Fig 1.
(a) Binocular fusion produces the sensation that a disc, Q, is floating in front of the background plane, P. The screen was 300mm away from nasion of participants. RDS with different disparities used in the experiment: for (b), the disparity was 1.5° and for (c) the disparity was 0.1°. Compared with (c), the depth of focus of the stereoscopic vision is lager for (b).
Fig 2.
Experimental design and distribution of optodes.
(a) The optodes covered on the scalp. (b) Task design following a slow event-related paradigm; each trial was composed of: a fixation (0.7s), RDS viewing and maintenance (14s), black screen (14s), subjective assessment (the duration depended on the participants’ response) and a shut-eye rest (13.3s). (c) The estimated cortical locations of 31 fNIRS channels. (d) The estimated cortical locations of the 20 optodes.
Fig 3.
Illustration of the data processing.
(a): a bad trial sample with large background noise; (b): extracted brain activity from the raw data; (c): fitting the data by two normal distribution functions.
Fig 4.
Statistical results of subjective assessments.
Correlations between subjective assessments and different binocular disparities (a, b, c, d). The evaluation of sustainability is correlated with the perception of stereopsis (e) and the degree of discomfort (f). Subsequent post-hoc pairwise comparisons (with FDR controlled) reveal that the differences of evaluations are all significant or marginally significant (Pmax = 0.052) among different disparities in (a), (b) and (d). The evaluation of RDS with 0.7° disparity is significantly larger than RDS with 1.1° in (c). *: Significance at the 0.05 level. **: Significance at the 0.01 level.
Fig 5.
Haemodynamic response and beta values.
(a): Comparison of HbO and Hb response to stereopsis. Averaged HbO signals (red solid curve) show larger magnitude and thus better sensitivity. (b): The topography of averaged beta values for 8 disparities from all participants (n = 11). The topography shows that the occipital cortex is spatially correlated with stereoscopic vision and that the activation pattern is associated with eye dominance.
Table 1.
The MNI coordinates of channels in two ROIs.
Fig 6.
Haemodynamic changes in both ROIs.
Haemodynamic changes in the left ROI (blue bar) and right ROI (orange bar) for RDSs with 8 disparities for all participants (n = 11). Error bars represent the standard error of the mean across all participants. Statistical analysis indicates that there is a left lateralization of the activation pattern; furthermore, haemodynamic response to an RDS with 0.5° disparity is significantly stronger than to an RDS with 1.1°.
Table 2.
The mean and standard deviation of haemodynamic change in different ROIs (mean ± standard deviation).
Table 3.
The correlation between subjective assessments and fNIRS data at group level (n = 11).
Fig 7.
Comparison between different baseline correction methods.
Good (a) and bad (b) trial samples after preprocessing of fNIRS data. Solid cyan curves: the time window for RDSs viewing and black screen (28s for each trial). Black curves: subjective assessments and shut-eye rests (13.3s). The averaged response of good (c) and bad (e) trial samples using traditional time-course analysis (zero order baseline corrections). The averaged response of good (d) and bad (f) trial samples using curvilinear fitting baseline corrections.