Neural Basis of Navigation in Urban Terrain

7/31/2019 Neural Basis of Navigation in Urban Terrain

1/4

electrode in the flowing bath solution, which wereconnected to a GeneClamp 500 amplifier (AxonInstruments, Foster City, CA) interfaced with acomputer. A drop of water at 0C, which was add-ed to the recording chamber at the end of eachexperiment, produced the expected large depolar-ization mediated by Ca2 and Cl channels [B. D.Lewis, C. Karlin-Neumann, R. W. Davis, E. P. Spal-ding, Plant Physiol. 114, 1327 (1997)], a test ofproper cell impalement.

15. D. J. Walker, R. A. Leigh, A. J. Miller, Proc. Natl.Acad. Sci. U.S.A. 93, 10510 (1996).

16. The shunt resistance between an inserted electrode

and the membrane permits a depolarizing current toflow[T. H.Goldsmith andM. H.M. Goldsmith, Planta143, 267 (1978)], the effect of which may be to un-derestimate V

mby 20 to 50 mV [W. Gassmann and

J. I. Schroeder, Plant Physiol. 105, 1399 (1994)].Corrections of that magnitude would shift each ofour V

mmeasurements, the most positive of which

was 193 mV, to values more negative than 230mV.

17. Plants were grown vertically on Petri plates contain-ing 1.5% (w/v) agar, 1 mM KCl, and 1 mM CaCl

2in

continuous light for 4 to 5 days. Root protoplastswere prepared by cutting the root about 150 mfrom the tip with a micromanipulator-mounted razor.

The cut seedlings were infiltrated with an enzymesolution containing Cellulysin (12 mg/ml) (Calbio-chem), Pectinase (2 mg/ml) (Sigma), and bovine se-rum albumin (5 mg/ml) (Sigma) dissolved in 10 mMKCl, 1 mM CaCl

2, 5 mM MES, and 300 mM sorbitol

{pH 5.2 with 1,3-bis[tris(hydroxymethyl)methyl-amino]propane (BTP)} using a vacuum produced bya faucet aspirator. After 2 hours of incubation, theseedlings were rinsed in the solution without en-zymes and stored at 4C or placed in a 500- lrecordingchamber containing 10 mM CaCl

2, 3 0 m M

KCl, 5 mM Hepes, and 120 to180 mM sorbitol (pH7.0 with BTP). Protoplasts of nonepidermal cellsemerged from thecut endof theotherwise undigest-ed root. Patch pipettes were filled with 130 mM K-glutamate, 2 mM EGTA, 5 mM Hepes, and 4 mMMgadenosine triphosphate (Mg-ATP) (pH 7.0 withBTP). Patch-clamp equipmentand procedureswereas described [M. H. Cho and E. P. Spalding Proc.Natl. Acad. Sci. U.S.A. 93, 8134 (1996)]. Our proce-dures led to observations of inward currents in allpatch-clamped wild-type cells. Lower percentagesreported by others (5) may be explained by differentgrowth conditions, differentvoltageprotocols, or use

of differenttissue.We encountered morevariability inthe voltage dependence of the inward currents thanhas been reported for heterologously expressed in-wardrectifiers [F.Gaymard et al.,J. Biol. Chem.271,22863 (1996)]. This variability is responsible for theappearance of weak rectification in the averagedwhole-cell I-V curve in Fig. 3E. Channel gating mayhavebeen affectedby cytoplasmic components thatwashed out in some patch-clamped protoplasts.

18. B. D. Lewis and E. P. Spalding, data not shown.19. Patch pipettes were filled with 390 mM K-gluta-

mate, 2 mM EGTA, 5 mM Hepes, and 4 mM Mg-ATP (pH 7.0 with BTP). The bath contained 130mM KCl, 10 mM CaCl

2, and 5 mM Hepes (pH 7.0

with BTP). Tail currents were evoked by clampingVm

at 200 mV before stepping to a series of morepositive potentials.

