Spatio-temporal Variation of Binocular Contrast Summation

Purpose: This study reports on the dependence of binocular contrast summation on the spatio-temporal properties of the stimulus in normal human visual systems. Methods: Monocular and binocular contrast detection thresholds were measured in 6 subjects with normal binocular function. Stimuli were Gabor (2.13 x 2.13 degrees) carrier gratings windowed with a 0.75 sigma Gaussian envelope. Contrast detection thresholds were measured for 3 spatial frequencies (0.5, 3, 9cpd) and 3 temporal frequencies (2, 8 and 16Hz) using a two-down-one-up descending staircase procedure presented with a 2 Interval Forced Choice method. Threshold was calculated as the mean of the last 6 reversals. Binocular summation ratio was calculated as the average of the monocular contrast detection threshold of each eye divided by the binocular contrast detection threshold. Results: Binocular summation ratios varied significantly across spatial frequency (F(2,267) = 5.634, p = 0.004), but not with temporal frequencies over the 3 octave range investigated in this study (F(2,267) = 1.516, p = 0.222). Mean binocular summation ratios ranged from 1.48 (+/-0.06, SEM) to 1.84 (+/-0.07, SEM). Neither the quadratic summation model nor the two-stage model in their current forms were capable of accounting for the spatial frequency dependent change in binocular summation. Conclusions: Binocular summation depends on the spatial frequency of the stimulus; being maximal for spatial frequencies about 3cpd. The quadratic summation and the two-stage models at minimum, require model parameters that should scale with the spatial frequency composition of the stimulus.