Prediction of contrast sensitivity in the presence of glare

Abstract

Purpose : The objective of this study is to introduce a model which uses the measured contrast sensitivity function (CSF) without a glare source to predict the CSF in the presence of a glare source.

Methods : The CSF was measured in 100 trials with the quick CSF method at three different mean luminance levels (48, 42 and 33.6 cd/m²) with and without a glare source on five healthy subjects. The different luminance levels were obtained using calibrated photographic filters. The position of the glare source was 2.5 degrees away from the contrast stimuli and the illuminance of the glare source was 12 lux at pupil of the eye. The area under the logarithm of the CSF curve (AULCSF) was used as outcome parameter. Furthermore, the stray light parameter at an angle of 2.5 degrees was measured. The reduction of CSF with a glare source was predicted from the measured CSF without a glare source through the calculation of a factor defined as the mean luminance of the contrast test divided by the sum of the mean luminance of the contrast test and the veiling luminance induced by the glare source. The veiling luminance was determined by the stray light parameter of the subjects at 2.5 degrees, the strength of the glare source and the angular position of the glare source with respect to the contrast test. The predicted AULCSF with glare source was compared to the measured AULCSF with a glare source. The found difference in AULCSF was compared to the precision of the CSF test (0.1 AULCSF units) to assess the quality of the prediction.

Results : The average measured stray light parameter of the subjects was 1.1 log(s). The measured AULCSF ranged from 2.0 to 2.4 and from 1.8 to 2.1 AULCSF units for the measurements without and with a glare source, respectively. The differences between the model prediction and measured AULCSF for the luminance levels 48, 42 and 33.6 cd/m² were 0.05, 0.05 and 0.03 AULCSF units respectively. The prediction error was within the precision of an individual contrast sensitivity measurement using 100 trials.

Conclusions : The described prediction model is capable of estimating the CSF with glare based on the measured contrast sensitivity function without glare for the given subset of five healthy subjects.

This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.