Takahiro Doi (1), Seiji Tanabe (2), Ichiro Fujita (3), (4)

1. Department of Neuroscience, University of Pennsylvania, Philadelphia, PA.
2. Department of Neuroscience, Albert Einstein College of Medicine, Bronx, NY.
3. Graduate School of Frontier Biosciences, Center for Information and Neural Networks, Osaka University, Suita, Osaka, Japan.
4. National Institute of Information and Communications Technology, Suita, Osaka, Japan.

The main finding of the paper by Hibbard et al. (2014) is that subjects do not perceive stereoscopic depth, reversed or otherwise, with a wide range of anticorrelated random-dot stereograms (ACRDS). What makes this finding so interesting is that conditions seem to be limited in which ACRDS gives rise to a reversed perception of depth (Doi et al., 2011; Doi et al., 2013; Read and Eagle, 2000; Tanabe et al., 2008). Thus, we are faced with the pressing question: what conditions are necessary to perceive reversed depth with ACRDS? The answer to this question will lead to important insights into the processing of stereoscopic information. Hibbard et al. discuss a number of possible conditions in their paper. We want to add two more possibilities to this discussion.

We think it is critical, if not absolutely necessary (see Read and Eagle, 2000), to interleave ACRDS with its correlated counterpart (CRDS) in the same block of trials. The reason behind this idea is the standard model of perceptual decision-making (Shadlen et al., 1996). The responses of near-preferring neurons and the responses of far-preferring neurons are compared to produce a decision variable for near/far discrimination (Doi et al., 2011; Prince et al., 2000; Shiozaki et al., 2012; Uka and DeAngelis, 2004). By interleaving CRDS trials, strong sensory signals are evoked in occasional trials. We think that these signals help to keep this task strategy fixed throughout the block. The decision circuitry uses sensory signals to learn the strategy over time (Law and Gold, 2008). It is conceivable that the sensory signal elicited with ACRDS is too weak, and that the strategy easily drifts off during the course of a block without the interleaved strong signals. Even a small drift could wipe out disparity discrimination with ACRDS, in a way similar to the data in this paper.

The CRDS trials only need to comprise a small proportion of a block in order to fix the task strategy. Reliable measurement of psychophysical thresholds can be made by interleaving weak to strong sensory signals. Three studies from our laboratory applied this task design to test stereoscopic depth perception in a variety of RDSs (e.g., interocular delay for Tanabe et al., 2008; graded anticorrelation for Doi et al., 2011 and Doi et al., 2013), and consistently found reversal of depth with ACRDS (A total of 18 subjects when combined across 3 studies; 3 authors and 15 naives). The specific properties of the task probably helped to fix the subject’s strategy in those three studies.

We also point out that Hibbard et al. (2014) used a gap of 0.35° between the center region and the surrounding reference. A gap between two adjacent regions of a CRDS is known to reduce discrimination performance (Read et al., 2010). Hibbard et al. (2014) showed exactly this evidence of reduced performance, where discrimination was not quite perfect for CRDS (Figure 2). The same reduction may well have brought the discrimination to chance level in ACRDS. Thus, the absence of a gap might be another condition to evoke reversed depth perception in ACRDS.

We acknowledge that the explanation proposed above lacks evidence at the moment. But we hope the idea will help generate new quantitative predictions, and eventually lead our community to a better, more detailed understanding of the mechanisms underlying stereoscopic depth perception through further experiments.


References:
Doi, T., Takano, M., & Fujita, I. (2013). Temporal channels and disparity representations in stereoscopic depth perception. Journal of Vision, 13(13), 26: 1-25.

Doi, T., Tanabe, S., & Fujita, I. (2011). Matching and correlation computations in stereoscopic depth perception. Journal of Vision, 11(3), 1:1-16.

Hibbard, P. B., Scott-Brown, K.C., Haigh, E.C., & Adrain, M. (2014). Depth perception not found in human observers for static or dynamic anti-correlated random dot stereograms. PLoS One, 9(1): e84087.
Law, C. T., & Gold, J. I. (2008). Neural correlates of perceptual learning in a sensory-motor, but not a sensory, cortical area. Nature Neuroscience, 11(4), 505-513.

Prince, S. J., Pointon, A. D., Cumm ing, B. G., & Parker, A. J. (2000). The precision of single neuron responses in cortical area V1 during stereoscopic depth judgments. The Journal of Neuroscience, 20(9), 3387-3400.

Read, J. C., & Eagle, R. A. (2000). Reversed stereo depth and motion direction with anti-correlated stimuli. Vision Research, 40(24), 3345-3358.

Read, J.C., Phillipson, G.P., Serrano-Pedraza, I., Milner, A.D., Parker, A.J. (2010) Stereoscopic vision in the absence of the lateral occipital cortex. PLoS One, 5(9): e12608.

Shadlen, M. N., Britten, K. H., Newsome, W. T., & Movshon, J. A. (1996). A computational analysis of the relationship between neuronal and behavioral responses to visual motion. The Journal of Neuroscience, 16(4), 1486-1510.

Shiozaki, H. M., Tanabe, S., Doi, T., & Fujita, I. (2012). Neural activity in cortical area V4 underlies fine disparity discrimination. The Journal of Neuroscience, 32(11), 3830-3841.

Tanabe, S., Yasuoka, S., & Fujita, I. (2008). Disparity-energy signals in perceived stereoscopic depth. Journal of Vision, 8(3), 22: 1-10.

Uka, T., & DeAngelis, G. C. (2004). Contribution of area MT to stereoscopic depth perception: choice-related response modulations reflect task strategy. Neuron, 42(2), 297-310.