20. Plantswere grown in continuous light on media con-taining KCl at the concentrations indicated plus

0.8% (w/v) agarose (Bio-Rad, Hercules, CA), 0.5%(w/v) sucrose, 2.5 mM NaNO

3, 2.5 mM CaNO

3, 2

mM NH4

H2

PO4

, 2 mM MgSO4

, 0.1 mM FeNaEDTA,25 M CaCl

2, 25 M H

3BO

3, 2 M ZnSO

4, 2 M

MnSO4

, 0.5M CuSO4

, 0.2M Na2

MoO4

, and0.01M CoCl

2, adjusted to pH 5.7 with NaOH.

21. Of 104 seedlings tested, 23 were phenotypicallyscored as mutant and 88 were scored as wild type,giving x2 0.346 (based on the expected ratio ofthree wild type to one mutant); P 0.05.

22. Wild-type and akt1-1 plants were grown on agarcontaining MS salts and 5% (w/v) sucrose in con-stant lightfor 14 days.Roots wereexcised,weighed,and washed for 16 to 17 hours in a K-free desorp-tion solution (DS) containing 1.5 mM CaCl

2adjusted

to pH 5.7 with CaOH. Roots were incubated for 10min in DS containing 4 mM NH

4Cl and 10 M, 100

M,or1 mMRb(86Rb)Cland then washedtwotimesfor 10 min each in ice-cold DS. Radioactivity wasmeasured by detection of Cerenkov radiation.

23. J. I. Schroeder, J. M. Ward, W. Gassmann, Annu.Rev. Biophys. Biomol. Struct. 23, 441 (1994); W.Gassmann, J. M. Ward, J. I. Schroeder, Plant Cell5,1491 (1993).

24. Y. Cao, A. D. M. Glass, N. M. Crawford, PlantPhysiol. 102, 983 (1993); F. R. Vale, W. A. Jackson,R.J. Volk,ibid.84, 1416 (1987); F.R. Vale,R. J.Volk,W. A. Jackson, Planta 173, 424 (1988).

25. Supported by the Department of Energy (DOE)/NUSDA Collaborative Research in Plant Biology Pgram (BIR-9220331), funds to R.E.H. from the NUniversity of Wisconsin Cellular and Molecular Biol

Training Grant (GM07215), to M.R.S. from DOE (DFG02-88ER13938) and NSF (DCB-90-04068), andE.P.S. from the NASA/NSF Network for ResearchPlant Sensory Systems (IBN-9416016). We thankJanczewski and R. Meister for technical assistanand P. Krysan, W. Robertson, J. Satterlee, and

Young for helpful comments on the manuscript.

26 November 1997; accepted 27 March 1998

Knowing Where and Getting There: A HumanNavigation Network

Eleanor A. Maguire,* Neil Burgess, James G. Donnett,Richard S. J. Frackowiak, Christopher D. Frith, John OKeefe

The neural basis of navigation by humans was investigated with functional neuroimagiof brain activity during navigation in a familiar, yet complex virtual reality town. Activatiof the right hippocampus was strongly associated with knowing accurately where placwere located and navigating accurately between them. Getting to those places quickwas strongly associated with activation of therightcaudate nucleus. These tworight-si

brain structures function in the context of associated activity in right inferior parietal anbilateral medial parietal regions that support egocentric movement through the virtutown, and activity in other left-side regions (hippocampus, frontal cortex) probabinvolved in nonspatial aspects of navigation. These findings outline a network of braareas that support navigation in humans and link the functions of these regions physiological observations in other mammals.

