Arjan Hura, MD, points out that one of the most rewarding aspects of artificial intelligence integration in laser cataract surgery is how to boost surgeons' trust. (Image credit: Adobestock/Zabhie)
Surgeons strive for accuracy, efficiency and consistency in every step they perform. Integrating AI (AI) into laser-assisted cataract surgery (LACS) is a major step forward in achieving these goals. In my experience, the robotic cataract laser system (Ally; Lensar, Inc) incorporates two AI-driven features (image analysis and laser fragmentation optimization) to streamline the planning and execution of the surgery.
3D lens reconstruction and image segmentation accuracy
AI plays a key role in segmenting diagnostic imaging, another area where accuracy is most important. The platform utilizes deep neural networks trained with thousands of Scheimpflug images to accurately depict the cornea and lens capsules, even in the presence of dense cataracts.
Studies have shown that AI-driven segmentation can repeatedly identify boundaries of these structures at the pixel level, reconstructing surfaces with significant accuracy, and reducing the risk of errors during cataract surgery.1
Furthermore, AI-assisted pupil and limb segmentation further enhances intraoperative alignment and iris registration, improving the predictability of toric IOL placement and astigmatism correction.2 Although manual marking may continue to be performed as a backup, the automated alignment feature of the robotic cataract laser system greatly streamlines the process and provides an additional layer of accuracy.
AI-driven customization for cataract density
The challenge with LACS is to determine the ideal laser energy and fragmentation pattern for each eye. Traditionally, surgeons rely on predefined settings such as standard or dense cataract modes to determine energy levels. Although effective, this approach lacks nuance and often requires intraoperative adjustment. Alternatively, robotic laser cataract surgery can be used to analyze a vast dataset of cataract images to automatically adjust energy settings based on cataract density, minimizing faco energy while optimizing efficiency.
Like many surgeons, they use predefined laser fragmentation patterns depending on the specific case. This is a case that includes light adjustable lenses where routine cataracts, dense nuclei, or precise capsule incision sizing is important. Rather than manually adjusting these settings for each case, a robotic cataract laser system allows you to select predefined patterns and allow AI to fine-tune the laser energy parameters in real time. Currently, fewer intraoperative adjustments have been achieved, faster case times, more clear postoperative cornea and more consistent throughout the procedure.
One of the most rewarding aspects of AI integration in laser cataract surgery is how to increase surgeon trust. Knowing that my surgical parameters are continuously optimized based on actual data allows me to focus on the nuances of each case rather than on the micromanagement settings. This efficiency leads to a smoother surgical experience not only for me but for my patients.
Arjan Hura, MD
E: arjan.hura@gmail.com
Hura is a private practice at the Maloney-Shamie-Hura Vision Institute in Los Angeles, California. He is Lensar, Inc. I am a consultant for.
reference
-
Morley D, Evans M. Scheimpflug image segmentation using Deep Learning. Presentation: 2024 Annual Meeting of the Association for Visual and Ophthalmology Research. May 5-9, 2024. Seattle, Washington.
-
Morley D, Evans M. Multi-device student, limbus, and eyelid segmentation using Deep Learning. Presentation: 2024 Annual Meeting of the Association for Visual and Ophthalmology Research. May 5-9, 2024. Seattle, Washington.
