predict—and I think that it could predict—which patients are going
to be rapid progressers, which patients might need surgery, what would
be an appropriate target intraocular pressure for treatment of the disease,
etc. So, there’s tremendous amount that I believe AI still can do in our field.

Leave a Reply

Your email address will not be published. Required fields are marked *