A Clinical Decision Algorithm Based On Multimodal Corneal And Visual Quality Assessment For Refractive Laser Eligibility Screening

Department of Ophthalmology, Istanbul Medipol University, Istanbul, Turkey

In this retrospective study, we reviewed consecutive patients undergoing refractive laser eligibility screening and applied a rule-based clinical decision algorithm. The algorithm incorporated corneal tomography parameters (BAD-D, thinnest corneal thickness, keratometry), refractive error, estimated ablation depth, and visual quality factors. Procedure-specific safety thresholds (residual stromal bed, ablation ratio) were applied according to surgical technique. Each eye was analyzed independently and classified as suitable, conditional, or not suitable. Algorithm classifications were compared with clinician decisions, with subgroup analysis in operated eyes.

Eighty eyes from 40 consecutive patients were analyzed. The algorithm classified 61 eyes (76.3%) as suitable, 11 (13.8%) as conditional, and 8 (10.0%) as not suitable. Overall agreement was 67.5%. In the surgical subgroup (32 eyes), agreement increased to 90.6%, with 29 categorized as suitable and two as conditional. One operated eye was classified as not suitable due to residual stromal bed and ablation depth thresholds. Disagreements mainly occurred between the suitable and conditional categories, reflecting conservative safety limits or additional clinical factors not incorporated into the model.