Abstract
Purpose :
The purpose of this study is to validate the lens thickness (LT) measurement process by anterior segment OCT images, manually and with artificial intelligence.
Methods :
A retrospective chart review of patients that underwent uncomplicated FLACS at the Illinois Eye and ear Infirmary was done. 329 anterior segment OCT images of the LenSx Femtosecond platform were reviewed. Lens thickness measurements were done by 4 physicians and interobserver agreement was determined. Using artificial intelligence (AI), the same measurements were made and subsequently compared.
Results :
Of the 329 anterior segment images, 36 were excluded because bad quality. Mean LT was 520.3 ± 549.6 µm. Interobserver agreement was high across all 4 physicians. Afterward, LT measurements were compared with AI results. The LT measured by anterior segment OCT showed excellent agreement.
Conclusions :
Sensitivity and specificity were high for AI LT measurements. This can improve refractive outcomes when used in biometric formulas.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.