Using Machine Learning to Predict Vision Outcomes in Cataract Surgery


Mark Packer, MD, discusses Machine Learning and Predicting Vision Outcomes After Cataract Surgery with Sherrill Stevenson, Group Editorial Director, Ophthalmology Times, at the ASCRS Annual Meeting in San Diego.

video transcript

Editor’s note: This transcript has been edited for clarity.

Cheryl Stevenson:

Dr. Mark Packer, who will speak at this year’s ASCRS, will also participate. Hello Dr. Packard. It was great to see you again.

Mark Packer, MD:

Nice to meet you, Cheryl.

Cheryl Stevenson:

Yes, tell us a little bit about machine learning and vision, a story about predicting visual acuity after cataract surgery.

Mark Packer, MD:

Sure, you know, humans are fallible, and surgeons hate to admit it, but they tend to make mistakes from time to time. And as you know, one of the mistakes we make is always extrapolating from the latest experience. So if you have a patient who is very dissatisfied with a multifocal IOL, suddenly you will be more cautious with the next patient, and probably the next.

And vice versa can happen as well. If I had a patient who was so happy with his toric multifocal lenses that he no longer needed glasses and left for Hawaii in the morning completely transformed, I would think: Wow that was the best thing I’ve ever done. And now suddenly everyone looks like a candidate. And even for someone like me who’s been doing multifocal IOLs for longer than I’d like to admit, this can still pose a problem. That is just human nature.

And what we’re trying to do with our ophthalmology program is bring a little bit of objectivity into it. Of course, when we talk about IOL power calculations, we leave it up to the algorithm and let the algorithm do the work. One of the things we’ve been able to do with Oculotics is actually improve the way we calculate degrees. So, for example, instead of just looking at the power of the lens, we can actually look at the actual optical properties of the lens, the modulation transfer function, and relate that to what the patient wants in terms of spectacles. independence.

But the real brainchild here is the idea of ‚Äč‚Äčincorporating post-operative patient feedback into the decision-making process. Part of that actually includes providing apps and apps that allow patients to provide feedback on their satisfaction. This is essentially done by completing a short 25-item questionnaire called the VFQ-25. In the 1990s, Rand Corporation measured visual function and how satisfied people were with their vision, whether they should worry about it, how they felt about their vision, and whether they were comfortable driving at night. and so on. .

So if we can incorporate that feedback into our decision-making, now, instead of me going to the next room, with a fresh mind of what happened today, I can actually see every patient. will incorporate the knowledge of I’ve been doing surgeries ever since I started using this system and how he coped with these different IOLs.

So the machine learning algorithm actually takes this patient’s feedback and uses it to describe personal belongings such as hobbies, what they do recreationally, what their job is, what their visual needs are, etc. Can be combined with preoperative features. . Also, anatomical factors, i.e. axial length, anterior chamber depth, corneal curvature, all taken together, begin to select an interocular lens that actually fits the patient, not just biometrics. You will be able to But we also talked about their personal characteristics and how they really felt about the outcome of the surgery.

This is how I think machine learning can help us, and hopefully help surgeons understand premium IOLs faster. Because some of us have had years of experience before we can really confidently choose which patients are right for us. For premium lenses, especially multifocal extended depth of focus lenses, you know, there are visual side effects and limitations, but there are also significant advantages. Using machine learning is therefore expected to help young surgeons build confidence faster and increase patient adoption of these premium lenses.



Source link

Leave a Reply

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