Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Brief Communication
  • Published:

Human entorhinal cortex represents visual space using a boundary-anchored grid

Nature Neuroscience volume 21, pages 191–194 (2018)Cite this article

Subjects

Abstract

When participants performed a visual search task, functional MRI responses in entorhinal cortex exhibited a sixfold periodic modulation by gaze-movement direction. The orientation of this modulation was determined by the shape and orientation of the bounded search space. These results indicate that human entorhinal cortex represents visual space using a boundary-anchored grid, analogous to that used by rodents to represent navigable space.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Visual grid-like representation in human EC.
Fig. 2: Visual grid orientation is anchored to the search display geometry.
Fig. 3: Visual grid orientation rotates in concert with rotation of the search display.

Similar content being viewed by others

References

  1. Hafting, T., Fyhn, M., Molden, S., Moser, M.-B. & Moser, E. I. Nature 436, 801–806 (2005).

    Article  CAS  PubMed  Google Scholar 

  2. Krupic, J., Bauza, M., Burton, S., Barry, C. & O’Keefe, J. Nature 518, 232–235 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. Stensola, T., Stensola, H., Moser, M.-B. & Moser, E. I. Nature 518, 207–212 (2015).

    Article  CAS  PubMed  Google Scholar 

  4. Killian, N. J., Jutras, M. J. & Buffalo, E. A. Nature 491, 761–764 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  5. Doeller, C. F., Barry, C. & Burgess, N. Nature 463, 657–661 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Jacobs, J. et al. Nat. Neurosci. 16, 1188–1190 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Irwin, D. E., Yantis, S. & Jonides, J. Atten. Percept. Psychophys. 34, 49–57 (1983).

    Article  CAS  Google Scholar 

  8. Gardner, J. L., Merriam, E. P., Movshon, J. A. & Heeger, D. J. J. Neurosci. 28, 3988–3999 (2008).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Golomb, J. D. & Kanwisher, N. Cereb. Cortex 22, 2794-2810 (2012).

    Google Scholar 

  10. Duhamel, J.-R., Colby, C. L. & Goldberg, M. E. Science 255, 90–92 (1992).

    Article  CAS  PubMed  Google Scholar 

  11. Galletti, C., Battaglini, P. P. & Fattori, P. Exp. Brain Res. 96, 221–229 (1993).

    Article  CAS  PubMed  Google Scholar 

  12. Snyder, L. H., Grieve, K. L., Brotchie, P. & Andersen, R. A. Nature 394, 887–891 (1998).

    Article  CAS  PubMed  Google Scholar 

  13. Dean, H. L. & Platt, M. L. J. Neurosci. 26, 1117–1127 (2006).

    Article  CAS  PubMed  Google Scholar 

  14. Rolls, E. T. Hippocampus 9, 467–480 (1999).

    Article  CAS  PubMed  Google Scholar 

  15. McNaughton, B. L., Battaglia, F. P., Jensen, O., Moser, E. I. & Moser, M.-B. Nat. Rev. Neurosci. 7, 663–678 (2006).

    Article  CAS  PubMed  Google Scholar 

  16. Meister, M. L. & Buffalo, E. A. Neurobiol. Learn. Mem. 134 Pt A, 135–144 (2016).

    Article  PubMed  Google Scholar 

  17. Killian, N. J., Potter, S. M. & Buffalo, E. A. Proc. Natl. Acad. Sci. USA 112, 15743–15748 (2015).

    CAS  PubMed  PubMed Central  Google Scholar 

  18. Horner, A. J., Bisby, J. A., Zotow, E., Bush, D. & Burgess, N. Curr. Biol. 26, 842–847 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Constantinescu, A. O., O’Reilly, J. X. & Behrens, T. E. J. Science 352, 1464–1468 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Bellmund, J. L., Deuker, L., Navarro Schröder, T. & Doeller, C. F. eLife 5, e17089 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

