Purchase this article with an account.
Austin Roorda, Pavan Tiruveedhula, Qiang Yang, David W. Arathorn, Joseph Carroll; Measurements of retinal sensitivity on a cellular scale. Journal of Vision 2009;9(14):82. doi: 10.1167/9.14.82.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: To use the adaptive optics (AO) retinal microstimulator to map the sensitivity with the resolution of single cones in a human retina whose trichromatic cone mosaic has been mapped.
Methods: Real time AO imaging of the cone mosaic at IR wavelengths allows for eye motion tracking and facilitates visible light delivery to targeted retinal locations. This particular experiment was performed in a region of one subject's retina where the spectral identity (L, M, or S) of the cones had been previously mapped at the University of Rochester using spatially localized retinal densitometry (Hofer et al., J Neurosci. 2005). The protocol involved measuring the frequency-of-seeing (FOS) of an AO-corrected spot delivered to a grid of targeted retinal locations across the cone mosaic. The stimulus wavelength was 680 nm, which had a relative ratio of L:M:S cone sensitivity of 1:0.084:∼0. The power was set so that the overall FOS was about 53%. Careful correction of longitudinal chromatic aberration and measurement of transverse chromatic aberration was done throughout the experiment.
Results: After 4854 trials we generated a FOS map that spanned about 113 cones. A minimum of ten YES or NO responses in a binned region spanning 3X3 image pixels (0.65 arcmin) were required to confidently plot the FOS. A correlation of the FOS by cone type showed significant differences in sensitivity between the L, M and S cone classes, but these are much less extreme than the predictions because of blurring by optics and the extent of the cone aperture. To explore the correlation further, we generated a map of expected sensitivity across the mosaic considering blurring by the optics (diffraction), cone acceptance apertures (MacLeod, Williams & Makous, Vision Res. 1992) and L:M:S sensitivity. Pixel-by-pixel comparison of the expected sensitivity vs FOS was highly significant (p < 0.000001), but not very deterministic as to the cone type (R2 = 0.08).
Conclusions: The AO retinal microstimulator can measure retinal sensitivity at targeted retinal locations on the scale of a single cone. Experimental retinal sensitivity correlates with the L, M and S cone types that tile the retina.
This PDF is available to Subscribers Only