The contribution of virtual reality to the diagnosis of spatial navigation disorders and to the study of the role of navigational aids: A systematic literature review.

Review

The contribution of virtual reality to the diagnosis of spatial navigation disorders and to the study of the role of navigational aids: A systematic literature review

M. Cogne´

*, M. Taillade

, B. N’Kaoua

, A. Tarruella

, E. Klinger

, F. Larrue

, H. Sauze´on

, P.-A. Joseph

, E. Sorita

aEA4136handicapetsyste`menerveux,universite´ deBordeaux,33076Bordeaux,France

bServicedeme´decinephysiqueetdere´adaptation,centrehospitalieruniversitaire,33076Bordeaux,France

cInstitutnationalderechercheeninformatiqueetautomatique(INRIA),33405Talencecedex,France

dInstitutdeformationenergothe´rapie,centrehospitalieruniversitaire,33076Bordeaux,France

eLaboratoireinteractionsnume´riquessante´ handicap,ESIEA,53000Laval,France

fLaboratoireBordelaisderechercheeninformatique(LaBRI),universite´ deBordeaux,33045Bordeaux,France

1. Introduction

Findingone’swayaroundanenvironment,whetherornotitis familiar,isacommonsituationforhumans.Peoplefrequentlyneed tomovefromonepointtoanother,accordingtoatrajectorythat mayormaynotbeknowninadvance,usingdifferentelements– externalaids,landmarksorothersourcesofrelevantinformation.

This ability is essential, because it underpins the person’s autonomyandhisorherparticipationinsociety.Spatialnavigation hasraisedinterest,andhasbeenthesubjectofmuchresearchin mammals (including humans) on account of its particular integrativeroleinlinkingfunctionsintheareasofneurophysiolo- gy,learning,memoryandcognition[1].

Foranumberofyears,numerousclinicaldescriptionsofbrain- injuredpatients presentingselectivelossesoftheabilitytofind theirwayhavebeenundertaken.Thisdisorderhasbeentermed

‘‘topographicaldisorientation’’[2].Butthereiswidediversityin the types of topographical disorientation observed in these ARTICLE INFO

Articlehistory:

Received9April2015 Accepted23December2015 Keywords:

Spatialnavigation Virtualreality Brain-damagedpatients Ageing

Stimuli Aid

ABSTRACT

Introduction:Spatialnavigation,whichinvolveshighercognitivefunctions,isfrequentlyimplementedin dailyactivities,andiscriticaltotheparticipationofhumanbeingsinmainstreamenvironments.Virtual realityisanexpandingtool,whichenablesononehandtheassessmentofthecognitivefunctions involvedinspatialnavigation,andontheothertherehabilitationofpatientswithspatialnavigation difficulties.Topographicaldisorientationisafrequentdeficitamongpatientssufferingfromneurological diseases. The use of virtual environments enables the information incorporated into the virtual environmenttobemanipulatedempirically.Buttheimpactofmanipulationsseemsdifferaccordingto theirnature(quantity,occurrence,andcharacteristicsofthestimuli)andthetargetpopulation.

Methods:Weperformedasystematicreviewofresearchonvirtualspatialnavigationcoveringtheperiod from2005to2015.Wefocusedfirstonthecontributionofvirtualspatialnavigationforpatientswith braininjuryorschizophrenia,orinthecontextofageinganddementia,andthenontheimpactofvisual orauditorystimulionvirtualspatialnavigation.

Results:Onthebasisof6521abstractsidentifiedin2databases(PubmedandScopus)withthekeywords

«navigation»and«virtual»,1103abstractswereselectedbyaddingthekeywords‘‘ageing’’,‘‘dementia’’,

‘‘braininjury’’,‘‘stroke’’,‘‘schizophrenia’’,‘‘aid’’,‘‘help’’,‘‘stimulus’’and‘‘cue’’;Amongthese,63articles wereincludedinthepresentqualitativeanalysis.

Conclusion: Unlikepencil-and-paper tests, virtual reality is useful to assess large-scalenavigation strategiesinpatientswithbraininjuryorschizophrenia,orinthecontextofageinganddementia.Better knowledgeaboutboththeimpactofthedifferentaidsandthecognitiveprocessesinvolvedisessential fortheuseofaidsinneurorehabilitation.

ß2016PublishedbyElsevierMassonSAS.

* Correspondingauthor.EA4136handicapetsyste`menerveux,universite´ de Bordeaux,33076Bordeaux,France.

E-mailaddress:melaniecogne@hotmail.fr(M.Cogne´).

Availableonlineat

ScienceDirect

www.sciencedirect.com

http://dx.doi.org/10.1016/j.rehab.2015.12.004 1877-0657/ß2016PublishedbyElsevierMassonSAS.

patients, and also in the type of brain injury incurred [3]. In addition,theevaluationsusedtocharacterisethedisordershave beendiverseandnon-standardised[4].

