Testing melanopsin and rod intrusion in large-field cone fundamentals: A statistical approach

BARRIONUEVO, PABLO A.; SANDOVAL SALINAS, MARÍA LEONOR; CAO, DINGCAI

Heraklion

Congreso; 26th Symposium of the International Colour Vision Society; 2022

International Colour Vision Society

In the CIE (2006) recommendations, the differences between 2o and 10o cone fundamentals (obtained at corneal level) are explained by optical density differences due to macular pigment and cone photopigments. However, since the discovery of melanopsin expressing cells in the human retina, evidence of their intrusion in brightness and color vision for extrafoveal stimuli is increasing. Furthermore, rod contribution to color vision has been largely shown. The objective of this study was to assess the intrusion of melanopsin and rods in large-field L, M, and S cone fundamentals. We predicted 10o fundamentals values based on 2o fundamentals values to assess whether rod and/or melanopsin contribution are necessary for a better prediction. Four nested regression models were tested: rods + melanopsin + cones (M1), melanopsin + cones (M2), rods + cones (M3), and only cones (M4). The models were assessed with likelihood ratio tests for comparison of model fits. Analyses were conducted considering CIE and Smith & Pokorny’s (1975, S&P) fundamentals. Also, this approach was implemented considering own experimental color matching data for large- and small-fields, and incorporating the optical filtering proposed by CIE recommendations to the computation. For S fundamental, M1 model was better (p < 0.001) than M2, M3, and M4; M2 was better than M4 (p < 0.05); and M3 was similar than M4 (p = 0.52, CIE; p = 0.35, S&P). For M fundamental, M1 was better than M2 and M4 (p < 0.001), and similar (p = 0.06, CIE) or better (p < 0.001, S&P) than M3. For L fundamental, M1 was similar than M2 (p = 0.43, CIE; p = 0.09, S&P), better than M4 (p ¡ 0.001, CIE and S&P), and better (p < 0.05, CIE) or similar (p = 0.43, S&P) than M3. Incorporation of optical filtering effect to experimental data didn’t explain completely the differences between small- and largefield conditions. Instead, incorporation of the filters effect in addition with rods and melanopsin produced a better fit. The difference between 10o and 2o S fundamentals is explained by a combination of rods and melanopsin, for both CIE and S&P curves; also, the inclusion of melanopsin to the only cone model improved the fitting but the inclusion of rods did not. Less agreement between CIE and S&P sets was obtained for M and L fundamentals. This approach suggests that melanopsin and rods may play a role in the large field sensitivity functions, especially for the S fundamental.