Where am I? Where are other places in theenvironment? How do I get there? Ques-tions such as these reflect the essentialfunctions of a navigation system. The neu-ral basis of way-finding activity has beenextensively studied. Spatially tuned neurons

found in the hippocampal formation of free-

ly moving rats [place cells coding for therats location (1) and head direction cellscoding for its orientation (2)] support theidea that this part of the brain provides anallocentric (world-centered) representationof locations, or cognitive map (3). Theposterior parietal lobe has been implicatedin providing complementary egocentricrepresentations of locations (centered onparts of the body) (4). Other brain regions,such as the dorsal striatum (5), have alsobeen identified as possible elements of anavigation system. In humans, there has

been much evidence for the involvement ofthe hippocampus in episodic memory, thememory for events set in their spatio-tem-

poral context (3, 6). By contrast, the rolethe hippocampus in human navigation hremained controversial, and the wider neral network supporting human navigationeven less well understood. We attack thissue by combining functional neuroima

ing with a quantitative characterizationhuman navigation within a complex virtureality environment.

We used positron emission tomograp(PET) (7) to scan subjects while they naigated to locations in a familiar virtual ality town using their internal represention of the town built up during a contuous period of exploration immediately bfore scanning (Fig. 1A). In one navigaticondition, the subjects could head directoward the goal (nav1), while in the oth(nav2), direct routes were precluded closing some of the doors and placing

barrier to block one of the roads, forcing tsubjects to take detours. Navigation wcompared to a task in which subjects movthrough the town following a trail of arrowthus not needing to refer to an internrepresentation of the town. An additiontask requiring the identification of featuin static scenes from the town was includfor contrast with the three dynamic tas(8).

We first investigated which brain gions were involved in successful navig

E. A. Maguire, R. S. J. Frackowiak, C. D. Frith, WellcomeDepartment of Cognitive Neurology, Institute of Neurolo-gy, University College London, 12 Queen Square, Lon-don WC1N 3BG, UK.N. Burgess, J. G. Donnett, J. OKeefe, Department of

Anatomy and Developmental Biology and Institute ofCognitive Neuroscience, University College London,Gower Street, London WC1E 6BT, UK.

*To whom correspondence should be addressed. E-mail:[email protected]

REPORT

www.sciencemag.org SCIENCE VOL. 280 8 MAY 1998 9


2/4

tion in both direct and detour way-finding.The trials in these conditions were dividedinto successful ones in which the correctdestination was reached (22/30 trials innav1 and 21/30 in nav2 across the 10 sub-jects) and those in which it was not. Thesuccessful trials compared to the arrows taskshowed significant activation of the righthippocampus (Fig. 1B), as did the compar-ison of the successful trials with the unsuc-cessful trials. This latter comparison alsorevealed activation in the left hippocampus,left lateral temporal cortex, left frontal cor-tex, and in the thalamus (Fig. 1B).

In order to explore in greater depth therelationship between regional cerebralblood flow (rCBF) and behavior during di-rect way-finding (nav1), we derived a quan-

titative measure of the accuracy of headingtoward the goal (9). The accuracy of head-ing measured across all trials in nav1 covar-ied significantly with rCBF in the righthippocampus and the right inferior parietalcortex (Fig. 2). Not only is the right hip-pocampus more active during navigationthan trail-following, but the more accuratethe navigation, the more active it is.

These results are consistent with ourinterpretation that the right hippocampusand inferior parietal cortex cooperate toenable navigation to an unseen goal: Thehippocampus provides an allocentric (envi-ronment-based) representation of spacethat allows the computation of the direc-tion from any start location to any goallocation, and the right inferior parietal cor-tex uses this information to compute thecorrect body turns to enable movement to-ward the goal given the relative (egocen-tric, body-centered) location of obstacles inthe way (doorways to go through, barriersacross roads, and so forth) and the current

heading direction. Because the parietal cor-tex takes account of information in addi-

tion to the allocentric direction to the goal,it would not have as high a correlation withthe accuracy of heading toward the goal asthe hippocampal formation (consistentwith our findings; Fig. 2). Similarly, rCBFin the right inferior parietal cortex wouldnot be significantly different in the trail-

following and way-finding conditions, be-cause both tasks have similar egocentricrequirements, and no such difference wafound (Fig. 1B). However, differences inright inferior parietal activity would be ex-pected when subtracting the static condi-tion from either trail-following or way-find-ing. This comparison did indeed show righinferior parietal activation, along with bi-lateral activation of medial parietal areaswhich we assume are also involved in ego

-

centric aspects of movement, for exampleprocessing the optic flow generated by themovement (Fig. 3B) (10).

