Abstract
Purpose:Visual performance in the mesopic range is affected by changes in the spatial and temporal properties of the retina and / or changes in the quality of the retinal image as a result of increased aberrations and scattered light. The aim of the study was to establish the extent to which retinal and / or optical factors set the limits of visual performance at low light levels and to develop a test of visual performance in the mesopic range that reflects the quality of the optics of the eye, when the pupil size is large, and not the limits of resolution of the retina.
Methods: A closed-loop system was developed capable of maintaining a constant retinal illuminance, independent of pupil size. The experiments were carried out on a calibrated visual display using spectrally calibrated neutral density filters. A new version of the P_SCAN 100 system (Clinical Vision Science, 2, 131–141, 1987) was used to measure the area of the pupil every 20 ms. This signal was low-pass filtered and then used to adjust the luminance of the display monitor to maintain the specified retinal illuminance. The retinal illuminances investigated were in the range 2 to −1 log trolands. Pupil size, ocular aberrations, chromatic sensitivity and “functional” contrast acuity (ASEM, 74,551–559, 2003) were measured as at a number of discrete retinal illuminances in this range. Retinal image quality was evaluated using a WASCA wavefront analyser (Wavefront Sciences). The rms wavefront aberration was computed for the subject's mean pupil size at each light level investigated.
Results: Measurements of scattered light for a number of different pupil sizes show that the scattering of light is non-uniform over the pupil with the far periphery contributing more scattered light than the centre region. The rms wavefront aberrations were also found to increase rapidly with decreasing light level in the mesopic range. This was paralleled by a massive increase in contrast acuity thresholds with decreasing light level and a rapid loss of both red-green and blue-yellow chromatic sensitivity. There was, however, little correlation between rms wavefront aberration and contrast acuity thresholds in the low mesopic range suggesting that, at least in the subjects investigated, the limit of visual performance over most of the range was set by the poor resolving power of the retina and not the optics of the eye. These data were used to design a new test for contrast acuity assessment that reflects changes in retinal image quality as a result of increased aberrations and scattered light (when the pupil size is large ∼ > 5.5 mm). The stimulus configuration was optimised to ensure that the pupil size remained large without compromising too much the resolving power of the retina.
Conclusions: The size of the pupil affects contrast acuity thresholds, significantly, in at least two different ways. For stimulus conditions that generate a large pupil size, but where vision is still dominated by cone photoreceptors signals, increased aberrations and scattered light can cause a large reduction in contrast acuity. We propose a new test that makes it possible to assess mesopic visual performance at large pupil sizes when vision is not limited by the resolving power of the retina. These restricted range of stimulus conditions benefits from reduced higher order aberrations at large pupil size since the limiting factor is the quality of the retinal image. In the low mesopic and scotopic ranges vision is dominated by rod signals and contrast acuity thresholds are much increased. The limits of visual performance are now largely determined by the resolving power of the retina. Visual performance in this range is strongly affected by even small changes in retinal illuminance caused by inter subject differences or within subject fluctuations in pupil size.