Elsevier

NeuroImage

Volume 59, Issue 2, 16 January 2012, Pages 1924-1931
NeuroImage

The neural encoding of guesses in the human brain

https://doi.org/10.1016/j.neuroimage.2011.08.106Get rights and content

Abstract

Human perception depends heavily on the quality of sensory information. When objects are hard to see we often believe ourselves to be purely guessing. Here we investigated whether such guesses use brain networks involved in perceptual decision making or independent networks. We used a combination of fMRI and pattern classification to test how visibility affects the signals, which determine choices. We found that decisions regarding clearly visible objects are predicted by signals in sensory brain regions, whereas different regions in parietal cortex became predictive when subjects were shown invisible objects and believed themselves to be purely guessing. This parietal network was highly overlapping with regions, which have previously been shown to encode free decisions. Thus, the brain might use a dedicated network for determining choices when insufficient sensory information is available.

Highlights

► We decoded object stimuli and choices from fMRI signals in the human brain. ► Our study revealed different networks for choices under high and low visibility. ► Without sufficient visual information, guesses were encoded in the precuneus. ► The brain's parietal guessing network overlaps with the free decision network. ► This network might act as a generator for internal random decision outcomes.

Introduction

Under ideal circumstances we can obtain sufficient information to identify objects in our visual environment. But how do we make decisions about objects when they are barely visible? There are two possible mechanisms that could explain human choices about objects that are difficult to see. One view holds that random noise fluctuations in the sensory system determine subjects' trial-by-trial choices (Shadlen et al., 1996, Swets, 1961). This would suggest that sensory regions play an important role in decisions, even for invisible stimuli. On the other hand, subjects frequently report to be “purely guessing” and “randomly choosing” when they are forced to categorise objects that are invisible. This could suggest that choices are based on specific brain networks involved in guessing and free choices (Elliott et al., 1997, Elliott et al., 1999, Haggard, 2008, Jenkins et al., 2000, Paulus et al., 2001, Soon et al., 2008). We thus sought to directly investigate whether strong differences in visibility have an effect on which brain regions determine a subject's choice. For this, we used a combination of multivariate pattern recognition (Haynes and Rees, 2006, Kriegeskorte et al., 2006, Norman et al., 2006) and functional magnetic resonance imaging (fMRI). Although fMRI is limited in its ability to resolve processes at the single-neuron level, it is uniquely suited to investigate large-scale changes in cortical networks under different stimulation conditions. By using pattern recognition we were able to overcome a limitation of conventional fMRI analyses that assess local differences in activation levels for each brain location individually. Pattern recognition allows extracting the full information encoded in fine-grained, local patterns of brain activity that are often overlooked by conventional fMRI analyses (Haynes and Rees, 2006, Norman et al., 2006). It has been shown that pattern recognition can successfully decode object categories from fMRI signals in the lateral occipital complex (LOC) (Cox and Savoy, 2003, Haxby et al., 2001, Williams et al., 2007) and can be used to assess prediction of perceptual choices (Serences and Boynton, 2007) as well as high-level cognitive representations such as rules and intentions (Bode and Haynes, 2009, Soon et al., 2008). Here, we used a searchlight decoding approach (Kriegeskorte et al., 2006, Soon et al., 2008) to search throughout the entire brain in an unbiased fashion for regions predicting sensory information and choice-related information, similar to single-cell studies of perceptual decision making (Britten et al., 1996, Parker and Newsome, 1998). On each trial subjects saw flashed sequences of object images embedded in scrambled masks with a timing that rendered their visibility either high or low (Grill-Spector et al., 2000). Our main aim was to assess any differential predictive information in cortical networks under low visibility when subjects were guessing versus under high visibility when subjects were extracting visual information from the sensory stimulus.

Section snippets

Participants

19 right-handed subjects with normal or corrected to normal visual acuity gave written informed consent and participated in the study. The experiment was approved by the local ethics committee and was conducted according to the Declaration of Helsinki. Data from two subjects were excluded due to excessive head movement. Data from another three subjects were excluded because their categorisation performance did not differ between the two visibility conditions, suggesting a lack of attention to

Results

For all categories the visibility was higher in the long target exposure duration as opposed to the short exposure duration (all p < 0.001; Fig. 1C; high visibility hit rates pianos 89%, chairs 93%, noise 79%; low visibility hit rates pianos 34%, chairs 45%, noise 46%; chance level 33%; see Supplementary Fig. 1 and Supplementary Table 1 for individual results). The same category was never perceived better under “low visibility” compared to “high visibility” by any of the subjects. In line with

Discussion

Our study assessed both stimulus-related and choice-related information in cortical networks during perceptual decision making and guessing. First, we confirmed that the specific stimuli subjects were viewing under high visibility could be reliably decoded from spatial patterns of brain signals in the LOC, a region involved in high-level processing of object information (Cox and Savoy, 2003, Haxby et al., 2001). However, there was a marked difference between conditions of high and low

Conclusion

In summary, we demonstrated that the brain might switch between two different decision networks depending on whether visual stimuli are easy or difficult to see. For clearly visible objects the brain might use the information contained in specialised sensory brain regions. In the absence of sufficient perceptual input when guessing is required, the decision relies on neural populations in the precuneus that are different from those for perceptual decisions under high visibility. The precuneus

Acknowledgments

This work was supported by the German Research Foundation (DFG Grant HA 5336/1-1); the Bernstein Computational Neuroscience Program of the German Federal Ministry of Education and Research (BMBF Grant 01GQ0411); the Excellency Initiative of the German Federal Ministry of Education and Research (DFG Grant GSC86/1-2009) and the Max Planck Society. S.B. received the OHBM Students Travel Award 2008. The authors thank Tobias Donner for valuable comments on the manuscript and Anna H. He for comments

References (54)

  • C.K. Abbey et al.