Activity in the left hippocampus, al-though associated with successful naviga-tion, does not covary significantly with oumeasure of the accuracy of navigation. Weinterpret this as a role in actively maintain-ing the memory trace of the appropriatedestination during navigation or recollect-ing specific paths taken during learning thatlead to the goal but are not necessarilydirect. Either role would be consistent with

the known involvement of the left hip-pocampus in episodic memory for person-

ally experienced events (6).These results are consistent with previ-

ous reports of the involvement of the hip-pocampal or parietal areas in topographicamemory (11) and provide a more preciseinterpretation of their roles in the actuaperformance of navigation.

A

B

Fig. 1. (A) Example of view from inside the virtual

town. (B) The comparison between successful

navigation (nav1 and nav2) compared to following

a trail of arrows. PET data are superimposed onto

theaveraged MRI of the10 subjects at thevoxel of

peak activation in the right hippocampus dis-

played in the coronal plane. The color scale of the

activation pertains to the significance level of the z

scores with the peak of the activation in white.

Coordinates in stereotactic space (x,y,z, respec-

tively) and z scores of the activations are: right

hippocampus (30, 16, 22; z 3.74) and left

tail of caudate (28, 16, 28; z 3.05). Otherareas activated in this comparison but not dis-

played on this plane were: left occipital area 18

(24, 102, 2; z 3.50) and left superior fron-tal gyrus (22, 52, 22;z 3.65). The comparisonbetween successful versus lost trials showed the

following activations: right hippocampus (30,

20, 16; z 3.61); left hippocampus (16,26, 6; z 4.10); left superior temporal gyrus(54, 30, 14; z 3.86); left inferior temporalgyrus (52, 50, 12; z 5.23); left inferior fron-tal gyrus (46, 22, 6;z 4.59); and rightthalamus(6, 6, 12; z 4.25).

A B

C

Fig. 2. (A) The virtual en-

vironment is shown from

an aerial perspective,

demonstrating the com-

plexity of the town andthe many possible paths

between the various

places. Subjects navi-

gation during scanning

on the navigation task

was analyzed in terms of

accuracy in degrees (9).

Three trajectories be-

tween the screens atA and B (18 m

apart) are shown from the range of

subjects behavior: an accurate tra-

jectory (yellow, accuracy 155.3),an inaccurate but successful trajec-

tory (green, accuracy 127.9),and an inaccurate lost trajectory

(red, accuracy 70.6). (B) ThePET data from the correlational

analysis of rCBF (22) and accuracy

are superimposed onto the aver-

aged MRI of the 10 subjects at the

voxel of peak activation in the right

hippocampus displayed in the

coronal plane. Areas activated in this comparison were right hippocampus (36, 12, 20; z 3.66displayed in figure) and right inferior parietal cortex (60, 30, 50; z 3.36). (C) Scatter plot of thecorrelation of rCBF values at the voxel of peak activation in the right hippocampus plotted against the

accuracy of navigation (r 0.56, P 0.002). The behavioral data for one trial of one subject (subject 6)was not available. The data points for each subject are plotted in different colors. The correlation of

accuracy of navigation and perfusion in the right inferior parietal cortex was r 0.43, P 0.02.

SCIENCE VOL. 280 8 MAY 1998 www.sciencemag.org922


3/4

Next, we looked at successful navigationrequiring detours (nav2) compared to suc-cessful navigation in the nav1 condition.This comparison revealed left frontal acti-vation (Fig. 3A). The increased require-ment for strategy switching in the presenceof obstacles (nav2) compared to direct way-finding (nav1) is consistent with f indings offrontal involvement in other studies withtasks making similar demands (12). Leftfrontal activation was also apparent whensuccessful navigation was compared to fol-lowing arrows or compared to unsuccessfulnavigation. These frontal activations areconsistent with a role for this region inplanning and decision making (13). It islikely that following the trail of arrows de-mands less planning than way-finding.

As well as comparing the dynamic andstatic tasks detailed above, we further char-acterized movement in the town in terms ofthe speed of motion, that is, the ratio be-tween distance traveled and time taken,producing an average speed measure in vir-

tual meters per second. In contrast to theareas whose activity correlated with naviga-

tional accuracy, the only area of regionalactivation that covaried significantly withspeed of navigation in the nav1 conditionwas in the right caudate nucleus; rCBF inthis region increased as speed increased (Fig3, C and D). This correlation with speed ofvirtual navigation was much more signifi-cant than that with simple motor responsevariables such as the rate or average dura-tion of keypad presses (Fig. 3D). It suggestsa higher function than the simple control ofthe physical movement of parts of the body,although its precise interpretation remainsopen. The dorsomedial caudate region re-ceives projections from the cortices adja-cent to the hippocampus in rats (14). Wesuggest that location within the environ-ment or spatial context might provide animportant source of information for the stri-atal control of higher-level aspects of cur-rent or planned movements and that thiscontrol is reflected in the amplitude orspeed of movement, rather than the direc-tion of movement.

In conclusion, our results outline thenetwork of brain regions supporting human

navigation and suggest specific roles each of these regions. They agree with, afurther illuminate, previous findings shoing that lesions of the right human hpocampus result in deficits of spatial meory (15) while those of the right inferparietal cortex result in deficits of the abity to represent or act on objects locatwith respect to the egocentric left-rigbody axis (16). Our interpretation of tparietal role in navigation agrees with neronal responses from inferior parietal cortin monkeys (in particular, area 7a and tlateral intraparietal area) implicating itthe translation of the location of stimfrom retinal to head- or body-centered cordinates (4), and with the connectionsarea 7a to the hippocampal formation [icluding the presubiculum which, at leastrats, codes for the current head directi(2)]. Our interpretation of the hippocamprole in navigation is concordant with neurnal responses in rats (3) and with modelshow the hippocampus guides rats navigati

(17), from which our measure of navigatioal accuracy was explicitly derived. Our fining that rCBF in the right caudate nuclecorrelates with the speed of navigationcompatible with its proposed role in motlearning (18) and the process by whimovements are reinforced [and hence, toccurrence of abulia after lesions of this gion (19)], and also with the more generhypothesis of a role in context recogniti(20). It also has relevance to the suggestiothat rats may use signals derived from cethat encode their speed of movement determine distances and that such speed cemight be located in one of the sub-corti

nuclei (21), perhaps in the basal gangliawe identified here. Although many detailsthe inputs and outputs of a human navigtion network remain to be specified, we hademonstrated the closest link yet betwehumans and other mammals in the neurimplementation of navigation.

REFERENCES AND NOTES___________________________

1. J. OKeefe and J. Dostrovsky, Brain Res. 34, 1(1971).

2. J. S. Taube, R. U. Muller, J. B. Ranck, J. Neuros10, 420 (1990).

3. J. OKeefe and L. Nadel, The Hippocampus aCognitive Map (Clarendon, Oxford, 1978).

4. R.A. Anderson, Philos. Trans.R. Soc. LondonSe352, 1421 (1997); H. O. Karnath, ibid., p. 1411;

Vallar, ibid., p. 1401.

5. S. I. Wiener, J. Neurosci. 13, 3802 (1993); B.Devan, E. H. Goad, H. L. Petri, Neurobiol. LeaMem. 66, 305 (1996); M. Potegal, Spatial AbilitiDevelopment and Physiological Foundations (Ademic Press, New York, 1982).

6. M. Kinsbourne andF. Wood, in Short-Term MemoD. Deutsch and J. A. Deutsch, Eds. (AcademPress, New York, 1975), pp. 257291; E. TulviElements of Episodic Memory (Clarendon, Oxfo1983); F. Vargha-Khadem et al., Science 277, 3(1997); B. Milner, Clin. Neurosurg. 19, 421 (1972

7. Ten healthy right-handed male subjects (mean a

A B

D

C

Fig. 3. (A) Activation of the left mid-

dle and superior frontal gyri when

successful detour navigation (nav2)

is compared to successful direct

(nav1) navigation, superimposed

onto the averaged MRI of the 10

subjects displayed in the transverse

plane. Areas activated in this com-

parison are: left superior frontal gy-

rus (10, 66, 8; z 3.57; displayedin figure); left middle frontal gyrus

(16, 26, 40; z 3.57); and rightcerebellum (24, 74, 46; z 4.10). (B) The SPM in the coronal

plane associated with comparisonof movement tasks (nav1, nav2, arrows) with the static scenes task. Displayed on a transparent brain to

facilitate viewing of all significant activations that are on different planes. Areas activated: left medial

parietal cortex (16,52,54;z 4.48); rightmedial parietal cortex (12,68,48;z 4.02); rightinferiorparietal lobe (56, 38, 32; z 6.30); left cerebellum (34, 40, 40; z 3.89); and right cerebellum(49, 36, 38; z 4.22). (C) Activity in the right caudate covaried significantly with speed of virtualmovement,displayed here in sagittal section on theaveragedMRI of the10 subjects at thevoxel of peak

activation in the right caudate (10, 14, 4; z 3.97). (D) Scatter plot of the correlation of rCBF (22)values at the voxel of peak activation in the right caudate plotted against the speed of navigation (r0.75, P 0.0001). Thebehavioral data for onetrialof onesubject (subject 6) wasnot available. Thedatapoints for each subject are plotted in different colors. The correlations between right caudate rCBF and

peripheral motor behaviors were much less significant (average duration of button press r 0.08, P0.67; rate of button pressing r 0.51, P 0.004).

REPORT

www.sciencemag.org SCIENCE VOL. 280 8 MAY 1998 9


4/4

36.5 years) participated in the study, approved bythe local hospital ethics committee and the UK Ad-ministration of Radioactive Substances AdvisoryCommittee. Each subject had 12 PET scans. Therewere four tasks each performed three times in acounterbalanced order. PET scans were obtainedusing a Siemens/CPS ECAT EXACT HR (model962) PET scanner. Scanning was performed withsepta retracted,in three-dimensional (3D)mode. Thefield of view of 15.5 cm in theaxialextent allowed thewhole brain to be studied. Volunteers received anH215O intravenous bolus (330 MBq) infused over

20s followedby a 20-s salineflush througha forearm

cannula. Data were acquired in a 90-s scan frame.There were 12 successive administrations of H

2

15O,each 8 min apart. The integrated radioactivity countsthat accumulated over the 90-s acquisition period,corrected for background, were used as an index ofregional cerebral blood flow. Attenuation correctionwas computed with a transmission scan beforeemissionscan acquisition. Images werereconstruct-ed into 128 pixels by 128 pixels in 63 planes with anin-plane resolution of 6.5 mm. In addition, high-res-olution magnetic resonance imaging (MRI) scanswere obtained with a 2.0-T Vision system (SiemensGmbH, Erlangen, Germany) using a T1-weighted 3Dgradient echo sequence. The image dimensionswere 256 voxels by 256 voxels by 256 voxels. Thevoxel sizewas1 mmby 1 mmby 2 mm. Imageswereanalyzed with Statistical Parametric Mapping(SPM96, Wellcome Department of Cognitive Neu-rology, London, UK; www.fil.ion.ucl.ac.uk) executed

in MATLAB (Mathworks, Sherborn, MA). All scanswere automatically realigned to the first scan andthen normalized using a nonlinear deformation [K. J.Friston et al., Hum. Brain Mapp. 2, 189 (1995)] intostandard stereotactic space [J. Talairach and P.

