Interocular interaction of contrast and luminance signals in human primary visual cortex

• Published in: NeuroImage (Orlando, Fla.) Fla.), 2018-02, Vol.167, p.23-30
• Main Authors: Chadnova, E., Reynaud, A., Clavagnier, S., Baker, D.H., Baillet, S., Hess, R.F.
• Format: Article
• Language: English
• Subjects: Binocular vision; Cortex (temporal); Experiments; Eye; Functional magnetic resonance imaging; Magnetoencephalography; Neuroimaging; Psychophysics; Retina; Visual cortex; Visual system
• ISSN: 1053-8119; 1095-9572
• DOI: 10.1016/j.neuroimage.2017.10.035

Interocular interaction in the visual system occurs under dichoptic conditions when contrast and luminance are imbalanced between the eyes. Human psychophysical investigations suggest that interocular interaction can be explained by a contrast normalization model. However, the neural processes that underlie such interactions are still unresolved. We set out to assess, for the first time, the proposed normalization model of interocular contrast interactions using magnetoencephalography (MEG) and to extend this model to incorporate interactions based on interocular luminance differences. We used MEG to record steady-state visual evoked responses (SSVER), and functional magnetic resonance imaging (fMRI) to obtain individual retinotopic maps that we used in combination with MEG source imaging in healthy participants. Binary noise stimuli were presented in monocular or dichoptic viewing and were frequency-tagged at 4 and 6 Hz. The contrast of the stimuli was modulated in a range between 0 and 32%. Monocularly, we reduced the luminance by placing a 1.5 ND filter over one eye in the maximal contrast condition. This ND filter reduces the mean light level by a factor of 30 without any alteration to the physical contrast. We observed in visual area V1 a monotonic increase in the magnitude of SSVERs with changes in contrast from 0 to 32%. For both eyes, dichoptic masking induced a decrease in SSVER signal power. This power decrease was well explained by the normalization model. Reducing mean luminance delayed monocular processing by approximately 38 ms in V1. The reduced luminance also decreased the masking ability of the eye under the filter. Predictions based on a temporal filtering model for the interocular luminance difference prior to the model's binocular combination stage were incorporated to update the normalization model. Our results demonstrate that the signals resulting from different contrast or luminance stimulation of the two eyes are combined in a way that can be explained by an interocular normalization model. •First MEG study to assess dichoptic masking with varying contrast and luminance.•First MEG work to quantify the cortical delay introduced by luminance reduction.•Binocular normalization model extended to include temporal domain.•Extended binocular normalization model now accounts for unequal visual input.