Abstract
A subset of retinal ganglion cells contains a photopigment, melanopsin, and is intrinsically photosensitive. Melanopsin is routinely described as a "non-visual" pigment, perhaps to highlight its role in functions like pupil dilation and circadian rhythms. However there is no decisive evidence as to whether light can or cannot be seen through the melanopsin pathway in healthy humans. We report measurements and analyses that estimate the influence of melanopsin absorptions on the visibility of lights presented in the healthy human peripheral retina. We ask whether light visibility in photopic conditions is better described as 3-pigment (cones only) or 4-pigment (cones and melanopsin). To analyze whether a fourth photopigment contributes to peripheral photopic visibility, it is necessary to deliver well-controlled light signals using at least four independent primaries. We built a display device capable of accurately delivering six independent primary lights (VSS 2011). Individual variability of inert lens pigments and photopigments optical density makes it difficult to specify a light absorbed by melanopsin but not by cones. Hence, we used a different approach that compares the 3- and 4-pigment predictions. A 3-pigment model predicts that the least visible stimulus (LVS) will cause no change in cone absorptions; with perfect calibration it will be invisible. In a 4-pigment model the LVS is determined by the neural pathway sensitivities and may cause cone absorptions. We collected detection thresholds to many four-primary stimuli. We could not reliably identify an invisible stimulus, but we could estimate a reliable LVS. We estimate that the LVS causes significant numbers of cone absorptions. Furthermore, the optimized 3-pigment model systematically underestimates the sensitivity of test lights designed to produce no cone absorptions. These results suggest a small but measurable peripheral sensitivity to absorptions in four, not three, types of photopigments.
Meeting abstract presented at VSS 2012