Abstract
Purpose:
To distinguish L-cones from M- and S-cones psychophysically in humans using an adaptive optics scanning laser ophthalmoscope (AOSLO).
Methods:
We used a multi-wavelength AOSLO to measure visual sensitivity to cone-sized stimuli under L-cone isolating conditions in two subjects with normal color vision. Increment thresholds were measured at 1.5° eccentricity, and stimulus delivery to targeted cones was achieved using high-speed retinal eye tracking. An image-based measure of transverse chromatic aberration (TCA) was obtained and used to ensure repeated stimulation of the cone of interest. The test stimulus (λ = 711 nm) was a square which subtended 27 seconds of arc (= 2.2 microns) on the retina. The background was comprised of dim infrared light from the imaging beam, light leaking through the acousto-optic modulator in the red stimulus channel, and an external 488 nm LED source viewed through a beam splitter. The luminances of the red and blue backgrounds were 2.5 and 50 cd/m2, respectively. Converting these luminances to cone contrasts, an L- versus M-cone sensitivity ratio of 24-to-1 was expected. Stimulus intensity of the red beam was controlled via 10-bit modulation and expressed as arbitrary units (au) ranging from zero to 1. Thus, for L-cone thresholds greater than 0.042 au, M- and S-cone thresholds would be predicted to lie beyond the upper modulation limit of the stimulus beam (i.e., >1.0 au).
Results:
Thresholds were measured for 88 cones in the two subjects. The stimulus was readily visible in one group of cones (n = 59; threshold = 0.55 ± 0.14SD au), and these were classified as candidate L-cones. A second group of cones (n = 18) had thresholds above 1.0 au, with even the brightest stimulus being not visible. In these cones, threshold values were normal when tested under achromatic conditions, ruling out underlying dysfunction. In the remaining cones, classification was ambiguous (n = 11). Finally, there were multiple sites in each subject where thresholds from neighboring cones yielded different spectral identities, supporting the notion that the spatial accuracy of our psychophysical apparatus is on the order of a single cone.
Conclusions:
With combined retinal eye tracking, high-order aberration correction, and TCA compensation, AOSLO can be used to identify the spectral class of single cones in vivo with a psychophysical approach.
Keywords: 471 color vision •
649 photoreceptors: visual performance