Elsevier

NeuroImage

Volume 42, Issue 4, 1 October 2008, Pages 1686-1697
NeuroImage

Visuospatial contextual processing in the parietal cortex: An fMRI investigation of the induced Roelofs effect

https://doi.org/10.1016/j.neuroimage.2008.06.016Get rights and content

Abstract

Neighboring contextual elements can dramatically affect the manner in which the brain processes the perceptual characteristics of an object. Indeed, many well-known visual illusions rely on misleading contextual cues to create misperceptions of size, length or orientation (e.g., in the Ebbinghaus, Muller-Lyer or rod-and-frame illusions, respectively). However, little is known about the brain regions underlying these integrative computations. The current study used fMRI to delineate the brain areas responsible for processing visuospatial contextual information. Participants were asked to determine whether a small target was positioned left or right of midline in the presence of an offset rectangle designed to induce a shift in the participant's perception of straight-ahead (the induced Roelofs effect). We found localized, bilateral regions in superior parietal cortex and precuneus that were specifically active when participants judged the target location in the presence of this shifted context; significantly less activation was present when a color judgment was made with identical stimuli, or when the location judgment was made without a Roelofs-inducing frame. We propose that this portion of parietal cortex is selectively involved in processing visuospatial contextual information. Additional findings support the notion that perceptual judgments of target location based on an egocentric frame of reference fall within the purview of the dorsal stream of visual processing, rather than the ventral stream.

Introduction

When perceiving and judging attributes such as the size, orientation or location of an object, other elements in the visual scene are often unconsciously taken into account by the observer. Well-documented illusions such as the Ebbinghaus illusion (Wundt, 1898), Müller-Lyer illusion (1889), and the rod-and-frame effect (Witkin and Asch, 1948) rely on extraneous cues to mislead our perceptual system about certain aspects of an attended object. Although visuospatial contextual integration is a factor in many perceptual tasks, few studies to date have explicitly addressed its neural underpinnings (though see Weidner and Fink, 2006, Murray et al., 2006). The current imaging study sought to uncover the neural locus of visuospatial contextual processing in an illusion known as the induced Roelofs effect (Bridgeman et al., 1997; see also Roelofs, 1935). In this illusion, observers in a darkened environment tend to misperceive the location of a target object when it is presented in the context of a large rectangular frame whose center is positioned left or right of the observer's midsagittal plane. For example, a left-shifted frame will cause an enclosed target to appear to lie further to the right, and vice-versa. Though the illusion was initially described by Bridgeman et al. (1997) as resulting from competition between separate allocentric and egocentric frames of reference, the effect has more recently been shown to be the specific result of a distortion of egocentric space. In this updated view (Dassonville et al., 2004, Dassonville and Bala, 2004), the induced Roelofs effect causes the observer's egocentric reference frame to be pulled in the direction of the large rectangle. Indeed, when the observers are asked to “look straight ahead,” the offset frame biases their reports toward the center of the rectangle. Thus, when a target's location is encoded within the observer's biased reference frame, it appears to be displaced in the opposite direction of the rectangular frame.

In the paradigm of the present study, participants were instructed to attend to and report either the location or color of a small target square. In half of the trials, the target was presented alone on a black background. For the remaining half, the target was presented inside a large, colored frame offset from the participant's midline. The context of this larger frame served to induce the Roelofs effect, as well as a color-contrast effect (i.e., a red frame caused the target to be perceived with more of a greenish tint and vice-versa; Webster et al., 2002). As the low-level aspects of the stimuli were identical across the location and color judgments, any observed differences in brain activation could be attributed to the specific judgments made concerning particular aspects of the target (e.g., noting position during the location judgment, and hue during the color judgment). The results provided an opportunity to assess the regions responsible for contextual processing in the two tasks.

