Abstract
The human fovea is equipped with neural mechanisms whose spatial grain is finer than the spatial and temporal blur introduced into the retinal image by optical aberrations and fixational eye motion, respectively. As a consequence, pre-retinal factors have the potential to play a significant role in shaping visual performance in the central retina. In this presentation, I will discuss experiments in which the spatial summation of light increments delivered to the human fovea was examined using a custom adaptive optics scanning laser ophthalmoscope that provided precise control over high-order ocular aberrations and stimulus motion on the retina. We found that these pre-retinal factors had relatively little impact on behavioral measurements of foveal summation. When pre-retinal factors were minimized, the area over which incident photons were pooled completely (i.e., Ricco's area) had an average diameter of 2.43 arcmin, encompassing between two and three dozen foveal cones. When the measurements were repeated in the presence of habitual aberrations and natural stimulus motions on the retina, Ricco's area did not increase significantly. Our results obtained across conditions were in good agreement with predictions generated by a computational observer that included a physiologically-realistic model of the front-end of the visual system coupled with a post-receptoral summing filter.