• Therapeutic Cataract & Refractive
  • Lens Technology
  • Glasses
  • Ptosis
  • Comprehensive Eye Exams
  • AMD
  • COVID-19
  • DME
  • Ocular Surface Disease
  • Optic Relief
  • Geographic Atrophy
  • Cornea
  • Conjunctivitis
  • Myopia
  • Presbyopia
  • Allergy
  • Nutrition
  • Pediatrics
  • Retina
  • Cataract
  • Contact Lenses
  • Lid and Lash
  • Dry Eye
  • Glaucoma
  • Refractive Surgery
  • Comanagement
  • Blepharitis
  • OCT
  • Patient Care
  • Diabetic Eye Disease
  • Technology

ASCRS 2023: Predicting vision outcomes in cataract surgery with machine learning


Mark Packer, MD, sat down with Sheryl Stevenson, Group Editorial Director, Ophthalmology Times®, to discuss his presentation on machine learning and predicting vision outcomes after cataract surgery at the ASCRS annual meeting in San Diego.

Mark Packer, MD, sat down with Sheryl Stevenson, Group Editorial Director, Ophthalmology Times®, to discuss his presentation on machine learning and predicting vision outcomes after cataract surgery at the 2023 ASCRS annual meeting in San Diego

Video transcript

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

Sheryl Stevenson:

We're joined by Dr. Mark Packer, who will be presenting at this year's ASCRS. Hello to Dr. Packard. Great to see you again.

Mark Packer, MD:

Good to see you, Sheryl.


Sure, tell us a little bit about your talk about machine learning, and visual, predicting vision outcomes after cataract surgery.


Sure, well, as we know, humans tend to be fallible, and even though surgeons don't like to admit it, they have been prone to make errors from time to time. And you know, one of the errors that we make is that we always extrapolate from our most recent experience. So if I just had a patient who was very unhappy with a multifocal IOL, all of a sudden, I'm going to be a lot more cautious with my next patient, and maybe the one after that, too.

And, the reverse can happen as well. If I just had a patient who was absolutely thrilled with their toric multifocal, and they never have to wear glasses again, and they're leaving for Hawaii in the morning, you know, getting a full makeover, I'm going to think, wow, that was the best thing I ever did. And now all of a sudden, everyone looks like a candidate. and even for someone like me, who has been doing multifocal IOL for longer than I care to admit, you know, this can still pose a problem. That's just human nature.

And, so what we're attempting to do with the oculotics program is to bring a little objectivity into the mix. Now, of course, we already do that, when we talked about IOL power calculations, we, we leave that up to algorithms and let them do the work. One of the things that we've been able to do with oculotics is actually improve upon the way that power calculations are done. So rather than just looking at the Dioptric power of a lens, for example, we're actually looking at the real optical properties of the lens, the modulation transfer function, in order to help correlate that with what a patient desires in terms of spectacle independence.

But the real brainchild here is the idea of incorporating patient feedback after surgery into the decision making process. So part of this is actually to give our patients and app that they can use to then provide feedback on their level of satisfaction, essentially, by filling out the VFQ-25, which is a simply, a 25 item questionnaire that was developed in the 1990s by RAND Corporation, to look at visual function and how satisfied people are with their vision, whether they have to worry about it, and how they feel about their vision, that sort of thing, whether they can drive at night comfortably and all that.

So if we can incorporate that feedback into our decision making, now instead of my going into the next room, you know, with fresh in my mind just what happened today, actually, I'll be incorporating the knowledge of every patient that I've operated on since I started using this system, and how they fared with these different IOLs.

So the machine learning algorithm can actually take this patient feedback and put that together with the preoperative characteristics such as, you know, personal items, such as hobbies, what they do for recreation, what their employment is, what kind of visual demands they have. And also anatomic factors, you know, the axial length, anterior chamber depth, corneal curvature, all of that, put that all together, and then we can begin to match inter ocular lens selection, actually to patients based not only on their biometry, but also on their personal characteristics, and how they actually felt about the results of their surgery.

So that's how I think machine learning can help us, and hopefully bring surgeons up to speed with premium IOLs more quickly because, you know, it's taken some of us years and years to gain the experience to really become confident in selecting which patients are right for premium lenses, particularly multifocal extended depth of focus lenses and that sort of thing where, you know, there are visual side effects, and there are limitations, but there also are great advantages. And so hopefully using machine learning can bring young surgeons up more quickly increase their confidence and allow them to increase the rate of adoption among their patients for these premium lenses.

Related Videos
Megan Cavet, PhD
© 2024 MJH Life Sciences

All rights reserved.