Weidner and Fink (2006) have previously examined visuospatial contextual processing in a task that involved a visual illusion, the Müller-Lyer illusion. Their findings suggest that the superior parietal and lateral occipital cortices play a role in visuospatial contextual processing. Given this, it might also be expected that the same areas are responsible for the contextual processing that leads to the Roelofs illusion. On the other hand, it might be expected that the contextual processing involved in a length judgment task (such as that in the Landmark task employed by Weidner and Fink, 2006) is different from that involved in an egocentric location task such as the one used here. Indeed, a previous study from our lab (Walter et al., in press) assessed individual differences in susceptibilities to these illusions, and found evidence that they were driven by separate underlying mechanisms. Given this, we predicted that the pattern of activation associated with contextual processing in the Roelofs illusion will differ substantially from that of the Müller-Lyer illusion.

The results of this study also allowed us to evaluate the utility of two major characterizations of function for the separate streams of visual processing, referred to henceforth as the “what vs. where” (Ungerleider and Mishkin, 1982) and “perception vs. action” hypotheses (Milner and Goodale, 1995). For example, the what vs. where framework suggests that there is a ventral stream of visual areas that is responsible for processing the visual information necessary for determining the identity of a viewed object, and a dorsal stream responsible for determining the object's location in space (Ungerleider and Mishkin, 1982). Within this framework, then, we would expect a predominantly ventral pattern of activation during the trials that require a color judgment, and a dorsal pattern during the trials that require a location judgment. The perception vs. action framework, on the other hand, suggests that the ventral stream is responsible for processing the visual information that leads to our perception of the objects that populate the visual world, whereas the dorsal stream is responsible for guiding actions (e.g., reaches, pointing movements, eye movements) toward these objects (Milner and Goodale, 1995). Both of the tasks employed here were perceptual in nature, in that they involved a perceptual judgment that was expressed with a symbolic button press (rather than an isomorphic goal-directed response aimed directly toward the stimulus itself). Thus, the perception vs. action framework leads to the prediction that both the color and location judgments will give rise to a predominantly ventral pattern of activation.

Section snippets

Participants

Sixteen right-handed participants (10 females; 18–32 years of age) gave their informed, written consent as per the University of Oregon Institutional Review Board. Participants were compensated either with money or course credit for an introductory psychology course, and were naïve to the goals of the study. An initial behavioral calibration session took approximately one hour, while the scanning session itself lasted 1.5–2 h.

Stimuli

During each trial, participants were presented with a small colored

Behavioral effects

Based on the results of earlier examinations of the Roelofs effect, it was expected that the location of the frame would affect the perceived location of the target, with, for example, left-shifted frames causing a rightward bias in the participant's perception of the target location. Indeed, a significant effect of frame location was seen in the Location task during pre-testing sessions (F(4, 60) = 5.041, p < 0.005, with a significant linear contrast, F(1, 15) = 7.610, p < 0.05; average gain of 3.0%),

Discussion

This study examined the brain areas responsible for judgments of location and color, as well as the areas that are involved in processing contextual cues associated with these judgments. We demonstrated brain activations that were uniquely related to decisions about a target's location versus its color, even when viewing identical visual stimuli. Specifically, judgments regarding location in our paradigm activated a network of parietal and frontal areas, relative to the tasks requiring a

Summary

The current study found that portions of posterior parietal cortex (BA 7) were selectively active when making decisions about the location of a target in the presence of contextual information. Though located in posterior parietal cortex, these areas are distinct from regions that are directly related to eye-movement and attentional shifts. This provides support for the idea that the function of these parietal areas extends beyond egocentric transformations and spatial attention, and includes a

Acknowledgments

We thank Scott Watrous and Mark Dow for technical support throughout this study. This work was derived from a portion of the doctoral dissertation of Elizabeth Walter, and was partially supported by an NIH Systems Physiology Training Grant (5-T32-GM07257) to EW and an award from the Telemedicine and Advanced Technology Research Center (DAMD17-01-1-0750). Preliminary results were presented at the 2006 Annual Meeting of the Vision Sciences Society.

References (41)

  • DassonvilleP. et al.

    The induced Roelofs effect: two visual systems or the shift of a single reference frame?

    Vis. Res.

    (2004)
  • WojciulikE. et al.

    The generality of parietal involvement in visual attention

    Neuron

    (1999)
  • BridgemanB. et al.

    Interaction of cognitive and sensorimotor maps of visual space

    Percept. Psychophys.

    (1997)
  • CeyteH. et al.

    Role of gravity-based information on the orientation and localization of the perceived body midline

    Exp. Brain Res.

    (2007)
  • ColbyC. et al.

    Space and attention in parietal cortex

    Annu. Rev. Neurosci.

    (1999)
  • CorenS. et al.

    An empirical taxonomy of visual illusions

    Percept. Psychophys.

    (1976)
  • CorenS. et al.

    Individual differences in visual-geometric illusions: Predictions from measures of spatial cognitive abilities

    Percept. Psychophys.

    (1987)
  • DassonvilleP. et al.

    Action, perception and the Roelofs effect: a mere illusion of dissociation

    PLoS Biology

    (2004)
  • DukelowS.P. et al.

    Distinguishing subregions of the human MT+ complex using visual fields and pursuit eye movements

    J. Neurophysiol.

    (2001)
  • DumoulinS. et al.

    A new anatomical landmark for reliable identification of human area V5/MT: a quantitative analysis of sulcal patterning

    Cerebral Cortex

    (2000)
  • DydeR.T. et al.

    Two illusions of perceived orientation: one fools all of the people some of the time; the other fools all of the people all of the time

    Exp. Brain Res.

    (2002)
  • GalatiG. et al.

    The neural basis of egocentric and allocentric coding of space in humans: a functional magnetic resonance study

    Exp. Brain Res.

    (2000)
  • GoebelR. et al.

    Analysis of functional image analysis contest (FIAC) data with Brainvoyager QX: from single-subject to cortically aligned group general linear model analysis and self-organizing group independent component analysis

    Hum. Brain Mapp.

    (2006)
  • HadjikhaniN. et al.

    Retinotopy and color sensitivity in human visual cortical area V8

    Nat. Neurosci.

    (1998)
  • KastnerS. et al.

    Mechanisms of visual attention in the human cortex

    Ann. Rev. Neurosci.

    (2000)
  • Lacadie, C.M., Fulbright, R.K., Rajeevan, N., Constable, R.T., Papademetris, X., in press. More accurate Talairach...
  • LancasterJ.L. et al.

    Automated Talairach atlas labels for functional brain mapping

    Hum. Brain Mapp.

    (2000)
  • MarendazC.

    Nature and dynamics of reference frames in visual search for orientation: Implications for early visual processing

    Psychol. Sci.

    (1998)
  • MilnerA.D. et al.

    The visual brain in action

    (1995)
  • Müller-LyerF.C.

    Optische Urteilstäuschungen [Deception of visual judgment]

    Dubois-Reymonds Archive für Anatomie und Physiologie

    (1889)
  • Cited by (31)

    • Defining the neural correlates of spontaneous theory of mind (ToM): An fMRI multi-study investigation

      2019, NeuroImage
      Citation Excerpt :

      The dorsal peak corresponded to the supramarginal gyrus (SMG), or BA40, a region that is part of the inferior parietal lobule (IPL) (Igelström and Graziano, 2017) and is likewise activated in false belief tasks (see the meta-analysis by Schurz et al., 2013). However, the SMG is also involved in other functions, such as spatial working memory, spatial attention and visuospatial processing (Silk et al., 2010; Walter and Dassonville, 2008). Therefore, our findings, while confirming the activation of the TPJ in false belief reasoning, also extend this role to other parietal and temporal areas with peaks in the SMG and MTG, respectively.

    • The Roelofs and induced Roelofs effects

      2018, Consciousness and Cognition
    View all citing articles on Scopus
    View full text