A physical model eye was constructed to test the quality of ophthalmic instruments. The accuracy and precision of two commercially available instruments were analysed. For these instruments, a particular model eye was obtained which mimicked the physical properties that would be usually measured e.g. corneal topography or optical path within the human eye. The model eye was designed using relatively simple optical components (e.g. plano-convex lenses) separated by appropriate intraocular distances taken from the literature. The dimensions of the model eye were known a priori: The lenses used in the construction of the model eye were characterised ac cording to values given in the manufacturers' data sheets and also through measurement using an interferometer. The distances between the lens surfaces were calculated using the interferometric data with reverse ray-tracing. Optical paths were calculated as the product of refractive index and axial distance. The errors inherent in mea suring these ocular parameters by different ophthalmic instruments can be considered as producing an erroneous value for the overall refractive power of the eye. The latter is a useful metric for comparing various ophthalmic devices where the direct comparison of quality is not possible or is not practical. For example, a 1% error in anterior corneal radius of curvature will have a more detrimental effect than the same error in posterior corneal radius, due to the relative differences in refractive indices at those surface boundaries. To quantify the error in ocular refractive power, a generic eye model was created in ZEMAX optical design software. The parametric errors were then used to compute the overall error in predicting ocular refractive power, thus highlighting the relative importance of individual errors. This work will help in future determination of acceptable levels of metrological errors in ocular instrumentation.
|