Abstract
Purpose:
Recent advances in optical coherence tomography (OCT) technology have led to the expansion of this technology into all areas of medicine. A large number of clinical studies have been conducted in areas such as glaucoma, macular degeneration, Alzheimer's disease, and others. In these studies patients with different refractive errors are studied with different OCT devices using a variety of imaging protocols that measure retinal thickness. Our purpose was to study the lateral and axial magnification characteristics in six different optical coherence tomography devices in a model eye and draw conclusions for clinical studies.
Methods:
We build an eye model with a simulated retina and an achromatic doublet lens. The axial length of the eye model was adjustable to yield a corresponding defocus of +4D to -10D from emmetropia. We imaged the model eye with six different optical coherence tomographs (CirrusOCT, Optos OCT, Optovue, SpectralisOCT, StratusOCT, and Topcon DRI). We measured the lateral dimension of features on the simulated retina with each instrument. Additionally, we measured the axial dimension of a microscope cover glass.
Results:
In the first study we analyzed lateral magnification. The SpectralisOCT was the only telecentric imaging system and showed no magnification error with axial length. The Topcon DRI and Optovue showed 12% and 13% variability over the defocus range, respectively. The CirrusOCT and StratusOCT showed 30% variability and the Optos OCT showed 45% variability over the defocus range. In the second study we analyzed axial magnification error. None of the instruments showed any axial magnification error.
Conclusions:
Optical coherence tomography has dramatically altered retinal imaging and anatomical understanding. Recently, the instruments have been used to collect cross-sectional normative databases in large number of patients. These studies rely on the accuracy of the instrument for different refractive powers and different axial lengths. It appears that several instruments exhibit lateral magnification errors that need to be considered in cross-sectional studies. Our studies indicate that none of the instruments exhibit axial magnification error.