Comparative postoperative topography pattern recognition analysis using axial vs tangential curvature maps

Abstract

Purpose: To determine prediction accuracy of patient refractive surgery status by novice reviewers based on topography pattern analysis using axial or tangential anterior curvature maps.

Setting: Four U.S. academic centers.

Design: Prospective case-control study.

Methods: Image evaluation was performed by novice reviewers (n = 52) at 4 academic institutions. Participants were shown 60 total images from 30 eyes presenting for cataract surgery evaluation with known refractive surgery status, including 12 eyes imaged with Placido-based topography and 18 eyes imaged with Scheimpflug-based tomography. There were 12 eyes with myopic ablations, 12 eyes with hyperopic ablations, and 6 eyes with no previous refractive surgery performed. Each eye was shown in both axial and tangential curvature from either device, reviewed as a single image at a time, and masked to the map type (axial vs tangential).

Results: For the 52 novice reviewers included, accuracy of pattern identification was 82.9% (517 of 624) for tangential vs 55.0% (343 of 624) for axial maps for eyes with myopic ablation (P < .00001), 90.9% (567 of 624) for tangential vs 58.3% (364 of 624) for axial maps for eyes with hyperopic ablation (P < .00001), and 15.4% (48 of 312) for tangential vs 62.8% (196 of 312) for axial maps for eyes with no ablation (P < .00001). There were no significant differences between Placido and Scheimpflug devices and no significant differences across groups based on year of training.

Conclusions: Tangential curvature maps yielded significantly better pattern recognition accuracy compared with axial maps after myopic and hyperopic corneal refractive surgery ablations for novice reviewers. Using tangential curvature maps, especially for challenging cases, should benefit post-LASIK intraocular lens (IOL) calculator selection and, thereby, improve IOL power calculation accuracy.