The role of melanopsin activation in peripheral color matching functions

Abstract

Trichromatic color theory is based on the notion that only three types of cones (L, M and S) mediate foveal color vision in photopic lighting levels. Consequently, three primaries are sufficient to match any spectral colors. Cone fundamentals were obtained by linear transformations of traditional color matching functions (CMFs), for example, 10° r ̅,g ̅,b ̅ from Stiles and Burch (Optica Acta, 1959). However, since the discovery of melanopsin expressing intrinsic photosensitive retinal ganglion cells (ipRGCs) in the peripheral retina, several works have demonstrated the involvement of the melanopsin photopigment in color vision. The goal of this study was to assess the intrusion of melanopsin activation in peripheral CMFs.We obtained CMFs for a set of monochromatic test lights (470nm-630nm) in photopic levels (~1780Td), which were matched with three intensity-adjustable LED primaries [peak wavelengths: 460nm (R), 520 nm (G) and 630 nm (B)]. Both set of lights (test and matching) were presented using a Maxwellian view optical arrangement and were foveally-fixated. The measurements were conducted in three spatial conditions: 1) ?peripheral? (10°-26° annular field); 2) ?large central? (10° circular field); 3) ?small central? (1.7° circular field). The test and matching stimuli were presented sequentially over time, with three seconds each. Observers adjusted primary lights to match the monochromatic test light color. Each observer repeated at least five matches for each test wavelength and stimulus condition. CMFs were normalized allowing for comparison with Stiles and Burch?s CMFs. Our results showed that the three spatial conditions had similar R matches. On the other hand, the small central condition had higher B matches than the large central and peripheral conditions, particularly at 480nm to 510nm. The small central condition had lower G matches than the peripheral condition at 490nm and 500nm. A control experiment aimed to rule out rod intrusion showed that bleaching did not change peripheral results for most of the test wavelengths, except at 490nm for the three matches and at 520nm for G and R matches. The peripheral CMFs were in a good agreement with Stiles and Burch?s CMFs (R match: r2 = 0.92; Green match: r2 = 0.82; B match: r2 = 0.97). Statistical modeling indicated that the inclusion of rod and melanopsin with foveal cone fundamentals (yielded from small central CMFs) help to improve the predictions of peripheral S- and M- but not L-cone sensitivity functions (yielded from peripheral CMFs). Similar results were obtained for CIE CMFs.CMFs for periphery are similar than CMFs for large central field but different for small central field CMFs. Since ipRGCs and rods are present in the periphery and absent in fovea, it is likely that CIE 10° color matching function involves both melanopsin and rod intrusions. Based on bleaching experiment, the difference between peripheral and 2° CMFs could be accounted mostly by melanopsin intrusion.