Abstract
Purpose: :
Adaptive optics (AO) retinal imaging tools enable the monitoring of individual photoreceptors over time. A feature of photoreceptors is that their reflectance varies over time, on scales ranging from seconds to hours. Light-evoked changes in photoreceptor reflectance have also been described. It has been suggested that such reflectance changes might be used to monitor photoreceptor structure and function, however the repeatability of these measurements has not been sufficiently characterized. Here we examined the repeatability of cone photoreceptor reflectance changes in the human retina.
Methods: :
To assess the repeatability of temporal variations in cone reflectance, 2 subjects were imaged using an AO scanning light ophthalmoscope (AOSLO). Images of the cones (0.7° from fixation) were obtained using a 775nm light source (13.4 μm coherence length) every ten minutes for two hours. The experiment was repeated within three days for both subjects. To assess the repeatability of light-evoked changes in photoreceptor reflectance, one subject was dark adapted then imaged using an AO flood-illumination camera at two separate time points. Videos of the cones (2.5° from fixation) were acquired at 167Hz for 0.33s using an 837nm light source (15.4 μm coherence length). After 0.10s, a 650nm stimulus was presented. Reflectance profiles for each cone were generated by plotting the reflectance of that cell as a function of time. The reflectance profiles for each cone were compared between the 2 imaging sessions within each experiment, using cross correlation and profile archetyping. A total of 1,565 and 928 cones were analyzed for both subjects in the temporal experiment and 1,224 cones were analyzed in the stimulus-evoked experiment.
Results: :
A cross correlation analysis between reflectance profiles in the temporal experiment showed 68.5% and 67.1% of the cones for the two subjects, respectively correlated. When using an archetype analysis to describe the temporal reflectance behavior of the cones, we found that 70.4% and 69.9% of the profile archetypes were conserved between imaging sessions for the two subjects, respectively. A cross correlation analysis between reflectance profiles in the stimulus-evoked experiment showed that 50.5% of the profiles correlated, while 70.0% of the profile archetypes were conserved between imaging sessions.
Conclusions: :
Cone reflectance behavior appears moderately repeatable for the majority of cells analyzed. The amplitude of these reflectance changes can be quite large, which confounds utilization of this signal to probe photoreceptor function. Whether or not reflectance signals elicited under different experimental conditions show similar repeatability remains to be determined.
Keywords: photoreceptors • imaging/image analysis: non-clinical • retina: distal (photoreceptors, horizontal cells, bipolar cells)