Related Experiment Video
Updated: May 30, 2025
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
Development of machine learning-based models for vault prediction in implantable collamer lens surgery according to
Timoteo González-Cruces1, Francisco Javier Aguilar-Salazar, Jordi Marfany Tort
1From the Department of Anterior Segment, Cornea and Refractive Surgery, Hospital Arruzafa, Cordoba, Spain (González-Cruces, Aguilar-Salazar, Sánchez-Ventosa, Villarrubia, Cerdán Palacios, Cano-Ortiz); Department of Anterior Segment and Refractive Surgery, Clínica Simon Oftalmología, Barcelona, Spain (Marfany Tort); Barraquer Ophthalmology Center, Barcelona, Spain (Mateu, Barraquer, Pardina); Department of Health and Biomedical Sciences, Universidad Loyola, Andalucía, Spain (Cano-Ortiz).
Purpose:
To develop a prediction model based on machine learning to calculate the postoperative vault and the ideal implantable collamer lens (ICL) size, considering for the first time the implantation orientation in a White population.
Setting:
Arruzafa Ophthalmological Hospital (Cordoba, Spain) and Barraquer Ophthalmology Center (Barcelona, Spain).
Design:
Multicenter, randomized, retrospective study.
Methods:
Anterior segment biometric data from 235 eyes of patients who underwent ICL lens implantation surgery were collected using the anterior segment optical coherence tomography CASIA II to train and validate 5 types of multiple regression models based on advanced machine learning techniques. To perform an external validation, a dataset of 45 observations was used.
Results:
The Pearson correlation coefficient between observed and predicted values was similar in the 5 models in the external validation, with least absolute shrinkage and selection operator regression being the highest ( r = 0.62, P < .001), followed by random forest regression model ( r = 0.60, P < .001) and backward stepwise regression ( r = 0.58, ρ < 0.001). In addition, the predictions generated by the different models showed closer agreement with the actual vault compared with the Nakamura formulas. Using the new methods, about 70% of the observations had a prediction error below 150 μm.
Conclusions:
Advanced forms of regressions models based on machine learning allow satisfactory calculation of the ideal lens size, offering greater precision to surgeons customizing surgery according to implant orientation.