  21. Hegarty, M., Richardson, A. E., Montello, D. R., Lovelace, K. & Subbiah, I. Intelligence 30, 425–447 (2002).

    Article  Google Scholar 

  22. Brainard, D. H. Spat. Vis. 10, 433–436 (1997).

    Article  CAS  PubMed  Google Scholar 

  23. Jenkinson, M., Bannister, P., Brady, M. & Smith, S. Neuroimage 17, 825–841 (2002).

    Article  PubMed  Google Scholar 

  24. Smith, S. M. Hum. Brain Mapp. 17, 143–155 (2002).

    Article  PubMed  Google Scholar 

  25. Kunz, L. et al. Science 350, 430–433 (2015).

    Article  CAS  PubMed  Google Scholar 

  26. Stangl, M., Shine, J. & Wolbers, T. Front. Neuroinform. 11, 47 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

  27. Woolrich, M. W., Ripley, B. D., Brady, M. & Smith, S. M. Neuroimage 14, 1370–1386 (2001).

    Article  CAS  PubMed  Google Scholar 

  28. Berens, P. J. Stat. Softw. 31, 1–21 (2009).

    Article  Google Scholar 

  29. Zar, J. H. Biostatistical Analysis. (Pearson Education India, New Delhi, 1999).

    Google Scholar 

Download references

Acknowledgements

We thank G. Aguirre for providing the eye tracking camera and J. Ryan and S. Liapis for help with data collection. This work was supported by US NIH grants EY-022350 and EY-027047 (R.A.E.) and NSF grant SBE-0541957 (R.A.E.) and Graduate Research Fellowship (J.B.J.).

Author information

Authors and Affiliations

  1. Department of Psychology, University of Pennsylvania, Philadelphia, PA, USA

    Joshua B. Julian, Alexandra T. Keinath, Giulia Frazzetta & Russell A. Epstein

Authors
  1. Joshua B. Julian
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Alexandra T. Keinath
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Giulia Frazzetta
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Russell A. Epstein
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

J.B.J., A.T.K. and R.A.E. designed the study. Data collection was performed by J.B.J., A.T.K. and G.F.; J.B.J. and A.T.K. analyzed data. The manuscript was drafted by J.B.J., A.T.K. and R.A.E.

Corresponding authors

Correspondence to Joshua B. Julian or Russell A. Epstein.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Integrated Supplementary Information

Supplementary Figure 1 Eye tracking and fMRI analysis methods.

a) Schematic of how eye tracking data were transformed into parametric modulators (PM) for the fMRI analyses, depicted here for a subset of the data from one participant. After extracting periods of fast gaze movements, gaze movement directions (θ) over time were split into halves by scan run. In the first half of the data, we identified the orientation of the putative visual grid axes in each participant’s bilateral EC. To do so, we i) equated all 6-fold symmetric gaze movement directions (θ modulo 60°), and then ii) modeled the fMRI data using a GLM with two PMs, cos(6θ) and sin(6θ), down-sampled to the fMRI acquisition rate. The weights on these two PMs were used to calculate the average EC grid orientation (φ; see Methods), which was subsequently used to predict the grid signal during the second half of the scan runs. In particular, for the second half of the data, we iii) equated all 6-fold symmetric gaze movement directions aligned to φ, and then iv) modeled the fMRI data using a GLM with one PM, cos(6θ- φ), to test for reliable 6-fold symmetric periodic modulation of the fMRI signal as a function of gaze movement direction aligned to φ (as in Fig. 1B), and v) modeled the fMRI data using a GLM with two PMs to assess the fMRI signal compared to baseline when gaze movements were aligned or misaligned to φ (as in Fig. 1C). b) Example single-trial viewing traces from six different participants (6–33 seconds in length). c) Within each fMRI acquisition, gaze tended to move in only a single direction, indicating that changes to gaze movement direction tended to occur on a longer timescale than individual gaze movements themselves; α denotes the modal gaze movement direction within a given TR. d) Histogram of gaze movement speeds across all fMRI acquisitions separately for each participant, for the periods of fast gaze movements that were used to test for grid-like coding of visual space.

Supplementary Figure 2 Additional evidence for visual grid-like coding in entorhinal cortex (EC).

