Purchase this article with an account.
S. Ortiz, S. Barbero, S. Marcos; COMPUTER SIMULATIONS OF OPTICAL COHERENCE TOMOGRAPHY A–SCANS: WHAT CAN WE LEARN ABOUT REFRACTIVE INDEX DISTRIBUTION? . Invest. Ophthalmol. Vis. Sci. 2004;45(13):2781.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: Optical Coherence Tomography (OCT) is an alternative technique to ultrasound biometry to measure intraocular distances. It relies on measurements of optical paths, and must assume the media refractive index. However, the cornea and the crystalline lens are complex structures, with several layers or gradient refractive index (GRIN) distribution. We have developed a computational method to simulate OCT A–scans. We have investigated the effects of complex index distribution on corneal and crystalline lens thickness, and under which conditions, information on refractive index can be retrieved. Methods: . A–scans from the anterior segment of a model eye were simulated in Matlab. We simulated interference patterns based on electromagnetic theory, chromatic dispersion and non–absorbent media. The ocular media were modelled as a continuum of layers (every 1 µm). We assumed Fresnel reflection and transmission of 1–D beams and considered backscattering within the inner layers. The light source bandwidth was varied between 10 and 100 nm. Corneal thickness, anterior chamber depth, lens thickness and indices of refraction were randomly varied among plausible values. We used reported values of refractive indices of corneal layers, and different models of index distribution within the crystalline lens (constant, onion–layer and polynomial GRIN). Results: For constant known values of index of refraction, the peak interdistances in the simulated A–scans reproduce the actual intraocular distances to 100% accuracy. This accuracy was pratically unchanged when more realistic index distributions were used in the simulation, while using constant average indices in the reconstructions. The peaks shape and intensity changed with light source bandwidth. A bandwidth higher than 25 nm allowed to resolve all corneal structures. Those peaks became masked by scattering noise produced by random variations within the corneal tissue equivalent to index changes in the 3rd decimal. An onion–layer model of the lens produced multiple peaks whose appearance depends on bandwidth, interlayer distance and "aliasing" effects. Lens GRIN produced a modulation in the background, with maximum values depending logarithmically of light source bandwidth. Conclusions: We developed a theoretical method to simulate realistic OCT A–scans. These simulations allow to make predictions for practical experimental configurations (light source bandwidth or delay–line resolution) for resolving corneal and lens structures, and explore the possibilities of OCT to provide information on refractive index distributions.
This PDF is available to Subscribers Only