Performance of Large Language Models for Structured Recognition and Refractive Prediction

This study compares three large language models accessed in their native configurations, without fine-tuning or external tools. For each examination, the original IOLMaster 700 report image was supplied without manual annotation or pre-processing. A standardized instruction required: (i) structured extraction of AL, ACD, LT, WTW, K1/K2 and axes, ΔK, TK1/TK2 and axes, and ΔTK; (ii) binary toric candidacy and T-code according to institutional ALCON mapping; and (iii) refractive recommendations (sphere, cylinder, implantation axis). Each model generated three independent outputs per case. De-identification and IRB oversight (waiver of consent) were implemented according to institutional policy. The unit of enrollment is participants (n=54), with outcomes analyzed per eye (162 eyes) and per model generation where applicable.

Inclusion criteria

-postoperative corrected distance visual acuity (CDVA) of 0.10 logMAR or better -an absolute IOL rotational stability of less than 10∘ at the 1-month follow-up examination

Exclusion criteria

• incomplete biometric data on the examination report;
• a history of previous ocular surgery or ocular trauma
• the occurrence of intraoperative complications, such as an anterior capsular tear or posterior capsular rupture
• the development of significant postoperative complications, including but not limited to severe intraocular infection or inadequate pupillary dilation.