Abstract
Purpose:
Optical coherence tomography (OCT) has become a standard in ophthalmic imaging. Many large clinical practices routinely use multiple instruments from different manufacturers for following patients. Clinical trials compare quantitative retinal measurements despite the use of multiple instruments from different manufacturers. In this study we assess the agreement on lateral and axial measurements performed with spectral domain OCT across instruments from a single manufacturer and across multiple manufacturers.
Methods:
A commercially available model eye (Rowe Technical Design, Orange County, CA) was designed with circular tissue phantom with the following dimensions: 4.8mm lateral diameter, 0.18-mm central thickness, and 0.30-mm peripheral thickness. The Model Eye was imaged on 2 instruments each from 4 platforms (Zeiss Cirrus HD-OCT, Heidelberg Spectralis, Bioptigen SDOIS, and hand-held Bioptigen Envisu). Two graders used built-in software from each instrument to perform manual measurements of the tissue phantom lateral diameter, central foveal thickness, and thickness at 1 mm from the foveal center. Intragrader and intra/inter-manufacturer reproducibility were assessed.
Results:
There were no significant differences between graders. The range between different manufacturer devices was 5171-5465 μm for lateral width, 161-196 μm for central thickness and 268-316 μm for 1mm from the center. The difference between instruments from the same manufacturer ranged from 8.3-14.2 μm for lateral width, 3.3 - 9.6 μm for central thickness and 3.0 - 9.1 μm at 1mm from the center. All differences between instruments from the same manufacturer were statistically significant using a paired t test (p = 0.042 - p<0.0001). The inter-manufacturer differences between instruments were statistically significant for all measurements (p<0.001, ANOVA).
Conclusions:
The differences between quantitative measurements of the same retinal dimensions were smaller between devices from the same manufacturer than between devices from different manufacturers, but all differences were significant. These differences should be considered when comparing patient data between visits and across different instruments.
Keywords: 552 imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) •
630 optical properties •
550 imaging/image analysis: clinical