Purpose:
To compare measurements of retinal thickness using two optical coherence tomography (OCT) instruments: Cirrus HD-OCT and Stratus OCT (both from Carl Zeiss Meditec, Dublin, CA).
Methods:
A total of 148 subjects from 2 sites with one Cirrus HD-OCT 200x200 macular cube scan and one Stratus OCT Fast Macular Thickness Scan, both acquired during the same visit, were evaluated. The study eye was selected randomly except when only one eye qualified for enrollment. Subjects were grouped by their retinal condition. The retinal thicknesses in all 9 corresponding subfields of both types of scans were compared.
Results:
The Stratus and Cirrus retinal thickness subfield averages were well correlated. They could be fit to a simple linear model with slope of unity (R² = 0.94). The intercept (average difference between Cirrus and Stratus) and the standard deviation (SD) of the residuals varied depending on the specific pathology. We estimate the expected repeatability limits of this model, assuming a normal distribution of residuals, as square root(2)*1.96*SD = 2.77 * SD. In a clinical setting, if the value changes by more than the repeatability limit value, there is 95% confidence that the change is due to more than variability due to instrument-to-instrument conversion.
Conclusions:
The difference between Stratus and Cirrus retinal thickness measurements is dependent upon the type of retinal pathology. It should therefore be possible to use a conversion factor specific to certain disease groups. Differences in residuals from this conversion should be taken into account when evaluating significance of change for individual cases.
Keywords: retina • imaging/image analysis: clinical • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)