Anterior segment optical coherence tomography scanning protocols and corneal thickness repeatability

Purpose: To examine the influence of anterior segment optical coherence tomography imaging protocols on the intraobserver and intrasession repeatability of epithelial, stromal, and total corneal thickness measurements. Methods: Repeated anterior segment optical coherence tomography (AS-OCT) images (Spectralis, Heidelberg) were obtained from 15 adults using single 8.3 mm wide horizontal line scans with an average of 2, 10, 20, 30, 50 and 100 B-scans. Volumetric scans consisting of nine 8.3 mm horizontal line scans encompassing a 1.3 mm vertical region were also captured (with 20 B-scans per line scan). Single point thickness measures (at the normal to the tangent of the anterior corneal surface) were compared with thickness measures averaged over the central 6 mm. The impact of B-scan averaging and intraobserver variability were examined for single line scans. For volumetric scans, the impact of the number of line scans upon intraobserver and intrasession variability were calculated. Results: Intraobserver repeatability did not vary significantly as a function of the number of averaged B-scans per line scan, but was lowest for 20-30 averaged B scans. For volumetric scans, increasing the number of line scans did increase scan duration (p < 0.001), with minimal impact upon the average scan quality index (p = 0.06). Averaging more than 3 line scans did not significantly improve intraobserver or intrasession repeatability for either single point or average thickness measurements. Conclusion: AS-OCT volumetric scans with 3 lines each consisting of 20 B-scans with measurements averaged over a central 6 mm of the cornea provide highly repeatable measures of epithelial, stromal and total corneal thickness (95 % LoA <= +/- 3.2 mu m for intraobserver repeatability and <= +/- 3.7 mu m for intrasession repeatability). This scanning protocol can provide reliable information when monitoring subtle changes in corneal thickness.