Tostartwith,testsofspatialcognitiononasmallscalewere elaborated,intheformof‘‘pencil-and-paper’’tests,suchasthe Rotation Mental Test [5] or the Guilford and Zimmermann spatialorientation test. Hegartyet al. [6]showed onlypartial correlation between paper-and-pencil visuo-spatial tests on a smallscaleandtestsinlarge-scalevirtualorrealenvironments.

Spatialnavigationin large spaces thusseems torequire skills otherthanthoseassessedinpaper-and-penciltests.Thisstudy concludedtopoorecologicalvalidityofpencil-and-papertests, andconfirmedtheusefulnessofassessingspatialnavigationina large-scaleenvironment,whetherrealofvirtual.Indeed,several authors, among whom Cushman et al. [7], have shown a correlationinnavigationperformancesbetweenrealandvirtual environments.

The progressincomputing andits widespread usefromthe 1990shaveenabledthemodellingofsimulatedenvironmentsthat areclosetoreality,knownasvirtualrealityenvironments.Virtual realityrapidly demonstrated its potentialusefulness in clinical practice,in particular in the areas of cognitiveevaluation and rehabilitation,andwellasforthestudyofcognitiveprocessesin humans[4].Theevaluationofnavigationinavirtualenvironment hasvariousadvantages overevaluation inrealsituations [8]: it enablesbetterscientificmonitoring,itcanallowfordeficitsand disabilitiesthatsometimesmake evaluationproblematicinreal situations,itislesscostly,andalsosafer.Fromafunctionaland behaviouralviewpoint,itenablesimmediatefeedbackonperfor- mancesinvarioussensoryformsandmodes,controlledadminis- tration of cues so as to improve performances in a learning approach, and it also enables ‘‘time-out’’ to make room for discussionandtheinclusionofdifferentmethods.

Followingthepublicationofnumerousstudiesthathaveshown theadvantagesofthevirtualsettingintheareaofspatialcognition, itseemedimportanttoustoperformanup-to-datereviewonthe contributionsof the useof virtualreality in theexploration of spatial navigation, especially in populations with cognitive impairments. These patients are particularly likely to present spatialcognitiondisorders,whicharehighlydamagingforsocial participationintheirordinaryenvironments[2,3].Inaddition,the clinicaluseofvirtualenvironments,becauseitenablesmorestrict experimentalcontrolthanworkinginarealenvironment,makesit possibletomeasuretheinfluenceofdifferentcomponentsbrought intoplayonspatialnavigationperformances.Thusenvironmental cuesareusedassensorycuesofthebottom-uptype,andprovide scopeforinvestigation,wheretheuseofvirtualrealityhasmuchto contribute.Thereviewpresentedherehastwoobjectives:thefirst istodetailtheclinicaluseofvirtualrealityinspatialnavigationin populationswithcognitivedisorders;thesecondistoprovidean updateontheliteratureontheimpactonvirtualspatialnavigation ofusingvisualandauditoryenvironmentalcues.

2. Methods

Thisreviewwasconductedontwomedicaldatabases,PubMed andScopus(fortheScopusdatabase,onlyarticlesintheareaofthe health sciences were searched). The following Mesh keywords wereused:‘‘navigation’’, ‘‘virtual’’, ‘‘ageing’’,‘‘dementia’’,‘‘brain injury’’, ‘‘schizophrenia’’, ‘‘stroke’’, ‘‘cue’’, ‘‘help’’, ‘‘aid’’, and

‘‘stimulus’’. The basic search algorithm was ‘‘navigation AND virtual’’towhichalltheotherkeywordswereassociatedinturn.

The processing of results was performed according to Prisma guidelines.Articles publishedin English in the area of the life sciencesfrom2005toJune2015wereretained.Thearticleswere

selectedonthebasisoftitleandabstract.Thefinalselectionwas discussedbythethreeauthorsofthepresentarticle(MC,PAJ,ES) withrespecttothetwoobjectivesdefinedabove.Aflowdiagram (Fig.1)illustratesthearticleselectionprocedure.

Inall,44articlescomplyingwiththefirstobjectiveand19with thesecondobjective wereretained. Themaincharacteristics of thesearticlesaresummedupinTables1and2usingthefollowing characteristics:firstauthor,virtualenvironmentused,population (numbersandtype)studyobjectiveandmainresult.

