Eyeworld

74 | EYEWORLD | SPRING 2024 R EFRACTIVE of EVO surgeons I talk to don't use toric if it's under 2 D; they perform an arcuate incision or on-axis incision instead. They'd rather avoid the issues of the lens potentially rotating unless they need 3 or 4 D." Many surgeons are using artificial intelli- gence to improve sizing, Dr. Packer said. Roger Zaldivar, MD, is doing just that. "The last few years, we've finally realized the role and impor- tance that UBM has on ICL selection." While he's had UBM in his clinic for some time, he said that there are difficulties in applying UBM to every single patient and lack of standardization. "Today, since we are living a true AI revolution, there are different means to standardize what we're doing and improve and get the best imag- es," he said. "We're now understanding exactly what we have to measure, and we've trained a deep learning model based on UBM, and we are using different brands of UBM, not just one." Dr. Zaldivar's method—ICL Guru on the Revai platform—processes different videos and clips, extracting the best images automatically selected by the algorithm. That fits the algo- rithm to give you a recommendation based on the whole eye configuration, he said. "We have trained the model with more than 12,500 images, which is very robust, and we are putting thousands of images in each week," he said. "Each time, we are getting less and less outliers. The level of predictability is impressive. We've never experienced this amount of predictability." For example, he said roughly 100% of patients with the 13.7 ICL size, the most difficult ICL be- cause it's the biggest, were within 250 microns of predictability, with an absolute error around 83 microns. Dr. Zaldivar said the main concept of this method is that it's a deep learning algorithm that still uses some parameters of anterior chamber OCT to double check. "It's the combi- nation of the two that allows us to be so sharp," he said. It's important to also consider the concept of a safe vault, he said. "We have been insist- ing in this concept for a long time. Finally, we are happy with everyone else understanding this complexity," he said. While many publica- tions note a safe vault of 250–750 microns, Dr. Zaldivar said a safe vault absolutely depends on each patient and each eye configuration. "If you have a big eye, the margins are much higher. You could have 1,000 vault, and it could be perfect. It would be absolutely depending on the angle of the anterior chamber and lens rise," he said, adding that vault is also considered in his nomogram. Dr. Zaldivar has extensive experience with using ICLs and said he hasn't explanted one for many years. "With this methodology, we're getting predictability in numbers that are really impressive." He's also noticed a change in com- fort with the procedure. "What has changed is the behavior of how we analyze which patient can get an ICL," he said. "We're using every single power of ICL. We don't even hesitate with PRK; we just put in an ICL. We're so comfort- able using this as a standard procedure." Many physicians have experience in this area, Dr. Zaldivar said, and experience helps to deal with outliers, but he said his predictabil- ity method could prove very beneficial in the future. continued from page 73 Another tool for ICL sizing ArcScan is another tool that can be used for imaging ICL sizing, Dr. Nikpoor said. There's a nomogram that can be used on iclsizing.com, she said. It may help simplify things because it's similar to UBM, and a lot of it is automated. However, she added that it is a large expense. Dr. Nikpoor doesn't personally use the ArcScan because she said she's seen so much success with her method of using UBM and white-to-white. "For people who are high volume and have physical space, I think it can help make the preop process a lot more streamlined and take a lot of the nervousness that people have about sizing out of the equation." References 1. Packer M. Meta-analysis and review: effectiveness, safety, and central port design of the intraocular collamer lens. Clin Ophthalmol. 2016; 10:1059–1077. 2. Yiming Y, et al. Evaluation of ciliary body morphology and position of the implantable collamer lens in low-vault eyes using ultrasound biomicroscopy. J Cataract Refract Surg. 2023; 49:1133–1139. 3. Packer M. The Implantable Collamer Lens with a central port: review of the literature. Clin Ophthalmol. 2018; 12:2427–2438. Contact Nikpoor: drneda@alohalaser.com Packer: mark@markpackerconsulting.com Zaldivar: zaldivarroger@gmail.com

• Blank Page
• Blank Page (1)