A model of cone photoreceptor reflectivity

Abstract

Purpose : Adaptive optics scanning laser ophthalmoscopy (AOSLO) is used to image photoreceptors in vivo. Image formation was modelled using the finite difference beam propagation method (FDBPM). The model gives a possible explanation for the variation of cone reflectivity seen in AOLSO images.

Methods : Using FDBPM, light was propagated down a single simplified cone receptor, as we previously presented (ARVO 2017). The cone model consisted of 3 segments: myoid, ellipsoid, and outer segment. Where there is a transition from one refractive index to another, the reflected field was calculated using the Fresnel equations and propagated back (Fig. 1). The returned fields were then used to calculate the power falling on a detector via convolution. As there is evidence from OCT that junctions at the inner/outer segment interface (IS/OS) and at the OS tip consist of multiple reflectors, small extracellular gaps at these locations were built into the model. Interference effects between reflections were also included. By moving the input beam across the cone, the scanning of an AOSLO system is mimicked and en-face cone images could be created. The model was also used to calculate AO-OCT intensity profiles.
The distribution of modelled cone reflectivity was compared to a distribution measured in vivo, captured during 325 psychophysical threshold tests in 3 subjects. Modelled images were generated using randomly selected gap sizes for each cone drawn from gap estimates (Jonnal et al., 2017).

Results : The modelled OCT profile had broadened widths for the IS/OS and OS tip reflections and had typical relative amplitudes. Variation in modelled cone reflectivity was similar to the values seen in AOSLO images (Fig. 2). The intensity profile of each cone’s reflection appeared Gaussian, like those imaged invivo.

Conclusions : FDBPM modelling of cone photoreceptor reflectivity is consistent with observations in both OCT and AOLSO. The agreement occurs only if multiple reflections at the IS/OS and OS tip are present and interference effects between those reflections are taken into account.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.

 

Fig. 1. Normalized electric field amplitudes at the external limiting membrane (top) and along the length of a photoreceptor (bottom) for an incoming beam and for reflections off two surfaces.

 

Fig. 2. (A) Normalized distributions of modelled and AOSLO cone reflectivities. (B) Modelled AOSLO image. (C) AOSLO image of a cone mosaic from a normal human subject at 1.5° eccentricity.