Related Experiment Video
Updated: Sep 15, 2025
Author Spotlight: Unraveling the Molecular Mechanisms in PCO and Fibrosis Following Cataract Surgery
Published on: December 1, 2023
VAULT-OCT: vault accuracy using deep learning technology-an artificial intelligence model for predicting implantable
Matthew T Hirabayashi1, Gurpal S Virdi, Taj A Nasser
1From the Parkhurst NuVision LASIK Eye Surgery, San Antonio, Texas (Hirabayashi, Parkhurst); University of Missouri Columbia School of Medicine, Columbia, Missouri (Virdi); Mason Eye Institute, Columbia, Missouri (Virdi); Tylock George Eye Care and Laser, Dallas, Texas (Nasser); Mueller Vision, Fort Worth, Texas (Nasser); Texas State University, San Marcos, Texas (Abramson).
A new deep learning model, VAULT-OCT, accurately predicts the postoperative vault of phakic Implantable Collamer Lenses (ICLs) using pre-operative OCT scans. This AI tool aids in precise ICL sizing and improves surgical outcomes.
Area of Science:
- Ophthalmology
- Medical Imaging
- Artificial Intelligence
Background:
- Phakic Implantable Collamer Lenses (ICLs) are used for refractive error correction.
- Accurate prediction of postoperative vault is crucial for ICL success.
- Pre-operative anterior segment optical coherence tomography (AS-OCT) is a key imaging modality.
Purpose of the Study:
- To develop and validate VAULT-OCT, a deep learning model for predicting postoperative ICL vault.
- To assess the accuracy of VAULT-OCT using pre-operative AS-OCT images.
- To determine the feasibility of using AS-OCT for ICL sizing decisions.
Main Methods:
- A retrospective study included 324 eyes from 162 patients undergoing ICL implantation.
- The VAULT-OCT neural network was trained on pre-operative AS-OCT images and postoperative vault measurements.
- Model performance was evaluated using Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE).
Main Results:
- VAULT-OCT demonstrated high accuracy in predicting ICL vault across different lens sizes.
- Mean Absolute Error (MAE) ranged from 21.7 µm to 98.1 µm.
- A high percentage of predictions (89.1%–100%) fell within a clinically acceptable 200 µm margin.
Conclusions:
- The OCT-based deep learning model, VAULT-OCT, accurately predicts postoperative ICL vault.
- The model's performance suggests its feasibility for guiding ICL sizing decisions.
- This technology has the potential to enhance refractive surgery outcomes.
