Accuracy of a novel ultrasound biomicroscope for measuring corneal epithelial thickness and agreement with an SD-OCT based device.

Purpose: To assess the accuracy of corneal epithelial thickness (ET) measurements in healthy and post transepithelial photorefractive keratectomy (post-T-PRK) eyes using a novel ultrasound biomicroscope (Insight 100, ArcScan, Colorado, USA) and to compare the agreement with the measurements provided by a spectral-domain optical coherence tomographer (RTVue-XR, Optovue, Fremont, CA).

Methods: In total, 32 healthy and 38 post-T-PRK eyes were included. Three consecutive measurements of right eye were acquired with the Insight 100 VHF-UBM and RTVue-XR SD-OCT. Accuracy was analyzed using intraclass correlation coefficient (ICC), test-retest repeatability (TRT), within-subject coefficient of variation (CoV), and within-subject standard deviation (S_w). Agreement was assessed by Bland-Altman analysis and paired t-tests.

Results: The precision of the normal eyes were higher than post-T-PRK eyes. Central epithelial thickness (CET) had the best repeatability and reproducibility, CoV was 0.5%∼1.71%. The S_w and TRT was 0.35µm and 0.79µm in normal eyes, 1.02µm and 1.92µm in post-T-PRK eyes. ET_superior-6mm has poor repeatability in normal eyes. However, post-T-PRK eyes have the worst repeatability at ET _inferior-3mm and ET _temporal-3mm. Agreement of CET was excellent in both groups. The widest 95% LoA range appeared at ET_superior-6mm of normal and post-T-PRK eyes, with 0.72∼8.74µm and -9.51∼7.32µm, respectively.

Conclusion: The Insight 100 provides good repeatability and reproducibility for ET measurements in normal and post-T-PRK eyes. CET can be used interchangeably between Insight 100 and RTVue-XR, however, it is not recommended for other positions.