Free Papers

Does BrAIn Cornea empower us to predict the risk of corneal ectasia after keratorefractive surgery?

Free Paper Details

First Author: A.Marta PORTUGAL

Co Author(s):    P. Baptista   A. Abreu   S. Monteiro   M. Pinto           

Abstract Details

Purpose:

The purpose of this study was to analyse different scores to predict corneal ectasia after keratorefractive surgery in surgery candidates with suspicious tomographies and evaluate their correlation with the results of the Brazilian Study Group of Artificial Intelligence and Corneal Analysis (BrAIN) online platform.

Setting:

This retrospective study was conducted at ophthalmology department of a tertiary referral centre in Oporto, Portugal - Centro Hospitalar Universitário do Porto.

Methods:

The eyes of candidates for keratorefractive surgery with a borderline “D” value on the Belin/Ambrosio Enhanced Ectasia Display (BAD-D) evaluated in our department between January 2019 and February 2020 were included. We evaluated the demographic data, spherical equivalent refraction (SE) and tomographic parameters (Oculus Pentacam®). Ablation depth and residual stromal bed thickness were calculated by LASER Excimer Wavelight® EX500 (ablation zone=6.5). The Randleman Ectasia Risk Score System (RERSS), the Percent of Tissue Altered Index (PTA) and the ectasia risk by BrAIN Cornea online platform for Laser-Assisted in Situ Keratomileusis (LASIK) (BrAIN-LASIK) and Photorefractive keratectomy (PRK) (BrAIN-PRK) were calculated and evaluated.

Results:

The sample included 29 patients(58 eyes). The mean age was 31.34±6.04years and 62.1% were female. PTA was <40% (low incidence of ectasia) in 98.3% of cases. However, RERSS was classified as “low-risk” in 8.6%, “moderate-risk” in 25.9% and “high-risk” in 65.5% of cases. BrAIN-LASIK revealed higher risk of ectasia than BrAIN-PRK, except 3 eyes(p<0.001). BrAIN score was <20% in 1.7% for LASIK and in 48.3% for PRK. It was >50% in 91.3% and 32.7% of eyes if LASIK and PRK were performed respectively. We found positive correlations between RERSS and BrAIN-LASIK(r=0.332,p=0.011) and BrAIN-PRK(r=0.400,p=0.002); and between PTA and BRAIN-LASIK(r=0.461,p<0.001) and BrAIN-PRK(r=0.449,p<0.001).

Conclusions:

This study provides an insight into different scores in borderline cases based on clinical, and on pentacam parameters. BrAIN score have positive correlations among RERSS and PTA. BrAIN platform allows to distinguish risk between LASIK and PRK and was able to more objectively quantify it, by using artificial intelligence. The authors believe that this platform can give us more information and more strongly support our clinical decisions.

Financial Disclosure:

-