Evaluating Repeatability Of Corneal Curvature And Thickness On Ms-39 Anterior Segment Oct Through The Measurement Range

The MS-39 (CSO, Italy) is a novel device which combines high resolution anterior segment optical coherence tomography with Placido disk topography. To accurately determine progression of ectasia, ‘machine noise’ for the measurement device must be accurately defined. The purpose of this study was to comprehensively evaluate the repeatability of the MS-39 tomographer across a large dataset for several measurement parameters.

The Early Keratoconus Clinic (EKC), Moorfields Eye Hospital, London, United Kingdom.

A retrospective study was conducted on all consecutive patients’ data from the MS-39 from September 2019 to March 2023. The mean, standard deviation within-subject (Sw) and repeatability coefficient (RC) were calculated for Kmax, K1, K2, thinnest corneal thickness (TCT) and thinnest epithelial thickness (TET). The only exclusion criteria were less than 2 repeated measurements under the same conditions (intra-observer) and/or poor-quality scans (as per manufacturer’s definition). A robust mixed-linear effects model was performed to derive within-individual variance using RStudio data analysis software. The calculated sample size required for <5% uncertainty in defining the repeatability parameters was n=768 patients.     

A total of n=16,817 patients were included in the final analysis with n=115,796 observations. The overall mean (diopters), sW and RCs were: Kmax (52.8D, 0.3D and 0.8D), K1 (44.7D, 0.2D and 0.6D), K2 (48.5D, 0.2D and 0.6D), TCT (477uM, 2.7uM and 7.8uM) and TET (42.1uM, 1.6uM and 4.8uM). The sW and RC for Kmax stratified by severity of keratoconus were: <48D (0.2D and 0.6D), 48-53D (0.25D and 0.7D), 53-58D (0.3D and 0.8D) and >58D (1D, 2.9D).

This is the first large scale powered study defining the repeatability (intra-observer) parameters on the MS-39 anterior segment optical coherence tomographer. Corneal curvatures and thicknesses taken with the MS-39 are highly repeatable. Future studies will define other parameters of clinical interest and devise strategies to further improve scan quality.