Using Functional Near Infrared Spectroscopy (fNIRS) to Study Dynamic Stereoscopic Depth Perception

Abstract

The parietal cortex has been widely implicated in the processing of depth perception by many neuroimaging studies, yet functional near infrared spectroscopy (fNIRS) has been an under-utilised tool to examine the relationship of oxy- ([HbO]) and de-oxyhaemoglobin ([HbR]) in perception. Here we examine the haemodynamic response (HDR) to the processing of induced depth stimulation using dynamic random-dot-stereograms (RDS). We used fNIRS to measure the HDR associated with depth perception in healthy young adults (n = 13, mean age 24). Using a blocked design, absolute values of [HbO] and [HbR] were recorded across parieto-occipital and occipital cortices, in response to dynamic RDS. Control and test images were identical except for the horizontal shift in pixels in the RDS that resulted in binocular disparity and induced the percept of a 3D sine wave that 'popped out' of the test stimulus. The control stimulus had zero disparity and induced a 'flat' percept. All participants had stereoacuity within normal clinical limits and successfully perceived the depth in the dynamic RDS. Results showed a significant effect of this complex visual stimulation in the right parieto-occipital cortex (p < 0.01, η(2) = 0.54). The test stimulus elicited a significant increase in [HbO] during depth perception compared to the control image (p < 0.001, 99.99 % CI [0.008-0.294]). The similarity between the two stimuli may have resulted in the HDR of the occipital cortex showing no significant increase or decrease of cerebral oxygenation levels during depth stimulation. Cerebral oxygenation measures of [HbO] confirmed the strong association of the right parieto-occipital cortex with processing depth perception. Our study demonstrates the validity of fNIRS to investigate [HbO] and [HbR] during high-level visual processing of complex stimuli.

Keywords: Binocular disparity; Depth perception; Haemodynamic response; Random dot stereogram; fNIRS.

Fig. 1

Image of the depth stimuli (dynamic RDS) to be viewed with red–green anaglyph glasses. a Shows the fused ‘flat’ control stimulus (dynamic RDS with zero disparity, perceived as ‘flat’) and b is the test stimulus (dynamic RDS with binocular disparity, induced depth percept of 3D vertical sinusoidal waves). c Sequence of experimental protocol timings between the conditions (Color figure online)

Fig. 2

Average HDR to the test stimulus (dynamic RDS with binocular disparity, induced depth percept, grey area), and the control ‘flat’ stimulus (dynamic RDS zero disparity, perceived as ‘flat’, white area) for the occipital and parietal–occipital cortices (a O1, b O2, c PO3, d PO4). [HbO] plotted in red and [HbR] in blue, mean ± SEM (Color figure online)

Fig. 3

Parieto-occipital grand average cortical responses comparing the test image during depth perception responses ([HbO] plotted in red, [HbR] in blue) to the control image at baseline ([HbO] plotted in dark grey, [HbR] in light grey). Means and SEM plotted. Significant differences between depth perception and baseline found for PO4 [HbO] only at p < 0.001 (Color figure online)

Fig. 4

Group averaged Z-transformed HDR of a [HbO] and b [HbR]. Test stimulus (dynamic RDS with binocular disparity, induced depth percept) in grey area, and control ‘flat’ stimulus (dynamic RDS, perceived as ‘flat’) in white area, for the parieto-occipital regions (PO3 black dashed line, PO4 red/blue solid line). Means and SEM plotted (Color figure online)