Evaluation of Three Advanced Language Models for Identifying Ophthalmic Assessments and Assisting in Preoperative Toric IOL Planning

1

Cataract surgery has evolved to meet patient demands for high-quality vision and spectacle independence, especially in cases of corneal astigmatism.
2

Preoperative ocular biometry accuracy is crucial, as errors in corneal measurement significantly impact toric IOL power calculations.
3

Large language models (LLMs) can assist in ophthalmology by processing complex datasets, potentially improving preoperative accuracy in toric IOL planning.
4

This study evaluates three LLMs for their effectiveness in extracting biometric parameters and supporting toric IOL planning in cataract surgery.
5

The evaluation involved extracting data from IOLMaster 700 reports and assessing the models' ability to calculate astigmatism and recommend IOL parameters.

1