Multicenter evaluation of AI-based intraocular lens formulas and interpretable machine learning for cataract refractive prediction
Abstract
Accurate intraocular lens (IOL) power selection remains challenging in cataract eyes with long axial length, altered corneal geometry, or previous corneal refractive surgery. This multicenter retrospective study evaluated conventional, post-refractive, and AI-based IOL formulas and developed interpretable machine learning models for postoperative refractive prediction. The development cohort included 1203 eyes from West China Hospital; the external validation cohort included 336 high-myopia eyes from Chengdu Aidi Eye Hospital, with 257 eyes available for cleaned counterfactual target-achievement analysis. Formula and model performance were assessed using mean absolute error, root mean square error, median absolute error, and predefined refractive error thresholds. In eyes without prior corneal refractive surgery, AI-based formulas outperformed conventional formulas, with Hill-RBF showing the lowest internal prediction error. Locally developed linear regression, support vector regression, and random forest models achieved comparable internal accuracy, but external validation showed reduced portability. SHAP analysis identified target refraction, axial length, and IOL power as dominant predictors. Post-refractive eyes, very long eyes, and flat corneas remained difficult. These findings support cautious, locally validated use of AI-based formulas and interpretable machine learning in refractive cataract surgery.
Related
- Three-Year Functional Vision, Longitudinal Refractive Outcomes, and Patient-Reported Outcomes After Bilateral Implantation of the Extend Bessel-Optics Intraocular Lens: An Independent Real-World Ambispective Study
- Effect of residual spherical equivalent on visual outcomes of trifocal intraocular lens implantation in patients with cataracts across different refractive states
- Visual, refractive and patient-reported outcomes after two trifocal intraocular lens designs: a 12-month comparative cohort study.
- Distribution and factors associated with refractive errors among young adults attending ophthalmic clinics in Iraq: a multicenter cross-sectional study
- Development and external evaluation of an interpretable machine-learning model for early prediction of organ failure in higher-risk acute pancreatitis patients: A multicentre cohort study
BibTeX
@article{yang2026multicenter,
title = {{Multicenter evaluation of AI-based intraocular lens formulas and interpretable machine learning for cataract refractive prediction}},
author = {Ning Yang and Jiancheng Mu and Yi Liu and Wanyue Guo and Chuhuan Sun and Bosen Peng and Junhan Xiang and Lixing Deng and Wei Fan},
journal = {Scientific Reports},
year = {2026},
month = oct,
doi = {10.1038/s41598-026-76755-8},
url = {https://www.nature.com/articles/s41598-026-76755-8},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Multicenter evaluation of AI-based intraocular lens formulas and interpretable machine learning for cataract refractive prediction
AU - Ning Yang
AU - Jiancheng Mu
AU - Yi Liu
AU - Wanyue Guo
AU - Chuhuan Sun
AU - Bosen Peng
AU - Junhan Xiang
AU - Lixing Deng
AU - Wei Fan
PY - 2026
DA - 2026/10/22
JO - Scientific Reports
DO - 10.1038/s41598-026-76755-8
UR - https://www.nature.com/articles/s41598-026-76755-8
AB - Accurate intraocular lens (IOL) power selection remains challenging in cataract eyes with long axial length, altered corneal geometry, or previous corneal refractive surgery. This multicenter retrospective study evaluated conventional, post-refractive, and AI-based IOL formulas and developed interpretable machine learning models for postoperative refractive prediction. The development cohort included 1203 eyes from West China Hospital; the external validation cohort included 336 high-myopia eyes from Chengdu Aidi Eye Hospital, with 257 eyes available for cleaned counterfactual target-achievement analysis. Formula and model performance were assessed using mean absolute error, root mean square error, median absolute error, and predefined refractive error thresholds. In eyes without prior corneal refractive surgery, AI-based formulas outperformed conventional formulas, with Hill-RBF showing the lowest internal prediction error. Locally developed linear regression, support vector regression, and random forest models achieved comparable internal accuracy, but external validation showed reduced portability. SHAP analysis identified target refraction, axial length, and IOL power as dominant predictors. Post-refractive eyes, very long eyes, and flat corneas remained difficult. These findings support cautious, locally validated use of AI-based formulas and interpretable machine learning in refractive cataract surgery.
SN - 2045-2322
ER -