Abstract
Abstract: :
Purpose: To learn about the mechanism for reflectivity of cone photoreceptors by comparing adaptive optics scanning laser ophthalmoscope images taken with coherent vs incoherent light. Methods: A coherent laser diode (LD, 660 nm, Hitachi) and a low–coherence superluminescent diode (SLD–26–HP, Superlum, Russia, 680 nm, 9 nm bandwidth or 17 µm coherence length in the eye) were both aligned into the same SLO system. The power output at the eye was equalized for each laser and the detector quantum efficiency (H7422–20 PMT module, Hamamatsu Corp.) was nearly identical for the two wavelengths. Short video sequences of the photoreceptor mosaic were acquired and about 20 frames were aligned and added for each condition. Images were analyzed by calculating the FFT and evaluating the visibility of Yellot’s ring, which occurs at the photoreceptor packing frequency. Results: Images from the SLD provided a more faithful representation of the cone mosaic and were less affected by artifacts caused by interference between the light emitted from adjacent cones, especially for small cones near the fovea. The photoreceptor mosaic appeared more contiguous and the power spectrum showed a well–defined Yellot’s ring, similar to those obtained from incoherent flood–illuminated AO ophthalmoscopes. Conclusions: The fact that a 17 µm coherence length SLD can render the light emitted from adjacent cones mutually incoherent suggests that light reemitted from cones must undergo multiple scattering in the layer or layers posterior to the cones before getting recoupled back into the cone fiber. A simple model which relies on random–length cones reflecting at the posterior termini of their outer segments would be unrealistic, given the large random variability in cone lengths that would be required. It is recommended that low coherence light sources, such as SLDs, be used for high resolution scanning laser ophthalmoscopes.
Keywords: photoreceptors • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)