Editor’s Column February 2025
Save as PDF
Artificial Intelligence in Optometry
Have you started using AI in your practice yet? Have you started using AI when managing your patients? I would imagine that most of us said no to the first question but paused after the second question. We might not (yet) use AI to help us code more efficiently or screen retinal photos. But we’ve all used search engines to look up diseases and diagnoses, and most search engines now incorporate (or will soon incorporate) AI into their search functions. Thus, in some respects we’re pretty close to using AI to manage our patients. I started to think about AI in optometry while reading our new Guess-the-Disease article by Shelly Kim, an optometrist at the James Lovell VA in Chicago.
I won’t spoil the answer for the disease in question in Dr. Kim’s column, but it’s one that involves lab testing. When I’m in clinic, every time I test for diseases which require labs I find myself looking them up. I’m always fairly certain I know which tests to order, but I never completely trust my memory and it’s comforting to be reassured by the literature. Early in my career I would look it up in textbooks. I spent a decent amount of money buying (at least) one textbook for almost every ocular specialty, and I would pull one down from the shelf when I needed it. Over the last five years I find myself ignoring the books and just searching for the information on the internet. More and more I read about eyecare not to learn new information, but just to refresh my memory on information I (supposedly) learned years ago.
What if I’m not doing myself any favors by just finding the answer to one small question and not reading more thoroughly? What if I knew about every condition only in bullet points? Would my intelligence prove artificial? Would I have false memories that I’m supposed to prescribe this dosage and not that dosage? I think trying to manage a patient’s condition by either memorizing facts or looking up facts is the wrong way to go about caring for that patient. To learn a subject completely, we should read full-length articles about the topic. Several of them. We need to use AI as a tool to improve our own skills but we also (as trite as it is) need to know enough about the condition to ask the right questions and to properly interpret the answers we’re given.
To better use artificial intelligence as a tool, we must improve our intelligence. What better way to do that than to read and write. Read about an entire subject; don’t do what I’ve done in the past and skim enough to find an answer and then stop. Start a chapter at the beginning and stop at the end. Find an in-depth article about a condition and read the whole thing. Once you’ve learned all you can, apply that information to your patient care. Write up what you’ve learned and tell others how to do it. AI has its machine learning, and we have ours: see one, do one, teach one.
Dr. Rett is the Editor-in-Chief of the Journal of Medical Optometry and the secretary for the American Board of Certification in Medical Optometry. He is the Chief of Optometry at VA Boston and sits on the national Field Advisory Board for Tele Eyecare at the VA. He is adjunct clinical faculty at several optometry schools, lectures nationally and enjoys writing about eyecare whenever and wherever.