a) fMRI activation as a function of gaze movement direction in a 2mm sphere centered on the peak EC voxel identified from the group analysis (Fig. 1B), plotted separately for each 15° interval relative to the grid orientation φ (left), and averaged over aligned and misaligned intervals (right). To ensure that any differences between directions were not induced by a gaze movement bias (see Supplementary Fig. 3), for this particular analysis we included dummy regressors of no interest in the GLM to remove fMRI variance associated with gaze movement directions that were oversampled relative to the least sampled gaze movement direction. The excluded fMRI acquisitions were randomly selected for each participant. This procedure ensured that the same numbers of fMRI acquisitions were used to estimate the response for each gaze movement direction for each participant. There was significantly greater fMRI signal aligned to the grid axes than misaligned (n=36 participants; t-test, t(35)=2.93, p=0.003) in the 2mm spherical peak EC voxel ROI identified from the group analysis (Fig. 1B), confirming that grid-like coding of visual space was not induced by a gaze movement bias. b) Orientation consistency across split-halves of the data in the 2mm spherical peak EC voxel ROI identified from the group analysis including all subjects (Fig. 1B) for 90°, 60°, and 45° periodicities, plotted separately for the square and rectangle display shape participants. For both groups (n=18 participants per group), there was significant orientation consistency across runs for 60° periodicity (t-tests: square: t(17)=2.26, p=0.019; rectangle: t(17)=2.48, p=0.012), but not 90° (square: t(17)=−2.32, p=0.808; rectangle: t(17)=−0.82, p=0.788) or 45° (square: t(17)=−2.83, p=0.840; rectangle: t(17)=−0.499, p=0.688). c) Average orientation consistency across all participants in left and right EC, depicted voxel-wise in the lateral-medial and posterior-anterior dimensions, after averaging across voxels in the superior-inferior dimension. The strongest reliable grid-like coding of visual space was found around the posterior-medial EC, which is believed to be homologous with dorsal-medial EC in the rodent (Schröder et al., 2015, eLife). Black colors denote regions outside of the group-level EC ROI. There was no significant difference in average EC grid-like coding between the left and right hemispheres (n=36; t-test, t(35)=0.97, p=0.340, two-tailed). Across participants, the magnitude of average grid-like coding was similar across hemispheres (Pearson correlation, r2=0.69, p=3.73x10−10, two-tailed). d) 60° periodic orientation consistency in EC was correlated with self-reported navigational ability (n=36 participants; Spearman rank correlation, rho=0.28, p=0.049), as assessed using the Santa Barbara Sense of Direction (SBSOD) questionnaire21. In contrast, orientation consistency for the non-grid 90° (rho=−0.10, p=0.719) or 45° (rho=−0.11, p=0.730) periodicities was not correlated with SBSOD. Higher SBSOD scores denote better navigators. Each dot represents one participant. Throughout the figure, all statistical tests are one-tailed unless otherwise noted. Error bars show ±1 SEM; * p < 0.05; ** p < 0.01.

Supplementary Figure 3 Grid-like coding of visual space is not confounded with biases in visual behavior.

a) Percentage of total gaze movements by direction in individual participants (top row) and on average across all participants (bottom row), across all scan runs for the square display participants and all upright-display runs for the rectangle display participants. Each possible gaze movement direction was sampled both within individual participants, and on average. b) Relative proportion of gaze movement directions modulo 90°, 60°, and 45°. Note that for the rectangle display participants, the proportion of gaze movement directions is computed here only for the upright display runs. Black lines denote means across all participants (n=36), and grey lines denote individual participants. Gaze tended to move disproportionately along the cardinal directions defined by the search display shape, as reflected by the peak centered around 0° in modulo 90° space. The presence of this 4-fold gaze direction bias could either reflect a common visual search strategy across participants, or could have been induced by the rectilinear shape of the displays. Critically, however, there was no 60° periodic bias in gaze movement directions that would be confounded with the presence of a specifically 60° periodic fMRI signal dependent on gaze movement direction. c) Participants searched all locations in the search displays in both the square and rectangle (upright and rotated) search displays. Average gaze sampling maps during gaze movement from all fMRI scan runs are shown. Participants largely constrained their search behavior to within the search display boundaries. The greater relative sampling around the center of each display was due to participants beginning each trial at a central fixation point. d) Difference in length of time with gaze movements aligned to each participant’s average EC grid orientation φ (within ±15° a φ axis) versus misaligned (more than ± 15° from all φ axes), plotted separately for the square and rectangular (upright and rotated) display participants (n=18 per group). There were significantly more gaze movements along misaligned than aligned directions for the square display participants (t-test; t(17)=2.99, p=0.008), though the magnitude of the difference was small relative to the scan run length (mean ± s.e.m.: 8.83 ± 2.95 seconds on average per scan run). There were no more gaze movements aligned than misaligned in the rectangular display participants, either upright (t(17)=0.95, p=0.360) or rotated (t(17)=1.63, p=0.122). e) Difference in gaze movement speed that passed the gaze motion speed threshold aligned versus misaligned to each individual participant’s average grid orientation, plotted separately for the square and rectangular display participants (n=18 per group). Gaze movement speed did not differ between aligned and misaligned for either the square or rectangle display subjects (t-tests: square: t(17)=0.44, p=0.666; rectangle upright: t(17)=0.55, p=0.588; rectangle rotated: t(17)=1.27, p=0.223). Throughout the figure, all statistical tests are two-tailed. All error bars show ±1 SEM. ** p < 0.01.