3. Results

3.1. Contributionsofspatialnavigationinvirtualrealityindifferent populationspresentingcognitivedisturbances

3.1.1. Spatialnavigationandnormalageing

Theeffectofageingonnavigationskillsinlargespaceshasbeen widely described [9]. Generally speaking, the remembrance of spatial configurations and landmarks during navigation loses precisionwithage[10],whileparadoxically,learningaroutefrom an aerial view appears, according to certain authors, to be preserved [11]. More specifically, different studies in virtual realitysettinghaveshownalterationsintheallocentricprocessing ofinformation(i.e.spatialconfigurations)amongelderlysubjects usingavirtualwatermaze[12],othermazemodels(wallmaze, radial maze)[13–16],or virtualcity districts [17,18].Similarly, otherstudiesusingmazesorothervirtualenvironmentsevidence difficultiesinacquiringspatialrepresentationsthoughttoresult from an age-linked decline in hippocampus function, which appears to spontaneously direct elderly subjects towards the useofegocentricstrategies[19–23].Arecentstudycomparinga virtualandarealcitydistricthowevershowedthatelderlysubjects encountergreaterdifficultiesthanyoungsubjectsinfindingtheir way,thisbeingrelatedtodeclineinexecutivefunctions,whilethe spatial knowledgeacquired in the virtualenvironment didnot significantly differ between the two groups. This age-linked executivedecline,affectingplanningandthechoiceofappropriate navigationstrategies,isconfirmedbydifferentstudies[9,12].How- evercertainelderlysubjectsretainbetterabilitiestoadapttheir strategies[14].Thisobservationisnothoweverconfirmedinthe studybyCarellietal.[13].Finally,forcertainauthors,failureto efficiently integrate sensory-motor information could also be implicated in the deterioration of spatial navigation skills [25–28]. Overall,virtual realitystudies indicatethat navigation disturbancesinthecourseofnormalageingismulti-determined, and has origins that can be mnesic (allocentric processing associated with hippocampus function), executive (strategic processing associated with front-lobe function) or related to sensory-motorintegration(integrationoftheitineraryassociated withtheretrosplenialcortex).

3.1.2. SpatialnavigationandAlzheimer-typedementia

Alteredlarge-scalespatialcognitioninAlzheimer-typedemen- tia (ATD) has been described in differentstudies using virtual reality[7,17,29,30].ThestudybyBallassenetal.[31]suggeststhat memoryofthetemporaltypeinthecourseofanavigationtask, whichtendstoberelatedtoallocentricstrategies,is aselective behaviouralmarkerforATD.Topographicaldisorientationisoften consideredtobeaprodromalsignofthispathology,andhasled authorstousevirtualenvironmentsinordertoobjectifytheearly signs and evidence the underlying neurological correlates [32–34].Wenigeretal.[30]retrospectivelycomparedthespatial navigationperformancesofagroupofpatientswithmildcognitive impairment(MCI)thatevolvedtowardsATDandagroupofMCI patients whodidnot evolvetowards ATD.Theauthors didnot

evidence any differences between these two populations in virtualrealitynavigationperformances,whatevertheconditions (allocentricoregocentric).Theydidhoweverobservethatthesize of the hippocampus was significantly smaller in the MCI-ATD groupthanintheMCIgroupthatdidnotprogresstodementia.In thisstudy,thelackofspecificityofthedifferentnavigationtasks assigned with respect to the different navigation strategies (allocentricoregocentric)possiblypreventedthedifferentiation ofthetwogroups.

Cushmanet al. [7] compared navigation in real and virtual environmentin populationsof young, healthy elderly,MCIand ATD subjects, and they found strong correlations between the resultsobtainedinthetwoenvironmentsforthefourgroups,thus demonstratingthe robustecologicalvalidity of virtualenviron- ments.Thisstudy,andalsothatbyZakzanietal.[17],showedthat ATDpatients,andtoalesserdegreeMCIpatients,exhibiteddeficits in recalling landmarks, itineraries and the configuration of the environmentexplored.However,incomplementarymanner,ina

more recent study, Kessels et al. [35] observed that implicit memory of objects placed at decisional locations in the path- findingtaskwasnotimpairedinATDpatientspresentingatrophy ofthemiddletemporallobeincomparisonwithhealthysubjects.

Inaddition,Burgessetal.[36]earliernotedselectiveimpairmentof theallocentriccomponentofspatialmemoryinATD,andDrzezga etal.[37],amongATDorMCIpatients,observedgreaterdifficulty ininhibitingtheinterventionofcerebralprocessesnon-relevantto thenavigationtaskcomparedtohealthysubjects.Theauthorsthus concludedthatspatialnavigationdisturbancesarerobustclinical signsforATDandMCI.Betterunderstandingoftheirmanifesta- tions,andtheelaborationoftasksspecifictoallocentricnavigation in virtual environments, could improve prognosisfor progress fromearlycognitiveimpairmenttodementia.Theuseofvirtual environmentsisinthisrespectaparticularlyvaluableapproachfor the understanding and evaluation of these phenomena that generallyshowthemselvesonalargescaleandinordinaryliving settings.

6521 records idenfied throughdatabase searchingPubmed and Scopus with the key words «navigaon» and«virtual»

85 recordsidenfiedby adding the key words 1018 records idenfiedby adding the key words

«ageing», «demena», «brain injury», «help », « aid», « smulus» and« cue»

« stroke» and« schizophrenia»

54 records screened aer duplicates 967 records screened aer duplicates removed removedand selecon of the period and selecon of the period 2005-2015 2005-2015

54 full-text arcles assessed for eligibility 90 full-text arcles assessed for eligibility

63studiesincludedin qualitave synthesis 10 arcles

excluded

71 arcles excluded

Fig.1.Flowdiagram.

Table1

Articlesselectedwiththekeywords‘‘ageing’’and‘‘dementia’’.

