When changes in ambient light level are rapid or of brief duration, the human visual system has the remarkable ability to adjust its sensitivity within seconds or tenths of a second rather than minutes. This type of adaptation, which is common in everyday experience (eg, the headlights on oncoming vehicles at night), is known variably as early light adaptation, early dark adaptation, transient disability glare (and recovery), or simply transient adaptation. In previous research, we measured loss in sensitivity at 2-time intervals, following the onset of a brief-duration, peripheral glare source that was intended to simulate the transient glare of oncoming headlights.

• Objective 1 was to determine detection thresholds in young and elderly individuals at each of several times following the onset of brief duration, fixed-ratio increments, and decrements in luminance relative to 2 different pre-adapted luminance levels.
• Objective 2 was to measure the time course of early dark adaptation using light levels sufficient to bleach 1% to 2% and 20% of the cone photopigment in normal young subjects and in 2 groups of elderly subjects, one with and one without early age-related macular degeneration (ARMD).

Specifically, the highest light level used to achieve Objective 1 represents a relatively low photopic level (100 c/m2). As such, there will be little bleaching of cone photopigment, and a critical question is whether a significant bleaching of photopigment is necessary to demonstrate an age-related and/or a disease-related difference in recovery rates during early dark adaptation.

• A comparison of threshold changes in response to luminance increments for the different age groups will provide an indication of the repeatability of previous studies of early light adaptation and our previous work on early glare adaptation.
• A comparison of threshold changes in response to luminance decrements for the different age groups will provide an indication of whether recovery times during early dark adaptation are different for the different age groups and the extent to which this may depend on absolute luminance level.

These results will also indicate whether the key factor determining visibility loss under transient conditions is indeed the ratio of change from one luminance level to another.

The purpose of this study is to test the feasibility of using a novel Degraded Texture Discrimination Test to detect the early neural dysfunctions caused by retinal diseases. Human subjects possess the capability of discriminating certain types of textures instantaneously and effortlessly. This ability depends on the correct registration of any conspicuous local features called textons. The types and spatial densities of textons in different regions determine whether these regions are perceived as being covered with different textures. When retinal neural cells degenerate, the types and densities of textons that the brain receives are altered. Consequentially, a patient’s ability to discriminate textures is impaired. Therefore, texture discrimination is suitable for detecting diffuse degeneration of individual neurons across a retina. However, because of a compensation mechanism of the visual system, a performance deficiency in discriminating intact textures may not be observed until a large proportion of retinal neurons have been lost. This is one of the reasons that clinicians usually fail to detect, and patients themselves fail to notice, the beginning of retinal neuron degeneration.

To increase the sensitivity of the texture discrimination test, texture stimuli will be degraded so that the loss of information in the stimuli will compound with the loss of information in a diseased retina to produce a magnified impact on the performance. The effect of a moderate amount of retinal neuron loss of vision can be magnified by the degradation of the stimulus so that it can be observed.

In our test, a computer generates two similarly degraded texture patches and displays them side-by-side for a short duration. These texture patches may or may not have built-in structural differences between them. The patient’s task is to use 2 computer keys to indicate whether a structural difference is perceived. Performances at several levels of texture degradation are measured. Comparison of patient performances with normal subject performances yields a quantitative estimation of the proportion of information lost on the diseased retina, data that can be derived through a simple mathematical formula. This quantity is proportional to the neural degeneration on the diseased retina and, thus, has important clinical value in detecting the retinal diseases, monitoring progression of retinal diseases, and evaluating the effect of therapeutic procedures.