What question did this study set out to answer?

To enhance the accuracy of intraocular lens power calculation using physics-informed machine learning techniques.

March 24, 2026Open Access

MIOLO Standard Formula v2.5 -- Calibrated IOL Power Calculation with Physics-Informed Machine Learning

Key Points

To enhance the accuracy of intraocular lens power calculation using physics-informed machine learning techniques.
Developed v2.5 of MIOLO Standard formula incorporating a 6-term calibration layer.
Validated across 28 cases with varying eye measurements (AL, K, ACD).
Implemented mechanisms for different eye types, including dampening for long myopic eyes.
Achieved 100% concordance with Barrett Universal II in validation cases.
Retained unchanged coefficients from base model v2.2.
Introduced innovative features for better performance in specific eye conditions.

Abstract

MIOLO Standard v2.5: Calibrated intraocular lens (IOL) power calculation formula using physics-informed machine learning. This version achieves 280/280 (100%) concordance with Barrett Universal II across 28 strategically selected validation cases spanning AL 21.5-29.0mm, K 39.9-46.9D, ACD 2.5-3.8mm. Key innovations include: (1) 6-term Barrett calibration layer with Kavg*ACD interaction and ACD quadratic modeling, (2) dampening mechanism for long myopic eyes with steep corneas, (3) fade-based short eye protection preserving superior hyperopic performance. Base model (v2.2) coefficients remain unchanged. Formula MLV transformation constitutes protected intellectual property. Supersedes v2.2 (DOI: 10.5281/zenodo.18099603).

MIOLO Standard Formula v2.5 -- Calibrated IOL Power Calculation with Physics-Informed Machine Learning

Key Points

Abstract

Cite This Study