Purchase this article with an account.
JM Hitt, FA Dodge, RB Barlow; Computation Modeling of Night Vision in Limulus . Invest. Ophthalmol. Vis. Sci. 2002;43(13):1370.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: Can a cell-based computational model of the Limulus lateral eye accurately predict optic nerve responses at night? Method: We recorded the activity of single optic nerve fibers in response to uniform illumination and modulated spots and sine waves. We controlled the sensitivity of the retina by exposing the eye to light during the day and simulating the nighttime state by electrical stimulation of the optic nerve efferent fibers. We converted our cell-based daytime retinal model into a nighttime model by incorporating the known circadian changes in retinal physiology. Fitting the model by power spectral analysis of optic nerve activity, we estimated the effective photon absorption rate. Results: Optic nerve activity under ambient daylight illumination (20 cd/m2) had a mean rate of 10 ips, a coefficient of variation (CV) of ∼0.05, and a broad power spectrum (half width ≷ 5 Hz). Shortly (2 to 5 min) after decreasing the level of illumination to near nighttime levels (2x10-3 cd/m2) optic nerve activity decreased to 4 ips, CV increased to 0.5, the power spectrum narrowed, and the dynamic response shifted (-π/3). After 30 minutes of adaptation to nighttime illumination, firing rate increased to 6 ips with no additional change in CV, power spectrum, or dynamic response, suggesting no change in photon absorption rate. We transformed the eye to the highly sensitive nighttime state and continued to record optic nerve activity under nighttime illumination levels. Under those conditions, the firing rate increased to 10 ips (equal to the daytime rate), the CV decreased to 0.3, the power spectrum narrowed (half width < 2 Hz), and the response phase lag increased (-2π/3 at 1 Hz). By incorporating increased photon catch and gain and decreased inhibition the computational model accurately predicts mean rate, variance, power spectra, as well as demonstrated the effects of inhibition on response phase. Conclusion: A cell-based model of the Limulus lateral eye can accurately predict the response properties of the eye day and night.
This PDF is available to Subscribers Only