Authors/year Mainaim n HC VE

setting*

Other investigation

Mainresult

Normalaging Normalagingeffectsonlargescalespatial cognition

Konishi(2013)[16] Toexaminetheeffectsofagingonthe hippocampusandcaudatenucleusduring navigation

Virtual maze

fRMI Agingprocessinvolvesashiftfrom usingthehippocampustowardthe caudatenucleusduringnavigation Taillade(2013)[24] Toassesswayfindingandspatiallearningin

relationtoexecutiveetmemorydecline

Virtual district

Memoryandexecutivedeclines hamperspatialnavigationinolder people

Adamo(2012)[26] Toexaminetheimpactofagingonpath integrationfromavalaiblesensorimotor informations

Virtual path

Agingimpactspathintegration.Result inVE*ispoorerthaninrealsetting Harris(2012)[25] Toexploreagedifferencesinpathintegration

bothwithandwithoutlandmarkinformation

Virtual landscape

InaVE*thatprovidedonlyopticflow information,olderparticipants exhibiteddeficitsinpathintegration Wiener(2012)[22] Toinvestigateage-relateddifferencesin

repeatingandretracingalearnedroute

Virtual maze

Theolderparticipantshadgreater problemsduringrouteretracingthan duringrouterepetition

Yamamotoetal.

(2012)[11]

Toinvestigatewhetherincreasingagehasequal consequencesforalltypesofspatiallearning

Virtual town

spatiallearningthroughexploratory navigationisparticularlyvulnerableto adverseeffectsofaging,whereas elderlyadultsmaybeabletomaintain theirmapreading

Head(2010)[10] Toexploreageeffectsonspatialnavigation abilitiesconsideringthemultiplecognitiveand neuralfactorsthatcontributetosuccessful navigation

Virtual maze

TDM Agedifferencesareobservedinboth wayfindingandroutelearning

Mahmood (2009)[27]

Toinvestigateage-relateddifferencesin participants’abilitytodeterminelineardistances, angularrotations,andangulardisplacement exclusivelyfromvisualinput

Virtual path

Thereareage-relateddeficitsinthe abilitytoperformvisualreturn-to- origintasks

Moffat(2007)[12] Toexamineagedifferencesinvirtual environmentnavigationandtoassesspossible relationshipsbetweennavigationandstructural integrityofhippocampalandextrahippocampal brainregions

Virtual maze

fMRI Individualdifferencesinregionalbrain volumesaswellasperformanceonthe testsofmemoryandexecutive functionscontributedtoagedifferences Moffat(2006)[23] toexamineagedifferencesinfunctionalbrain

activationduringvirtualenvironmentnavigation

Virtual maze

fMRI Thereisanagespecificneuralnetworks supportingspatialnavigation Lo¨vde´n(2005)[28] Toexaminetheimpactofimprovedsensorimotor

demandsonspatialnavigationamongelderly subjects

Virtual maze

Walkingsupportattenuatedage- relateddecrementsinnavigational learning

Normalaging Navigationalstrategiesadaptation

Harris(2014)[18] Totesttheabilityofyoungandoldparticipantsto switchfromlearnedroutestofindingnovel shortcuts

Virtual town

Olderparticipantsareimpairedinbothstrategyswitchingand allocentricprocessinginformations

Morganti(2014)[14] Totesttheabilitytoswitchfromallocentricto egocentricprocessing

Virtual maze

Virtualperformanceissignificativelyrelatedtotheoverall cognitivelevelandage

Bohbot(2012)[21] Toinvestigatenavigationalstrategiesacrossthelife span

Virtual maze

Theproportionofsubjectsusingspatialstrategydecreasesacross thelifespan,infavorofresponsestrategies

Harris(2012)[15] Toexploretheeffectsofageonstrategyswitching duringspatialnavigation

Virtual maze

Elderlysubjectsshowaspecific‘‘switch-to-place’’deficitthat couldaccountforapparentimpairmentsinbothnavigational strategyswitchingandallocentricprocessing

Rodgers(2012)[19] Toassessagedifferencesinstrategypreferencein humans

Virtual maze

Elderlypatientspreferentiallyuseegocentricstrategy Liu(2011)[20] Todeterminetheeffectofageandgenderupon

differentorientationstrategiesinvirtual environments

Virtual town

Age-relateddeclinesinnavigationarecommonacrossall orientationstrategiesandconfirmgender-specificeffectsin differentspatialdomains

Pathologicalaging Pathologicalagingonspatialcognition HC Tarnanas(2015)[34] Toinvestigatedeficitsinspatialnavigation,

prospectivememory,andexecutivefunction inMCI

25 Yes Virtual museum

EEG TheVEdiscriminatethepresenceofMCI

Bellassen(2012)[31] Tocomparetemporalmemoryamongvarious groupsfollowinganavigationtask

27 Yes Virtual maze

Temporalordermemorytestedinaspatial navigationtaskmayprovideaselective behavioralmarkerofAD*

Caffo` (2012)[33] Toinvestigatecategoricalspatialmemory deficitsusingavirtualnavigation-based reorientationtask

