Abstract
Purpose :
Adaptive optics (AO) observations of the human cone mosaic have shown that visible light stimuli induce changes in infrared reflectivity. Recently we demonstrated that the intrinsic reflectance response for a cone population is dependent on stimulus irradiance and wavelength and is related to cone phototransduction (Cooper et al, Biomed. Opt. Express, 2017). Here we examine whether we can measure an intrinsic reflectance response in individual cones.
Methods :
Three subjects were imaged using an AO scanning light ophthalmoscope. 1x1° videos were acquired 0.7° from the fovea with a 795nm imaging source. Each acquisition consisted of recording for 4s prestimulus, 2s while delivering a 550nm stimulus of varying retinal irradiance (0, 50, and 450nW/deg2), and 14s post-stimulus. Thirteen acquisitions (the first 3 of which were excluded to allow responses to stabilize) were obtained within a single run. Each run was preceded by 2 minutes of dark adaptation. Five runs were obtained for each stimulus irradiance, with run irradiances randomly interleaved. Every acquisition was split in half; the first and second halves were defined as stimulus and control sequences. All sequences were co-registered. Temporal reflectance signals were extracted from each cone and standardized to their pre-stimulus values. For each stimulus irradiance we calculated the standard deviation (SD) of the standardized reflectances at each time point across all stimulus sequences, subtracted the control sequences’ SD, and fit a piecewise-smooth function to the result. We took as each cone’s response the signed peak magnitude of that cone’s fit.
Results :
The average cone response for the 3 subjects was 0.08 (range 0.06-0.1) for 0nW/deg2, 0.9 (range 0.6-1.2) for 50nW/deg2, and 1.2 (range 0.8-1.5) for 450nW/deg2. Larger responses were found in 95% (range 92-97%) of cones at 450nW/deg2 compared to 0nW/deg2 and in 92% (range 90-95%) of cones at 50nW/deg2 compared to 0nW/deg2.
Conclusions :
Our measurements reveal an intrinsic reflectance response in ~95% of individual cones. It is unsurprising that ~5% of cones were unresponsive, as the stimulus wavelength does not substantially activate S-cones. The average individual cone response increases with irradiance, as would be expected for a functional signal. Our technique has the potential to enable simultaneous objective functional assessment of large numbers of individual cones.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.