Tournoux, Co-Planar Stereotactic Atlas of the Hu-man Brain ( Thieme, Stuttgart, Germany, 1988)] us-ing a template from the Montreal Neurological Insti-tute [A. C. Evans et al., in Proceedings of the IEEE-Nuclear Science Symposium and Medical ImagingConference, San Francisco, CA, 31 October to 6November 1993, L. A. Klaisner, Ed. (IEEE ServiceCenter, Piscataway, NJ,1993), pp.18131817]. Thestructural MRI scans were normalized into the samespace to allow for the superimposition of PET acti-vations onto an averaged structural image. Imagesweresmoothed usingan isotropicGaussiankernel of16 mm (full width at half maximum) to optimize thesignal-to-noise ratio and to adjust for intersubject

differences in gyral anatomy. Global variance be-tween conditions was removed, using analysis ofcovariance (ANCOVA). For each pixel in stereotacticspace, condition-specific adjusted rCBF values withan associated adjusted error variance were generat-ed. Areas of significant change in brain activity werethen determined, using appropriately weighted con-trasts between the task-specific scans and the tsta-tistic. The resulting sets of t values constituted the

statistical parametric map (SPM). Significance levelswere set at P 0.001 (uncorrected).

8. A commercially available computer game (DukeNukem 3D, 3D Realms Entertainment, Apogee Soft-ware Ltd., Garland, TX ) was used to present thevirtual reality town on a 120-MHz Pentium-basedpersonal computer, showing a color, 3D, fully tex-tured first-person view. The town was created usingthe editor provided (BUILD, Ken Silverman, 3DRealms Entertainment). The games recordand play-back functions were used to store subjects actionsand replay them for subsequent analysis. The townhad four streets and contained shops, bars, a cine-

ma, church, bank, train station, and video gamesarcade (Fig. 1A). Subjects could enter into and nav-igate through the buildings, as each room had atleast two entrances. The town contained smallscreens on the walls at various locations. Approach-inga screenand switchingit on causedit to display aview of another part of the town. Subjects controlledtheir movement within the environment by using akeypad with backward, forward, left turn, and rightturn buttons. A fifth button served to activatescreens. Before scanning acquisition, subjectsspentup to 60 min exploring the environment until they feltthat they hadlearned thespatiallayout of streets andbuilding interiors. A trail of arrows on the floor waspresent during exploration and in all conditions, butwas only relevant in the arrows condition. Subjectswere scanned under four conditions. (i) nav1: sub-

jects switch on a screen and navigate through thetownto the destination displayed. Whenthe destina-

tion is reached, the subject activates the screenfoundthere, whichdisplaysthe nextdestination,andso on; (ii) nav2: identical to nav1, except that somedoors have been closed, and a barrier has beenmoved to block a different street; (iii) arrows: sub-

jects move through the town following a trail of ar-rows on the floor. Subjects activated the screensencountered during the task, but the views of thetown displayed had no relevance to their task; (iv)scenes: static scenes from the town are presentedevery 2 s, subjects respond according to whetherthere is a screen in the scene or not.

9. For the nav1 condition, the subjects heading direc-tion

h(in degrees) and the direction of the current

destination d

were calculated at each meter alongthesubjects trajectory. The absolutedifference

h

d wasfound at each meter andwas averaged over

a trial to give A 180 h

d as a measure of

accuracy of heading. Similar analysis was not ap-

plied to the nav2 detour condition, because subjectsdid not know beforehand which doors would beclosed or where the barriers would be, and so couldnot be expected to plan an optimal route. Theoreti-cally, accuracy scores may vary from 0 (always mov-ing directly away from the current destination),through 90 (moving randomly)to 180 (always mov-ing directly toward the current destination). In prac-tice, an accuracy of 160 is hard to exceed because

of the cluttered nature of the environment (the accu-racy of one very well-practiced author, N.B., in thethree trials varied between 144.3 and 157.4). Thismeasure agrees with our subjective assessment otrials, was independent of the speed of navigationand is consistent with models of how the hippocampus directs navigation in rodents (17).

