Elsevier

Cognitive Brain Research

Volume 21, Issue 3, November 2004, Pages 401-411
Cognitive Brain Research

Research report
Neural foundations of emerging route knowledge in complex spatial environments

https://doi.org/10.1016/j.cogbrainres.2004.06.013Get rights and content

Abstract

Behavioral evidence suggests that spatial knowledge derived from ground-level navigation can consist of both route and survey knowledge. Neuroimaging and lesion studies aiming to identify the neural structures responsible for topographical learning in humans have yielded partially inconsistent results, probably due to the lack of an effective behavioral parameter allowing for a reliable distinction between different representations. Therefore, we employed a novel virtual reality environment that provides accuracy and reaction time measures precisely indicating the emergence of route vs. survey knowledge. Functional magnetic resonance imaging (fMRI) was used to localize brain regions involved in the acquisition of pure route knowledge in the form of associations between consecutive landmark views and the direction of intermediate movements. Participants were scanned during repeated encoding of the complex environment from a first-person, ground-level perspective. Behavioral data revealed emerging route knowledge in 11 out of 14 subjects. Overall comparisons between encoding and control conditions identified activation in medial frontal gyrus, retrosplenial cortex and posterior inferior parietal cortex. Most importantly, only posterior inferior parietal regions showed increasing activation across sessions, thus paralleling behavioral measures of route expertise. Given the established role of the posterior parietal cortex in spatial processing, this area is thought to provide the pivotal spatial link between two landmarks encountered in immediate temporal succession.

Introduction

Human spatial navigation is the expression of a complex ensemble of cognitive processes based on distinct spatial representations. Survey representations refer to an enduring, goal-independent representation of the environment. These representations—also termed cognitive maps or configurational knowledge—allow the individual to infer spatial relations between any two places irrespective of his own position. In contrast, route knowledge conveys information about a goal-directed, temporo-spatial sequence of environmental features. Salient objects (i.e. landmarks) in the environment are linked by their temporal occurrence during navigation and by spatial relations that determine movement decisions necessary to get from one landmark to the next (i.e. turn right). These spatial relations are encoded predominantly between neighboring objects [21], [42], presumably with respect to a place-dependent local reference system [40]. In addition, the connections between landmarks appear to be coded only in the direction experienced during navigation [45].

Even though different representations can be the result of experiencing an environment from different perspectives [46], [47], ground-level navigation by itself not only entails route knowledge [21], [34], [53], but can also result in survey representations [43], [51]. Hierarchical models have therefore conceptualized navigational learning as a succession of landmark, route and survey knowledge, the latter depending on a qualitative shift in the knowledge representation [48], [51]. However, the importance of both the applied learning strategy and general spatial abilities for the resultant representation has been emphasized repeatedly, arguing for the possibility to develop route or survey knowledge from the very beginning of a learning experience [4], [28].

Whereas many neuroimaging and lesion studies on spatial navigation have successfully investigated the retrieval of previously learned environments [1], [24], [26], [31], [32], [35], the neural foundations of learning complex spatial layouts have received less attention [3], [7], [33], [46]. Even though these studies have identified a network of areas including frontal, posterior parietal, retrosplenial and medial temporal regions the precise role of these structures remains controversial, particularly with respect to parietal function. For example, whereas Barrash et al. [7] did not find a systematic relationship between parietal lesions and route learning impairment, Shelton and Gabrieli [46] observed greater inferior parietal activity in an fMRI study for route as compared to survey encoding. One reason might be the lack of a behavioral measure allowing for a clear identification of the type of knowledge that was acquired, since many spatial memory tasks can be performed based on either route or survey representations [16]. In nonhuman primates, the importance of the posterior parietal cortex for route learning and retrieval has been demonstrated repeatedly [8], [52], arguing for its involvement in processing the spatial relationship between local environmental cues instead of representing the allocentric position of a goal or a fixed sequence of movements. In addition, subdivisions of the posterior parietal cortex have consistently been associated with general spatial abilities like mental rotation [6], [23], [25] or perspective taking [54] in humans, presumably reflecting the occurrence of mental spatial transformations. However, it remains unclear to what extent these areas are implicated in route learning as well.

In the present study, we aimed to characterize the neural system involved in one type of navigational learning, the acquisition of pure route knowledge. Route knowledge was conceptualized as an association between landmarks encountered in immediate temporal succession and the spatial relations connecting them. A complex virtual environment was designed enabling us to link behavioral performance directly to the gradual emergence of route knowledge. Reaction times and performance measures were obtained from judgments of spatial relations between pairs of buildings from adjacent intersections. With regard to the expected brain activation patterns, we formulated the following hypotheses: One important prerequisite for route learning consists of storing the temporal sequence of landmarks. This may entail involvement of medial frontal areas, given that they repeatedly have been associated with memory for temporal order and context [27], [50]. More importantly, the spatial relations between neighboring landmarks constitute the crucial links that can effectively guide navigational behavior. These relations presumably are encoded with respect to multiple local reference systems [40] that are imposed on distinct places within the environment. Considering the egocentric nature of these reference systems, we predicted a prominent involvement of the posterior parietal cortex in coding the spatial relations; an assumption based on its well-established role in the processing of spatial positions of external objects in various egocentric reference frames [5], [19], [20], [22].