51 Yes VE* VEmayrepresentanevaluationsupplementfor spatialmemorydeficitsinprodromalstagesof Alzheimer’sdementia

Nedelska(2012)[29] Todeterminewhetherallocentricspatial navigationimpairmentwouldbeproportional torighthippocampalvolumelossirrespective ofgeneralbrainatrophy

42 Yes Virtual maze

Therighthippocampusplaysacriticalrolein allocentricnavigation,particularlywhen cognitiveimpairmentispresent

3.1.3. Spatialnavigationandacquiredbraininjuryinadults Evaluationandrehabilitationusingvirtualrealityaretwomajor clinicalorientationsthathavebeenexploredinresearchonspatial navigation disturbances among subjects in the aftermath of acquired brain injury. Better comprehension of the navigation strategiesaffectedintheirrelationshipwiththebrainlesionsis likewiseanimportantlineofresearchinthispopulation.Different authors have thus shown that among adult patients having sustainedheadtrauma,therearedifficultiesinacquiringallocen- tricspatialrepresentationsorrepresentationsoftheconfiguration ofanenvironment[38,39].Thevirtualenvironmentsusedbythese authorstodissociateallocentricandegocentricspatialrepresen- tationsarebasedontheMorriswatertaskparadigm[40,41].More recently, among amnesic patients presenting hippocampus lesions,thesameteamconfirmedthatthesubjectswereassisted in their navigation towards a target by proximal landmarks (egocentricstrategy)butnotifthesemarkersweredistantfrom thetarget(allocentricstrategy)[42].Ontheotherhand,inarecent study,Wenigeretal.[43]demonstratedthatpatientswithleftor rightfocalparietallesionsfailedtolearnanitineraryinavirtual maze(egocentricstrategiescomingintoplayinalearningtaskof the route type), while their performance in a virtual park (allocentric strategies deployed to learn an itinerary of the configurationoftheenvironmenttype)wasnormal.Theseauthors alsoshowedthat subjects’rightprecuneus volumewassignifi- cantly predictive of performance in the virtual maze, and suggested that the parietal cortex could underpin egocentric strategies in spatial learningskills in large-scale environments.

Carellietal.[13]fortheirpartshowedthata groupof8brain- injuredpatients(3withalefthemispherelesion,2witha right hemisphere lesion, 2 with bilateral lesions) had difficulty implementingefficient spatialnavigation strategiesin a virtual mazefromamapofthemazeonwhichtheyhadpreviouslyplotted the way to the exit, suggesting difficulties in the transfer of knowledgeoftheconfigurationoftheenvironmenttypetofollow anitineraryoftheroutetype.VanderHammetal.[44],inacase- basedapproach,demonstratedadualdissociationbycomparing the spatial navigation performances between two patients presentingrightparietal-occipitallesions.Inthesetwopatients, thespatialstrategieswerenotaffectedinthesamemanner,since

onehaddifficultyinthetemporalorganisationoftheitinerarybut notintherecognitionofthescenes,andthereversewastruefor theother.Theseauthorsalsoconfirmedthatthemaincomplaints from thepatients concerning their topographicaldisorientation werenotdetectedbythetraditionalneuropsychologicaltestsas efficientlyaswiththevirtualnavigationtests.Inanoriginalstudy, Slobounov et al. [45] compared a group of athletes who had sustained mild braininjury witha matched controlgroup ina virtualnavigationtaskcallingonspatialmemory.Theseauthors observed that while performances did not show significant differencesbetweenthetwogroups,thebrainactivationrecorded undermagneticresonanceimagery(MRI)inthebraininjurygroup showedanincreaseinbrainactivity(significantlygreaterthanin thecontrolgroup)intheparietalcortex,therightdorsal-lateral prefrontalcortexandtherighthippocampus.Accordingtothese authors,differenttypesoflesionassociatedwithneuronelosses and local changes in cellular metabolism couldexplain certain focalhyperactivationsandtherecruitmentofadditionalcerebral resources.

Theevaluation ofpatients followinga strokeand presenting unilateral spatial neglect affecting navigationwas recently the subjectofworkbyBuxbaumetal.[46,47].Intheirfirststudy[46], theseauthorsevidencedthesensitivityofthevirtualtesttodetect unilateral spatial neglectproblemsthat that werenotdetected using the traditional neglect battery, and they showed strong correlationswithanavigationtaskinrealenvironment.Thesecond study [47]confirmed themetric qualities of thevirtual reality navigationteston70patientspresentingrightcerebralvascular lesions.

Fewstudiestodatehaveexploredtheuseofvirtualrealityin rehabilitationprogrammesfor patientsexhibiting topographical disorientation. In a very recent study, the feasibility of a rehabilitation programme using virtual reality among patients followingastrokewastested[48].Therehabilitationapproachwas of the top-down type, and showed that the use of virtual environments enabled patients to learn alternative navigation strategiessuitedtotheirparticularcognitivedeficits.

