Photopic spatial summation in the central retina assessed with adaptive optics

Abstract

Purpose : Anatomical and physiological evidence suggests that spatial pooling of cone signals in the primate fovea is minimal, but linking these findings to their perceptual correlates in humans is challenging, in part due to uncertainties introduced by optical aberrations and fixational eye movements. We used an adaptive optics scanning light ophthalmoscope (AOSLO) with high-speed retinal tracking to bypass these obstacles and examine the spatial characteristics of photopic signal integration in the central retina.

Methods : A multi-wavelength AOSLO was used to image the retina and deliver stimuli to three targeted locations along the temporal meridian between the fovea and 3.25 degrees. Three subjects with no known retinal pathology participated. Increment thresholds were assessed using monochromatic (λ = 550 nm) circular stimuli of varying diameter (0.4 to 9.3 arcmin) presented against a dim red background (8 cd/m²). At each size and eccentricity, thresholds were measured three times using 20-trial adaptive staircases. Parallel full-frame imaging in the infrared (λ = 790 nm) and stimulus channels enabled compensation for transverse chromatic aberration. Average log thresholds were plotted against log stimulus area and fit using a two-segment linear regression with the two slopes constrained to be -1 and 0; the spatial summation area was taken as the intersection of the two segments. When image resolution permitted, the number of cones comprising the summation area was counted directly from the AOSLO images.

Results : In the fovea, the summation area was 4.2 ± 0.7 arcmin² (mean ± 1 SD). Extant histological data (Curcio et al, 1990) indicate that this area corresponds to approximately 19 cones. At 2 and 3.25 degrees, summation areas increased to 7.8 ± 1.5 and 32.1 ± 11.2 arcmin². We determined directly from our AO images that, for our subjects, these areas correspond to 11 ± 1.5 and 24 ± 8.4 cones.

Conclusions : The spatial summation area for detection was measured in the central retina using diffraction-limited, retinally-stabilized stimuli. In the fovea, the majority of ganglion cells belong to the midget class and are thought to be driven by a single cone; however, the neural summation area in this region spans multiple cones, suggesting that either our stimulation paradigm favors detection by a coarser retinogeniculate pathway (e.g. magnocellular) or that additional summation occurs post-retinally.

This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.