Section snippets

Participants

Fourteen healthy male volunteers (mean age: 29.3 years, S.D.: 5.4, range: 23–35) with normal or corrected-to-normal vision gave written informed consent to participate in this study. The study was approved by the local ethics committee. All subjects understood the instructions without difficulty and none were aware of the hypotheses at the time of testing.

Experimental stimuli

In order to localize brain regions representing the emergence of route knowledge, a desktop virtual environment (Fig. 1) was constructed

Behavioral data

Participants were classified as route learners based on behavioral performance; in addition, we checked the maps drawn after fMRI scanning for global and/or local inconsistencies. According to our hypotheses, acquisition of route knowledge should manifest itself in performance improvements only for direct pairs. This behavioral pattern was observed in 11 out of 14 participants as reflected by response times and accuracy measures during retrieval (see Fig. 2). Whereas performance for both close

Discussion

In the present study, our aim was to determine the neural structures involved in route learning defined as temporo-spatial associations between consecutive landmarks. Therefore, we developed a behavioral paradigm that provided an objective distinction between route knowledge and survey representations. Eleven out of 14 participants showed evidence of pure route learning, reflected by increasing accuracy and decreasing response times during retrieval. Importantly, in these subjects, no

Acknowledgements

We thank the Physics and Methods group at NeuroImage Nord in Hamburg, Ron Paludan (www.railwaystation.com) for providing several 3D-models, and D. Waller, K.F. Richter, F. Binkofski, E. Schoell, A. McNamara and D. Gonzalo for suggestions on an earlier draft of this paper. This work was supported by the Volkswagenstiftung and the Bundesministerium für Bildung und Forschung (BMBF).

References (54)

  • K. Jordan et al.

    Cortical activations during the mental rotation of different visual objects

    NeuroImage

    (2001)
  • E. Mellet et al.

    Neural correlates of topographic mental exploration: the impact of route versus survey perspective learning

    NeuroImage

    (2000)
  • K.A. Paller et al.

    Observing the transformation of experience into memory

    Trends Cogn. Sci.

    (2002)
  • A.W. Siegel et al.

    The development of spatial representations of large-scale environments

  • M. Suzuki et al.

    Neural basis of temporal context memory: a functional MRI study

    NeuroImage

    (2002)
  • P.W. Thorndyke et al.

    Differences in spatial knowledge acquired from maps and navigation

    Cogn. Psychol.

    (1982)
  • J. Traverse et al.

    Impairments in route negotiation through a maze after dorsolateral frontal, inferior parietal or premotor lesions in cynomolgus monkeys

    Behav. Brain Res.

    (1986)
  • R.F. Wang et al.

    Human spatial representation: insights from animals

    Trends Cogn. Sci.

    (2002)
  • G.K. Aguirre et al.

    Environmental knowledge is subserved by separable dorsal/ventral neural areas

    J. Neurosci.

    (1997)
  • G.K. Aguirre et al.

    Topographical disorientation: a synthesis and taxonomy

    Brain

    (1999)
  • G.K. Aguirre et al.

    The parahippocampus subserves topographical learning in man

    Cereb. Cortex

    (1996)
  • G.L. Allen

    Spatial abilities, cognitive maps and wayfinding: bases for individual differences in spatial cognition and behavior

  • R.A. Andersen et al.

    Multimodal representation of space in the posterior parietal cortex and its use in planning movements

    Annu. Rev. Neurosci.

    (1997)
  • Y.E. Cohen et al.

    A common reference frame for movement plans in the posterior parietal cortex

    Nat. Rev., Neurosci.

    (2002)
  • M. Critchley

    The Parietal Lobes

    (1953)
  • L. Davachi et al.

    When keeping in mind supports later bringing to mind: neural markers of phonological rehearsal predict subsequent remembering

    J. Cogn. Neurosci.

    (2001)
  • A.C. Evans et al.

    3D statistical neuroanatomical models from 305 MRI volumes

    Proc. IEEE Nucl. Sci. Symp. Med. Imaging

    (1993)
  • Cited by (98)

    • Spatial orientation – a stable marker for vascular cognitive impairment?

      2023, Cerebral Circulation - Cognition and Behavior
    • In search of a naturalistic neuroimaging approach: Exploration of general feasibility through the case of VR-fMRI and application in the domain of episodic memory

      2022, Neuroscience and Biobehavioral Reviews
      Citation Excerpt :

      Both Parslow et al. (2004) (for both viewer-dependent and independent conditions) and Slobounov et al. (2010) found activation in the occipital regions, while only Slobounov et al. (2010) found activation in the somatosensory and visual regions, namely activation in the lateral extrastriate visual cortex extending into V2 during encoding and activation in the premotor cortex during retrieval. Other regions involved regions include the thalamic and cerebellar areas (Antonova et al., 2011; Parslow et al., 2004), with a differentiated activation in the cerebellum for Iglói et al. (2015) (left cerebellum for place-based responses; right cerebellum for sequence-based egocentric responses); RSC (Salgado-Pineda et al., 2016; Wolbers et al., 2004); precuneus (during both encoding and retrieval for Slobounov et al. (2010)); anterior cingulate and precentral gyri (Parslow et al., 2004). Furthermore, Iglói et al. (2015) noted that the two networks they observed were activated during training sequences.

    View all citing articles on Scopus
    View full text