Caglio et al. [49] used simulatednavigation tasks in virtual environmenttoimprovemnesicfunctionsinayounghead-trauma patient, thus using the multimodal cognitive nature of spatial Table1(Continued)

Pathologicalaging Pathologicalagingonspatialcognition HC Kessel(2011)[35] Toexaminememoryforobjectsrelevantfor

navigationinpatientswithAlzheimer’s dementia

21 Yes Virtual maze

SubjectswithAD*showimplicitmemoryfor directionallandmarks

Weniger(2011)[30] Tocomparespatialperformanceaccordingto corticalareasinvolvedinnavigationamong MCI*subjects

29 Yes Parkand maze

fMRI Variouscorticalareasarereducedamong patientsinrelationtospatialperformance Carelli(2011)[13] Toinvestigatehowthetransferfromsurvey

(map-like)torouterepresentationsisaffected afterabrainlesion

8 Yes Virtual

maze

Executivefunctionsandvisuo-spatialabilities deficitsappearedtobemorerelevantfor predictingpatients’resultsinperformingtheVR Mazetask

Pengas(2010)[32] Tocomparetopographicalmemorybetween variouspopulationofpatientswithAD*or MCI*

69 Yes Virtual town

Virtualtestdiscriminatesbetweenpatients withAD*,MCIandhealthysubjects Zakzanis(2009)[17] Toexamineage-andAlzheimer’sdisease-

relateddifferencesinroutelearningand memoryusingVR

2 Yes Virtual

town

Nodifferenceexistsinthelandmakrecognizing taskbetweenoldersubjectsandsubjectswith dementia

Cushman(2008)[7] Tocompareareal-worldnavigationtestwith aVEversionsimulatingthesamenavigational environment

26 Yes Virtual corridor

Closecorrelationsexistbetweenreal-worldand virtualnavigationaldeficitsthatincreased acrossgroups

Burgess(2006)[36] Toexplorethecaseofapatientthoughttobe intheveryearlystagesofAD

1 Yes Virtual

town

Thiscaseindicatesanewcategoryof topographicaldisorientationwithimpaired allocentricspatialmemory

Drzezga(2005)[37] Toevaluatebothcerebralactivationand deactivationinvariouspopulationincluding HC,MCI,andAD

21 Yes Virtual maze

PETsca DevelopmentofADappearstobecharacterized byaprogressiveimpairmentofcross-modal cerebraldeactivationfunctions

VE:virtualenvironment;AD:Alzheimerdementia;MCI:mildcognitiveimpairment;fMRI:functionalmagneticresonanceimaging.

navigation to improve impaired functions. In this single-case study,onthebasisoffunctionalMRIimagery,theauthorssuggest that this type of intensive training using virtual tasks might increasetheactivationofthehippocampusandpara-hippocampus regions.

3.1.4. Spatialnavigationinschizophrenia

Spatialnavigationdisordersamongschizophrenicpatientshave beenthefocusofspecificinterestintheliterature.Thesepatients present alteredspatial cognition,in particularregarding spatial strategiesoftheallocentrictype,whileegocentricstrategiesbased Table2

Articlesselectedwiththekeywords‘‘braininjury’’,‘‘stroke’’and‘‘schizophrenia’’.

Authors/year Mainaim n HC VE

setting*

Other investigation

Mainresult Braininjury&

stroke

Descriptionofprocessandbehavioral assessment

Buxbaum (2012)[48]

Toassessalargesampleofpatientswith theaimofdevelopingaclinicallyuseful versionoftheVirtualRealityLateralized AttentionTest(VRLAT)

70 Yes Virtual path

TheVRLATisasensitive,valid,andreliablemeasureof hemispatialneglect

Carelli(2011) [13]

Toinvestigatehowthetransferfrom survey(map-like)torouterepresentations isaffectedafterabrainlesion

8 Yes Virtual maze

Effectofbrainlesionsonthetransfermap/navigation

VanderHam (2010)[45]

Presentationoftwocaseswithnavigation problemsresultingfromparieto-occipital righthemispheredamage

2 X VE* Adoubledissociationbetweenspatialandtemporal deficitswasfound

Slobounov (2010)[46]

Toinvestigatethepossibilityofresidual functionaldeficitsinrecentlyconcussed butasymptomaticindividuals

15 Yes Virtual corridor

fMRI Concussedathletesshowedlargercorticalnetworks duringencodingthanHC

Goodrich-Hunsaker (2010)[43]

Toevaluatetheimportanceofthe hippocampusinspatiallearningand memory

5 Yes Virtual maze

fMRI Resultsconfirmtheroleofhippocampusinhuman navigationandprovideanewlinkbetweenits mnemonicandnavigationalroles

Weniger (2009)[44]

Tocomparesubjectswithbraininjuryand healthycontrolontwovirtualrealitytasks affordingtolearnavirtualpark(allocentric memory)andavirtualmaze(egocentric memory)

24 Yes Virtual maze andpark

fMRI Left-andright-sidedlesionedsubjectsdidnotdifferon taskperformancebutcomparedwithHC,subjectswith parietalcortexlesionswereimpairedlearningthe virtualmaze

Buxbaum (2008)[47]

Toreportdatafromparticipantswithright hemispherestrokeonanewVRnavigation test

9 X Virtual

path

TheVRtestdetectedlateralizedattentiondeficitsin participantswhoseperformancewaswithinthenormal rangeonotherneglecttests