    Classification image analysis: estimation and statistical inference for two-alternative forced-choice experiments

    J. Vis.

    (2002)
  • D.C. Bradley et al.

    Encoding of three-dimensional structure-from-motion by primate area MT neurons

    Nature

    (1998)
  • K.H. Britten et al.

    A relationship between behavioral choice and the visual responses of neurons in macaque MT

    Vis. Neurosci.

    (1996)
  • A.E. Cavanna et al.

    The precuneus: a review of its functional anatomy and behavioural correlates

    Brain

    (2006)
  • Cichy, R.M., Heinzle, J., Haynes, J.D., in press. Imagery and perception share cortical representations of content and...
  • J. Duncan

    An adaptive coding model of neural function in prefrontal cortex

    Nat. Rev. Neurosci.

    (2001)
  • R. Elliott et al.

    Ventromedial prefrontal cortex mediates guessing

    Neuropsychologia

    (1999)
  • J.I. Gold et al.

    Representation of a perceptual decision in developing oculomotor commands

    Nature

    (2000)
  • J.I. Gold et al.

    The neural basis of decision making

    Annu. Rev. Neurosci.

    (2007)
  • K. Grill-Spector et al.

    The dynamics of object-selective activation correlate with recognition performance in humans

    Nat. Neurosci.

    (2000)
  • P. Haggard

    Human volition: towards a neuroscience of will

    Nat. Rev. Neurosci.

    (2008)
  • J.V. Haxby et al.

    Distributed and overlapping representations of faces and objects in ventral temporal cortex

    Science

    (2001)
  • J.D. Haynes et al.

    Predicting the orientation of invisible stimuli from activity in human primary visual cortex

    Nat. Neurosci.

    (2005)
  • J.D. Haynes et al.

    Decoding mental states from brain activity in humans

    Nat. Rev. Neurosci.

    (2006)
  • H.R. Heekeren et al.

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

    Nature

    (2004)
  • H.R. Heekeren et al.

    The neural systems that mediate human perceptual decision making

    Nat. Rev. Neurosci.

    (2008)
  • I.H. Jenkins et al.

    Self-initiated versus externally triggered movements. II. The effect of movement predictability on regional cerebral blood flow

    Brain

    (2000)
  • Cited by (21)

    • In search of consciousness: Examining the temporal dynamics of conscious visual perception using MEG time-series data

      2019, Neuropsychologia
      Citation Excerpt :

      In fMRI studies, decoding studies have shown that it is possible to predict various types of stimulus-specific information such as stimulus category and stimulus location from brain activity (e.g., Carlson et al., 2003; Cox and Savoy, 2003; Haxby et al., 2001; Haynes, 2015; Kamitani and Tong, 2005; Kriegeskorte et al., 2006; Shinkareva et al., 2012). Moreover, the decoding performance has been shown to change as a function of consciousness (e.g., Williams et al., 2007; Bode et al., 2012). Williams et al., for example, found that patterns of activity in the Lateral Occipital Complex (LOC) could reliably predict object category only when the participants were consciously aware of the stimuli.

    • Dissociating neural variability related to stimulus quality and response times in perceptual decision-making

      2018, Neuropsychologia
      Citation Excerpt :

      After a short fixation period (white fixation cross for 200 ms), one of the 24 stimuli was presented at one of the three possible stimulus quality levels for 50 ms. This was followed by the checkerboard mask (250 ms), and subsequently by the response screen (2000 ms), which displayed the letters “P” (piano) and “C” (chair) on either the left or right hand side of the fixation cross (Fig. 1B). The mapping of categories and response buttons was randomised for each trial in order to de-correlate choices from motor responses (Bode et al., 2012a, 2012b, 2013). Participants indicated their response by pressing either the left or the right response button of an MRI-compatible Cedrus button box using the index- and middle finger of the right hand.

    • A role for the precuneus in thought-action fusion: Evidence from participants with significant obsessive-compulsive symptoms

      2014, NeuroImage: Clinical
      Citation Excerpt :

      In this way we can see that there may be a role for the precuneus in thought–action fusion, which involves making causal inferences which are not based on previous prediction success, i.e. in the absence of previous experience of thoughts causing events. Similarly, there is evidence to suggest that the precuneus is involved in ‘free’ decision making, that is, based on entirely internal reasoning or feeling rather than perceptually or evidence driven (Bode et al., 2012; Pesaran et al., 2008). In terms of agency, increased activation of the precuneus has been reported when participants take first-person perspective in a story (participant as agent) compared to third-person (other as agent) perspective (Vogeley et al., 2001).

    View all citing articles on Scopus
    View full text