10. B. M. DeJong, S. Shipp, B. Skidmore, R. S. J. Frackowiak, S. Zeki, Brain 117, 1039 (1994).

11. G. K. Aguirre, J. A. Detre, D. C. Alsop, MDEsposito, Cereb. Cortex 6, 823 (1996); E. AMaguire, R. S. J. Frackowiak, C. D. Frith, Proc. RSoc. London Ser. B 263, 1745 (1996); O. Ghaem e

al., Neuroreport8, 739 (1997); E.A. Maguire,R. S.JFrackowiak, C. D. Frith. J. Neurosci. 17, 7103(1997); G. K. Aguirre and M. DEsposito, ibid., p2512; E. A. Maguire, C. D. Frith, N. Burgess, J. GDonnett, J. OKeefe, J. Cogn. Neurosci. 10, 61(1998).

12. A. Bechara, D. Tranel, H. Damasio, A. R. DamasioCereb. Cortex6, 215 (1996); R. Elliott, C. D. FrithR. J. Dolan, Neuropsychologia 35, 1395 (1997).

13. A. R. Luria, Higher Cortical Functions in Man ( Tavistock, London, 1966); B. Milner, Arch. Neurol. 9, 90(1963); T. Shallice, Philos. Trans. R. Soc. LondonSer. B 298, 199 (1982).

14. A.E. Kelly andV. B.Domesick, Neuroscience7, 615(1982); L.W. Swanson andC. Kohler,J. Neurosci.63010 (1986); A. J. McGeorge and R. L. M. FaullNeuroscience 29, 503 (1989).

15. E. A. Maguire, T. Burke, J. Phillips, H. StauntonNeuropsychologia 34, 993 (1996); M. L. Smith andB. Milner, ibid. 19, 781 (1981).

16. N. Burgess, J. G. Donnett, K. J. Jeffery, J. OKeefePhilos. Trans. R. Soc. London Ser. B 352, 1397(1997).

17. J. OKeefe, Brain and Space, J. Paillard, Ed. (OxfordUniv. Press, Oxford, 1991), pp. 273295; N. Bur-gess and J. OKeefe, Hippocampus 6, 749 (1996).

18. R. E. Passingham, in Neural and Behavioral Approaches to Higher Brain Functions, S. Wise, Ed( Wiley, New York, 1987); R. E. Passingham, TheFrontal Lobes and Voluntary Action (Oxford UnivPress, Oxford, 1995).

19. K. P. Bhatia and C. D. Marsden, Brain 117, 859(1994).

20. J. C. Houk and S. P. Wise, Cereb. Cortex 5, 95(1995).

21. J. OKeefe, N. Burgess, J. Donnett, E. A. MaguirePhilos. Trans. R. Soc. London Ser. B, in press.

22. C. Buchel, R. S. J. Wise, C. J. Mummery, J. BPoline, K. J. Friston, Neuroimage 4, 60 (1996).

23. E.A.M., R.S.J.F., and C.D.F. are supported by theWellcome Trust, J.OK. and J.G.D. by the MedicaResearch Council, and N.B. by the Royal SocietyWe thank K. Friston, C. Price, and C. Buchel fohelpful comments.

11 December 1997; accepted 20 March 1998

SCIENCE VOL. 280 8 MAY 1998 www.sciencemag.org924

Neural Basis of Navigation in Urban Terrain

Documents

Transcript of Neural Basis of Navigation in Urban Terrain