Regional variation of contrast sensitivity across the retina of the achromat: sensitivity of human rod vision.

摘要

1. Detection thresholds for two-dimensional Gabor functions of varying spatial and temporal frequency were used to investigate the post-receptoral sensitivity across the retina of the typical and complete achromat. 2. Under photopic conditions there is no evidence for post-receptoral cone function at any retinal eccentricity investigated. Sensitivity saturates in a way consistent with known psychophysical and electrophysiological measures of rod saturation. This occurs in a unitary fashion across the retina. 3. Under scotopic conditions the regional fall-off in spatio-temporal sensitivity is similar for the achromat and duplex retina. This suggests that the rods in the achromat make normal neural connections. 4. Taken together this supports the contention that the typical and complete achromat is a functional rod monochromat and hence can be used to explore the sensitivity of the isolated rod post-receptoral mechanism under mesopic conditions where its sensitivity is optimal. This is where its contribution is most difficult to isolate in the duplex retina. 5. For the human rod mechanism, mesopic post-receptoral sensitivity for all spatio-temporal stimuli is optimal in the central region of the retina and falls off as a function of eccentricity. 6. For localized stimuli, peripheral spatial sensitivity is reduced evenly at all spatial frequencies compared with that of the central retina. A similar displacement of the spatial sensitivity function of the rod mechanism occurs as illuminance is reduced. 7. For localized stimuli, temporal acuity of the rod mechanism is around 20-25 Hz irrespective of retinal position. As the illuminance is further lowered dynamics of the rod pathway are reduced irrespective of retinal position and the sensitivity function maintains a bandpass shape. 8. The regional distribution sensitivity of the rod mechanism changes as illuminance is reduced from mesopic to scotopic levels.