ASSESSMENT OF VISUAL FUNCTION USING LUMINANCE NOISE

Inefficiency of visual processing may be partitioned into two broad categories: unreliability of "noisiness" of the nervous system and systematic error or "distortion". The goal of the proposed research is to psychophysically characterize the visual deficit of human observers in terms of intrinsic noise and systematic error. Results from this study have the potential to localize the source of a visual problem, to produce useful diagnostic indicies and to understand the underlying causes of visual deficits. The partitioning of inefficiency of visual detection into these two components may be accomplished by the measurement of contrast sensitivity (and its rate of change) in the presence of various levels of visual noise. This technique has the following advantages over conventional contrast sensitivity measures: 1) Contrast detection in high levels of noise can effectively by-pass the optics, yielding a measure of performance which is primarily dependent on the quality of the succeeding neural machinery irrespective of its noisiness. 2) Comparison of high noise performance with low noise performance leads to an index of "noisiness" 3) High levels of noise permit the study of high contrast vision with the same methods which have been applied to low contrast vision. 4) Human performance can be measured on an absolute scale. The measure called "efficiency", specifies the fraction of the available statistical information an observer is using. In order to assess the utility of this procedure, the study will restrict itself to two classes of patients: those with primarily optic media opacities, and those with optic neuritis. It is believed that the first class may show mainly systematic errors of vision, whereas the second may show increased noisiness of information transmission.