Supplementary Figure 4 Regional specificity of visual grid-like representations.

a) In a whole-brain analysis, we searched for voxels exhibiting sinusoidal modulation by gaze movement direction with 60° periodicity aligned to the EC grid orientation (n=36 participants; t-test). The results are plotted unmasked on the inflated cortical surface at an uncorrected statistical threshold for display purposes. No voxels survived whole-brain FWE correction. b) In an ROI analysis, we specifically tested for grid-like responses in ventral medial prefrontal cortex (vmPFC), a region previously shown to exhibit grid-like fMRI responses5,19 and located near frontal regions known to contain grid cells in humans6. We applied the split-half ROI analysis described in the main text to a bilateral vmPFC ROI, created based on a reverse inference meta-analysis for “vmPFC” using Neurosynth (thresholded at p<0.05, uncorrected) (Yarkoni et al., 2011, Nature methods). Left: Reliable grid-like coding of visual space was observed in right vmPFC (n=36; t-test, p<0.05, SVC in bilateral vmPFC; peak MNI coordinates: 7/50/19, peak Z=3.84). Right: Neither 90° nor 45° periodicities showed significant split-half orientation consistency (90°: data were not normally distributed, sign-test p=0.87; 45°: t-test, t(35)=0.46, p=0.228) in a 2mm spherical ROI centered on peak vmPFC grid-like coding (the magnitude of 60° consistency is plotted for comparison). c) Across participants, vmPFC and EC had similar grid orientations averaged across all scan runs (n=36; circular correlation, c=0.54, p=0.002, two-tailed), though grid orientations were not sufficiently similar to observe significant split-half reliable grid coding in vmPFC based on the EC grid orientation (i.e., Supplementary Fig. 4a). d) To confirm that the grid-like coding of visual space observed in EC was not due to a visual confound, we applied the same ROI analysis to a visual cortex ROI, which was anatomically defined for each participant (n=36) as the left and right occipital poles. Not surprisingly, EVC showed a strong response during periods of visual search compared to fixation (left; colored voxels are significant at p < 0.05, two-tailed, whole-brain FWE-corrected). There was no significant 60° (t-test, t(35)=1.05, p=0.151) or 45° (t(35)=1.22, p=0.116) periodic orientation consistency in EVC, but there was significant 90° (t(35)=2.58, p=0.007) periodic orientation consistency (right). Indeed, there was significantly greater 90° periodic orientation consistency in EVC than in EC (t(35)=2.43, p=0.020). The presence of a reliable 90° periodic signal in EVC likely reflects the cardinal direction gaze movement bias (Supplementary Fig. 3). Throughout the figure, all statistical tests are one-tailed unless otherwise noted. Error bars show ±1 SEM; ** p < 0.01; *** p < 0.001; ns, not significant.

Supplementary Figure 5 Clustering of grid orientations in entorhinal cortex (EC) within each participant.

Polar histograms show grid orientations (on the range 0º–60º) for all voxels in the individual subject bilateral EC ROIs, separately for each participant (one histogram per participant) for the square display (blue) and rectangular display (red) participants. Each histogram radius is scaled to the maximum number of voxels in a given bin for that histogram (range of max across subjects: 8%–51% of total EC voxels; the total number of functional voxels in EC ranged from 285–450 across participants). Grid orientations were significantly clustered across voxels in 28/36 participants (Critical value: * p < 0.05, Rayleigh’s test, adjusting degrees of freedom to account for spatial smoothing across voxels, on each participant). For the square participants, we also tested whether voxel-wise grid orientations were offset between 6° and 9° from the search display borders (Critical value: boxed-* p<0.05, V test for clustering around 6°–9° in 0.5° increments, Bonferroni corrected for 7 orientations tested).

Supplementary Figure 6 Whole-brain analysis of the rotation effect.

For the rotated-display runs in rectangle participants (n=18), we performed whole-brain analysis to identify voxels that exhibited greater 60° periodic sinusoidal modulation of the BOLD response for gaze movements aligned to φ+30° than for gaze movements aligned to φ (t-test, two-tailed). Results are shown unmasked on the inflated cortical surface at an uncorrected statistical threshold for display purposes. No voxels survived whole-brain FWE correction.

Supplementary information

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Julian, J.B., Keinath, A.T., Frazzetta, G. et al. Human entorhinal cortex represents visual space using a boundary-anchored grid. Nat Neurosci 21, 191–194 (2018). https://doi.org/10.1038/s41593-017-0049-1

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41593-017-0049-1

This article is cited by

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing