Elsevier

NeuroImage

Volume 51, Issue 1, 15 May 2010, Pages 329-335
NeuroImage

Neural correlates of stimulus-invariant decisions about motion in depth

https://doi.org/10.1016/j.neuroimage.2010.02.011Get rights and content

Abstract

Perceptual decision-making is a complicated, multi-stage process. Recently human neuroimaging studies implicated a set of regions, extending from the medial frontal cortex to the inferior parietal lobule that are involved in various steps of perceptual judgments. However, relatively little is known about the dependence of perceptual decisions on the visual stimulus itself. In the current study, we used functional magnetic resonance imaging to map neural activations while subjects performed a demanding 3D heading estimation task (heading slightly to the left or right of fixation). Subjects (n = 13) were presented a constantly expanding optic-flow stimulus, composed of disparate red–blue spheres, viewed stereoscopically through red–blue glasses. We varied task difficulty either by adding incoherently moving spheres to the stimuli, hence reducing the strength of the motion signal and thereby increasing the amount of noise or by reducing the relevant differential information by decreasing the deviation of the average trajectory of the spheres from straight ahead. BOLD signals were compared during “easy” and “hard” trials in both stimulation conditions to isolate the neural mechanisms underlying the decision process. We hypothesized that areas involved in perceptual decisions about motion should exhibit significantly different activation across both stimulus conditions. Our results indicate that during earlier, sensory-stimulation-related phases of decision-making the left dorsolateral prefrontal cortex, posterior cingulate and inferior parietal cortex showed more activation for the “easy” compared to the “hard” trials, while during later, response-related phases the bilateral precuneus and inferior parietal cortex, as well as the bilateral superior medial gyrus showed this pattern of activation. Our results suggest that a large, non-overlapping network of areas is involved in various steps of decisions regarding 3D motion.

Introduction

During everyday life we sometimes face easier and sometimes harder decisions. Consider for example the task of a goalkeeper in a game of soccer. On a clear sunny day he will have reliable information about the current position and velocity of the ball so as to optimally adjust his position in the goal. On a foggy, rainy day, however, his task will be more difficult, owing to the reduced reliability of the sensory information. These signal-to-noise differences in sensory input will place varying demands on the decision-making processes and will thus require more time to respond to changes of the ball's velocity and trajectory.

Recently, functional magnetic resonance imaging (fMRI) data suggest that judgments about such “easy” and “hard” decisions activate higher level decision-related cortical areas differentially. Based on macaque electrophysiological studies (Kim and Shadlen, 1999) it was proposed that neural activity in cortical areas involved in the active sensory accumulation steps of decision-making increases with increasing sensory evidence (for review see Heekeren et al., 2008, but also see Ho et al., 2009 for a recent demonstration of increased BOLD activation representing greater decision-making activity when a decision is hard). Accordingly, it was found that for decisions about objects (faces and houses), as well as about moving stimuli the left posterior dorsolateral prefrontal cortex (DLPFC) shows larger blood oxygen level dependent (BOLD) signal for easier than for harder trials and that this response pattern is even independent of the response modality (button press or saccadic eye movement; Heekeren et al., 2006), supporting its supramodal role in decision-making.

However, the difficulty of a perceptual decision does not only depend on the added noise that decreases the strength of the stimulus. Consider the example of the goalkeeper. If the ball heads directly towards one of the goalposts it is difficult to judge whether it will land in the goal or outside of it. Thus the amount of the differential information or the strength of the stimulus (i.e. the distance between the goalpost and the trajectory of the ball) will determine if a decision task is easy or difficult. According to recent models of decision-making (Mazurek et al., 2003) these two factors (i.e. the amount of differential information (in this case the direction of motion) and the strength of the stimulus (the amount of noise in the stimulus)) jointly determine the difficulty faced by the decision processes.

Hence, in our current experiments we tested the effect of differential information available in the stimulus and the effect of added noise. We adapted the rationale of Heekeren et al. (2006) hypothesizing that areas representing the evidence accumulation component of perceptual decisions at a supramodal level should respond more to “easy” than to “hard” trials, independent of the way the amount of sensory evidence is manipulated. In such regions, signal strength has been normalized to the prevailing noise level to give a direct measure of the weight of sensory evidence.

We used fMRI to measure the BOLD signal during a three-dimensional (3D) heading discrimination task, using an expanding optic-flow stimulus creating the percept of motion in depth (Kovács et al., 2008). We varied task difficulty either by adding incoherently moving spheres to the stimuli, hence reducing the strength of the motion signal and increasing the amount of noise or we reduced the relevant differential information by decreasing the deviation of the average trajectory of the spheres from straight ahead. BOLD signals were compared during “easy” and “hard” trials in both stimulation conditions. We asked the question: are the cortical areas known to be involved in perceptual decision-making active to the same extent in the two conditions? Our results indicate that the neural correlates of the decision-making process were invariant over the stimulus manipulations (i.e., coherency level, angle of deviation) suggesting a supramodal role for these neural substrates of decisions in motion.

Section snippets

Subjects

Seventeen naïve, healthy volunteers (10 females) with normal or corrected-to normal vision participated in the experiments (mean age: 24.9, SD = 3.3 years). All subjects gave written consent and were screened for MRI compatibility. The procedures were approved by the ethical committee of the University of Regensburg.

Display, stimuli and procedures

Stimuli are similar to those of Kovács et al. (2008), so here only a brief description is given. Fig. 1 depicts our experimental conditions and sample stimuli. Stimuli consisted of

Behavioral results

Analysis of variance with task difficulty (easy vs. hard) and stimulus type (COH vs. FOE) as within subject factors was applied to evaluate correct response rates. As expected a significant main effect of task difficulty (F(1,11) = 7.7, p < 0.01) was found with higher correct response rates for easy compared to difficult trials (Fig. 2). In contrast there was neither a main effect of stimulus type (F(1,11) = 3.1, p < 0.1) nor an interaction between level of difficulty and stimulus type (F(1,11) = 1.8, p < 

Discussion

The aim of the current experiment was to study the neural correlates of perceptual judgments during optic-flow stimulation, simulating 3D motion in depth. To study decision processes we adapted the paradigm of Heekeren et al. (2004), who showed that the degree of activation of areas related to the active sensory accumulation steps of decision-making is correlated with task difficulty. The central assumption of our hypothesis was that if a cortical area is associated with perceptual

Acknowledgments

This work was supported by a grant from the European Union (EU IST Cognitive Systems, project 027198 “Decisions in Motion”), by the Federal Ministry of Education and Research, Germany (BMBF 01GW0653) and by the Deutsche Forschungsgemeinschaft Grant KO 3918/1-1. The authors gratefully acknowledge startup funding from the Bavarian Research Foundation (Grant number: 570/03) and Siemens Medical Solutions.

References (36)

  • Van EssenD.C.

    A Population Average, Landmark and Surface based (PALS) atlas of human cerebral cortex

    Neuroimage

    (2005)
  • BlairK. et al.

    Choosing the lesser of two evils, the better of two goods: specifying the roles of ventromedial prefrontal cortex and dorsal anterior cingulate in object choice

    J. Neurosci.

    (2006)
  • BrettM. et al.

    Region of interest analysis using an SPM toolbox [Abstract]

  • FellowsL.K. et al.

    The role of ventromedial prefrontal cortex in decision making: judgment under uncertainty or judgment per se?

    Cereb. Cortex

    (2007)
  • FleckM.S. et al.

    Role of prefrontal and anterior cingulate regions in decision-making processes shared by memory and nonmemory tasks

    Cereb. Cortex

    (2006)
  • GläscherJ. et al.

    Determining a role for ventromedial prefrontal cortex in encoding action-based value signals during reward-related decision making

    Cereb. Cortex

    (2009)
  • HeekerenH.R. et al.

    A general mechanism for perceptual decision-making in the human brain

    Nature

    (2004)
  • HeekerenH.R. et al.

    Involvement of human left dorsolateral prefrontal cortex in perceptual decision making is independent of response modality

    Proc. Natl Acad. Sci. U. S. A.

    (2006)
  • Cited by (11)

    • Exercise-induced neuroplasticity: Balance training increases cortical thickness in visual and vestibular cortical regions

      2018, NeuroImage
      Citation Excerpt :

      Similar group differences were observed for the right posterior cingulate gyrus, a brain region proposed to be an anatomical interposition of the parieto-medial temporal pathway to the hippocampus (Kravitz et al., 2011). While receiving input from almost all sensory modalities, including the vestibular system (Hitier et al., 2014; Wirth et al., 2018), the posterior cingulate has been reported to be involved in the processing of depth motion cues (Kovács et al., 2010) and the perception of self-location (Guterstam et al., 2015). Noteworthy, the observed cluster coordinates are located at the borders of the retrosplenial cortex, being reciprocally connected with the subiculum, the parahippocampus and the enthorhinal cortex (Vann et al., 2009).

    • Gray matter volume is associated with rate of subsequent skill learning after a long term training intervention

      2014, NeuroImage
      Citation Excerpt :

      These areas have been previously implicated in visual spatial processing and spatial attention processes that would be important for learning to juggle. More specifically, the posterior cingulate cortex (PCC) is involved in spatial attention shifting and in navigation tasks (Hopfinger et al., 2000; Kovacs et al., 2010). The precuneus is a major association area that connects to other parietal regions such as the intraparietal sulcus (IPS) as well as premotor regions like the supplementary motor areas (SMA) (Cavanna and Trimble, 2006).

    • Modeling the BOLD correlates of competitive neural dynamics

      2014, Neural Networks
      Citation Excerpt :

      Our simulation results predict an increase in BOLD response with decreasing task difficulty (higher contrast) for voxels containing WTA networks like the one we used. Human imaging studies have found several regions which exhibit this pattern of BOLD activation during decision-making tasks including the DLPFC and inferior parietal lobule (Heekeren, Marrett, Ruff, Bandettini, & Ungerleider, 2006; Kovacs, Cziraki, & Greenlee, 2010). Our results suggest that these regions functionally correspond to the DLPFC and LIP in the monkey and contain neural circuits similar to the WTA network presented here.

    • Human visual and parietal cortex encode visual choices independent of motor plans

      2012, NeuroImage
      Citation Excerpt :

      While the exact nature of the involvement of parietal and visual cortex in this decision process awaits further study, importantly, the signals we identified in both regions cannot be explained by unspecific overall changes in activity during decision-making, because our analysis assessed the information about specific choices encoded in selective patterns of brain activity. Previous fMRI studies of perceptual decision-making have found a number of brain regions in which the BOLD signal correlated with the decision of the subject (Bankó et al., 2011; Heekeren et al., 2004, 2006; Ho et al., 2009; Kayser et al., 2010a, 2010b; Kovács et al., 2010; Liu and Pleskac, 2011; Noppeney et al., 2010; Tosoni et al., 2008; Woolgar et al., 2011). In line with this work we found that decreases in the strength of sensory evidence were paralleled by increases in fMRI response amplitudes in a wide network of cortical regions outside of visual cortex.

    View all citing articles on Scopus
    View full text