Livingstone (2007)[40]

ToassessamongsubjectswithTBIwhat arethespatialstrategiesimpaired

11 Yes Virtual maze

NavigationofTBIsurvivorswasnotimpairedwhenthe proximalcueswerepresentbutwasimpairedwhen proximalcueswereabsent

Skelton (2006)[41]

TotestsubjectswithTBIinavirtualMorris watermaze

17 Yes Virtual maze

TBIsurvivorsnavigatedtoavisibleplatformbutcould notlearnthelocationoftheinvisibleplatform BrainInjury

&Stroke

Interventionprograms

Claessen(2015)[49] Toinvestigatethefeasibilityofavirtual navigationtrainingtoinstructchronicstroke patientstoadoptanalternativenavigation strategy

6 X VE* Thenavigationstrategythatpatient

usecanbeinfluencedafterashort trainingprocedure

Caglio(2012)[50] Todescribeandevaluatetheefficacyofa navigationaltrainingprograminamawithTBI

1 X VE* fMRI Intensivetraininginvirtual

navigationaltasksmayresultinan enhancementofmemoryfunction inbrain-damagedadults Schizophrenia Descriptionofprocessandbehavioralassessment

Ledoux(2013)[53] Toscanpeoplewithschizophreniaandhealthy patientswhilebeingtestedonnavigatingina virtualtown

28 Yes Virtual

town

fMRI Between-groupcomparisons revealedsignificantlylessactivation amongpatientsrelativetocontrols intheleftmiddlefrontalgyrus,and rightandlefthippocampi Wilkins(2013)[52] Toinvestigatethestrategiesusedbypatients

withschizophreniatosolvenavigationtasks

21 Yes Virtual

maze

Importanceofconsidering individualstrategieswhen investigatingmemoryand navigationperformanceamong variousgroupsofpatientswith schizophrenia

Siemerkus(2012)[54] ToobserveinfMRIpatientswithpositive symptomswhilefindingtheirwayinavirtual maze

16 Yes Virtual

maze and park

fMRI Lesactivationsdesre´gionspertinentes sontplusmarque´esdanslegroupe controˆle

Weniger(2008)[51] Tocomparesubjectswithrecent-onset schizophreniawithhealthysubjectsontwo virtualrealitytasksaffordingthenavigationand learningofavirtualpark(allocentricmemory) andavirtualmaze(egocentricmemory)

25 Yes Virtual

maze and park

Themoreglobalneuralnetwork supportingegocentricspatial learningseemstobelessaffected thanthedeclarativehippocampal memorysysteminearlystagesof schizophrenia

VE:virtualenvironment;TBI:traumaticbraininjury.

onlandmark-directionassociationsappeartobepreserved,with normaltosubnormalperformancesincomparison withhealthy subjects[50–52].Theauthorsofthesestudiesusedvirtualmazes with landmarks outside the laze derived from the real world (buildings,trees,mountainsetc.)toobjectifythedissociationson navigationstrategies. Ledoux et al. [53],under functional MRI, observeda decrease on hippocampus and middle frontalgyrus activityinthesepatientsinaway-findingtask.Theysuggestedthat thesepatientshaddifficultyassociatingspatialinformationwith contextsinwhichitappeared,affectingtheformationofcognitive maps. A study by Simierkus et al. [54] under functional MRI however showed than in some of these patients specifically presentingpositivesymptoms,therewasdifficultyinrecruiting the brain areas required for egocentric learning of spatial information. The authorsindicate that this couldbe related to altered consciousness that these patients with positive schizo- phreniasymptomshaveof themselvesand oftheir relationship withtheenvironment.

3.1.5. Overview

Recentstudiesonlarge-scalespatialcognitionamongsubjects presentingcognitivedisturbancesshowfirstofalltheinterestof using virtual reality to assess navigation strategies, hitherto inadequatelyexploredusingtheclassicpencil-and-papertestson small scale. This research confirms that large-scale navigation requirestheconcurrentprocessingbythebrainofdifferenttypesof informationinthecourseofnavigation,enablingasubjecttofindhis orherwaytowardsagivendestination[55].Itshowscomplemen- tarilybetweenspatialrepresentationsoftheallocentrictype[55,56]

andthoseof the egocentrictype[57,58], andthe wayinwhich navigationstrategiescanbeselectivelyaffecteddependingonthe brainnetworksthathavebeendamaged(forareviewoftheneuro- anatomical determinants of large-scale spatial navigation see Chrastil [59]). One perspective afforded by the use of virtual environmentssimulatingrealityandappearinginthemostrecent researchconcerns thescope forimprovingthe specificity ofthe virtual tasks assigned to patients for diagnostic and predictive purposes. Large-scale spatial cognition is in itself an excellent integrativemodelofglobalbrainfunctioningthatisdirectlylinked tocomplexnavigationbehavioursineverydaylife[60].

