Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
We conducted a single-center, retrospective observational study to evaluate large language models (ChatGPT 4o, GPT-5, DeepSeek) for automated interpretation of de-identified IOLMaster 700 reports provided as raster images. Models produced structured biometric extraction, toric IOL recommendation, and refractive predictions (sphere, cylinder, axis). Primary outcomes included parameter-level agreement and refractive error metrics; secondary outcomes included decision-support performance for toric IOL selection and agreement on ordered T-codes. No clinical intervention was performed.
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.
Not provided
Not provided
Not provided
Not provided
Not provided
| Measure | Description | Time Frame |
|---|---|---|
| Refractive prediction error for sphere | Mean absolute error (MAE, diopters) of model-predicted sphere versus clinical reference | At index examination |
| Cohen's kappa with 95% CIs between model | Cohen's kappa with 95% CIs between model outputs and clinician-validated reference for per-parameter | At index examination (single time point) |
| Measure | Description | Time Frame |
|---|---|---|
| Cylinder prediction error | Mean absolute error (MAE, diopters) of model-predicted Cylinder | At index examination |
| Axis prediction error | Mean absolute error (MAE, diopters) of model-predicted Axis |
Not provided
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:
Not provided
Not provided
Not provided
cataract patients who had undergone phacoemulsification and implantation of an AcrySof IQ Toric monofocal intraocular lens
| Name | Role | Phone | Extension | |
|---|---|---|---|---|
| Xuanqiao Lin | Contact | +8615088920668 | 1532483480@qq.com |
Not provided
| Facility | Status | City | State | ZIP | Country | Contacts |
|---|---|---|---|---|---|---|
| Eye and ENT hospital of Fudan University | Recruiting | Shanghai | Shanghai Municipality | 200000 | China |
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
| ID | Term |
|---|---|
| D002386 | Cataract |
| D001251 | Astigmatism |
| ID | Term |
|---|---|
| D007905 | Lens Diseases |
| D005128 | Eye Diseases |
| D012030 | Refractive Errors |
Not provided
Not provided
Not provided
Not provided
Not provided
| At index examination |