Repeatability and reproducibility of retinal thickness measurements with spectral-domain optical coherence tomography...

To evaluate repeatability (test-retest) and reproducibility of retinal thickness measurements using spectral-domain optical coherence tomography with different densities of A-scans per B-scan and different frames per B-scan for real-time averaging.

Twelve healthy subjects were analyzed by spectral-domain optical coherence tomography. Raster lines analysis with 19 B-scans over the examined area centered on the macula was performed. Images were acquired both in standard-density (768 A-scans/B-scan) and high-density (1,536 A-scans/B-scan) modalities. Moreover, images were acquired using 2 and 20 frames/B-scan for real-time averaging. Each analysis was repeated twice to test for repeatability.

Intersession repeatability was good for all studied analysis protocols, with Lin concordance correlation coefficient values ranging between 0.88 and 1.00. Reproducibility assessment showed consistent retinal thickness measurements using variable scan density, with Bland-Altman limits of agreement of -6 μm to 6 μm in the central area. Reproducibility assessment showed consistent retinal thickness measurements using different number of frames used in the real-time averaging process, with Bland-Altman limits of agreement of -8 μm to 4 μm in the central area.

Spectral-domain optical coherence tomography is a very reliable tool for central retinal thickness assessment. Changes in the number of A-scans/B-scan and in frames used for real-time averaging do not affect repeatability and reproducibility.