3.2. Theimpactofvisualorauditoryenvironmentalcuesonvirtual spatialnavigation

Oneofthecentralquestionswhenusingvirtualenvironments toassessspatialnavigationfunctionsisthatofthedifferencesin thesensory-motorcuesbetweenrealandvirtualsettings.Indeed, themotorcontrolinvolvedinthemanagementofnavigationusing akeyboardorajoystickisnotthesameasthatimplementedin actuallywalking.Theproprioceptiveand vestibularinformation involvedinthesetwotypesofnavigationisalsodifferent.Grant andMagee[61]thusshowedthatparticipantsguidedalongareal routefoundtheplacesmoreeasilyinthereal-environmenttest thanthosethathadlearnttherouteinvirtualsetting(i.e.actually walking in situ, or navigating with a joystick). This was not however evidenced in another study [62]. Williamset al. [63]

observedthatjudgementerrorsinperspectiveweregreaterand latencieslongerinrepositioningandrotationtasksforsubjectsin virtual environment compared to those in real environment.

Walletetal.[64]showedthatpoorer-qualityvisualinformation (textured versus non-textured environment) in a virtual city districtledtoadeficitinthetransferofinformationinthesame, realcitysetting,independentlyfromwhethertheinitiallearning situationfortheitinerarywasactiveorpassive,whichisinfavour of the preponderance of visual processing of information over motorinformation.

One of the lines of research concerns the delivery of complementarysensory-motorinformation thatis poorerwhen virtualenvironmentsareused.Ruddleetal.[65]thusshowedthat theuseofatreadmillexerciserassociatedwithataskconsistingin localisingitemsinasupermarketdecreasedthedistancescovered andimprovedtheprecisionoftheestimatesofdistancesbetween products.Larrueetal.[66]demonstratedthatthereconstructionof atrajectoryinarealenvironmentwasperformedbetterwhenthe trajectoryhadbeenlearntineitherrealorvirtualsituationusinga treadmill and rotationsof theupper body, in comparisonwith learninginvirtualenvironmentwithajoystickorwithatreadmill butwithoutupperbodyrotation.However,arecentstudyonatask involvingasearchforgoodsinashoppingmallalsoshowedthat amongelderlypeopledifficultiesgeneratedbythecombinationof thevirtualenvironmentandthetreadmillhadanegativeimpact onperformance,wherethetaskrequiredfrequentshiftsinfieldof visioninordertomovetowardstheobjects[67].Indeed,adding sensory-motor information or walking during navigation puts elderlyorbrain-injuredsubjectsinadouble-tasksituation(motor taskinmovingaroundandbalancingandcognitivetaskforthe navigation parameters) [28], which may have reduced perfor- mances.Itdoeshoweverseemthatactivenavigation(motorand cognitive)isbeneficialfornavigationamongyoungsubjects[68].

Despite these differences in terms of sensory-motor cues, a certainnumber of studiesarein favourofthesimilarity ofthe processesinvolvedinlearningandfindingatrajectoryinrealand virtualenvironments[7,69,70].Theredoesindeedappeartobea compensationforthelossofidiotheticinformationinnavigationin virtualenvironments[71].

Nevertheless, whilevirtual reality enables the simulation of certainaspectsoftherealworld,it alsomakesitpossibletogo beyondreality[72].Thisisthecaseinparticularwhenstimuliare addedtoavirtualenvironmentinthesoftware.Thesestimulican be defined as additional information provided by the virtual systemthatincreasestheamountofinformationpresentinthe environment.Howevertheirabsencefromtheenvironmentdoes not prevent completion of the task. These stimuli can be contextual,thatistosaylinkedtotheperformanceofthetask, ornon-contextual,thatistosaytheybearnorelationshiptothe tasktobeperformed.

Thesesensorystimulicanbeofdifferenttypes–auditory,visual, tactile or olfactory. They can be administered separately or in combinationinthecourseoftheperformanceofthetask.

Differentstudieshaveassessedtheimpactofvisualorauditory environmentalcuesdeliveredbythevirtualsystemonnavigation orspatiallearningtasks.

3.2.1. Visualcuesandvirtualspatialnavigation(Table3)

Thevisualcuesusedwerefirstofallcreatedinamannerthat wasanalogouswithaidsavailablefornavigationinanunfamiliar realenvironment.Forinstance,DarkenandSibert[73]proposeda virtualmap,andtheyreportedasignificantimpactofthiselement on navigation performance in a virtual environment among healthy subjects.Sjo¨linder et al.[74]showed thatthe useofa mapinavirtualenvironmentwasusefultoimprovecomprehen- sionofthespatiallayoutofinformation,butitslowedparticipants in their progression on account of the mobilisation of greater cognitivecapacities, and also becauseof therepeated switches between the egocentric perspective of the participant and the allocentric perspective of the map. Parush and Berman [75]

showed that at theoutset of their test navigation witha map provedmoredifficultthannavigationwithalistofinstructionsfor movement,butthatthisdifferencebecamenon-significantasthe durationofnavigationwasprolonged.Likewise,theexistenceof landmarksmadethenavigationmoredifficultattheoutset,butthe difficultylessenedasthenavigationwenton.Thedeteriorationin