The bodily self: Insights from clinical and experimental research

HAL Id: hal-02121686

https://hal-amu.archives-ouvertes.fr/hal-02121686

Submitted on 6 May 2019

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

The bodily self: Insights from clinical and experimental research

Sebastian Dieguez, Christophe Lopez

To cite this version:

Sebastian Dieguez, Christophe Lopez. The bodily self: Insights from clinical and experimental re- search. Annals of Physical and Rehabilitation Medicine, Elsevier Masson, 2017, 60 (3), pp.198-207.

�10.1016/j.rehab.2016.04.007�. �hal-02121686�

Review

The bodily self: Insights from clinical and experimental research

Sebastian Dieguez

, Christophe Lopez

*

aLaboratoryforCognitiveandNeurologicalSciences,Unite´ deNeurologie,De´partementdeMe´decine,Universite´ deFribourg,Fribourg,Switzerland

bLaboratoiredeNeurosciencesInte´grativesetAdaptatives,UMR7260,CNRS,Aix-MarseilleUniversite´,CentreSaint-Charles,FR3C–CaseB,3,place Victor-Hugo,13331Marseillecedex03,France

1. Introduction

Inthisreviewarticle,wepresentanoverviewofclinicaland experimental approaches to the study of body representations (also called body schema, corporeal awareness or bodily self), providinginsightintotheirneuralandcognitivebases.

Thefirstsectionprovidesanoverviewofunilateralandnon- lateralizedneurologicaldisordersofthebodilyself,oftenfollowing damage to the right cerebral hemisphere. Here, we address hemiasomatognosia,feelingsofdisappearanceandtransformation ofbodyparts,supernumeraryphantomlimbs,somatoparaphrenia andautoscopicphenomenainvolvingtheentirebody.Disordersof thebodily selfhave been difficultto classify systematically,as notedhistorically by pioneeringresearchers[1–5] andin more recentworks[6–8].Indeed,thebodilyselfisinitselfaconceptually complextopic,becauseitsexperienceis inherentlymultimodal, subjective,and global. The bodilyself arisesfrom thedynamic integrationofbodilyand environmental visual,tactile,proprio- ceptive,vestibular,auditory,olfactory,visceralandmotorinfor- mation[9].Unlikeotherworldlyobjects,thebodyisthesourceof itsownperception,asubjectandanobjectatthesametime.The

humanbraincomputesbodilyinformationviadifferentmapsand networks,notablyareasoftactile,proprioceptive,vestibularand interoceptiveprojectionintheprimarysomatosensory,aswellas unimodal and heteromodal cortices, providing an unified and global representation of the lived body, which allows for experiencing it as a unique self and agent. To these sensory components,recentresearchhasaddedspatialandsocialfactors involvedinself-otherbodilyinteractions,andego-andaltercentric perspectivesonbodilyandactionperceptions[10,11].

Verbalandhigher-levelcognitiveaspectsofbodilyknowledge aresometimessubsumedundertheconceptof‘‘bodyimage,’’said tobeaconsciousandabstractrepresentationofthebody,involving for instance thenaming of body parts and general knowledge about human bodies. This classical distinction with ‘‘body schema,’’ which involves situated, directly experienced,uncon- scious,andnon-verbalaspectsofthebodilyselfhasbeenwidely discussedelsewhere (e.g.,[12,13]), andweadopttheviewhere thatasomatognosiapersepertainstothebody schemadomain [14,15].Assuch,wethinkthatdisorderssuchasautotopoagnosia (impairednamingandpointingofbodypartsondemand[16,17]), Gerstmann’ssyndrome(amongothersymptoms,impairednaming andpointingoffingersondemand[18,19]),orideomotorapraxia (impairedproductionofgoal-directedgesturesondemand[20])– all involving damage to the left parietal lobe – are clinically, phenomenologicallyandconceptuallydifferentfromdisordersof KEYWORDS

Bodilyself Asomatognosia Illusions Delusions Virtualreality

ABSTRACT

Thisreviewarticlesummarizesneuropsychologicaldescriptionsofabnormalbodyrepresentationsin brain-damagedpatientsandrecentneuroscientificinvestigationsoftheirsensorimotorunderpinningsin healthyparticipants.Thefirstpartofthearticledescribesunilateraldisordersofthebodilyself,suchas asomatognosia,feelingsofamputation,supernumeraryphantomlimbsandsomatoparaphrenia,aswell asdescriptionsofnon-lateralizeddisordersofthebodilyself,includingAliceinWonderlandsyndrome andautoscopichallucinations.Becausethesensorimotormechanismsofthesedisordersareunclear,we focusonclinicaldescriptionsandinsistontheimportanceofreportingclinicalcasestobetterunderstand thefullrangeofbodilydisordersencounteredinneurologicaldiseases.Thesecondpartofthearticle presentstheadvantagesofmergingneuroscientificapproachesofthebodilyselfwithimmersivevirtual reality,roboticsandneuroprostheticstofostertheunderstandingofthemultisensory,motorandneural mechanismsofbodilyrepresentations.

* Correspondingauthor.

E-mailaddress:christophe.lopez@univ-amu.fr(C.Lopez).

thebodyschemaandwillnotbeaddressedhere.Thus,thisreview focusesondisordersofthebodyschema,whicharepredominant afterrightbraindamage[21].

In the second section, we summarize current experimental investigationsin healthysubjects regardingthebodilyself. We focusonexperimentalparadigmsthathavecreatedmultisensory conflicts(ofteninvolvingvisualandtactilesignals)tomodulatethe bodyschema(e.g.,evokingavirtualAliceinWonderlandsyndrome [22]orthesenseofhavingachildbody[23]),thesenseofowning thebody(e.g.,evokingtherubberhandillusion[24]),andthesense of embodied self-location (e.g., evoking an out-of-body–like illusion[25]). These approaches have beenfoundpromising to betterunderstandthesensorimotormechanismsthatunderliea largerangeofbodilydisordersobservedfollowingbraindamageor duringepilepticseizuresormigraineepisodes(Table1).

Finally, we offer some concluding remarks highlighting the importanceof mergingapproaches from neuropsychology with modernneuroimaging techniquesand protocols from cognitive neuroscience,immersivevirtualreality,roboticsand neuropros- theticsforestablishingamorecomprehensivemodelofthehuman bodilyselfanditsdisorders.

2. Disordersofthebodilyself

2.1. Unilateraldisordersofthebodilyself 2.1.1. Hemiasomatognosia

Theterm‘‘hemiasomatognosia’’wascoinedbyFrenchneurol- ogistJeanLhermitte[3]torefertoaneglect, lackofinterest,or unawarenessofonepartorentirehalfofone’sbody.Suchpatients generallyignoretheirleftarmand/orleg;theybehaveandspeakas ifthese didnotexist. One of Zingerle’s [26]patients, withleft hemiplegia,didnotpaytheslightestattentiontohisleftside,never lookedatit,neverspokeaboutit.Allorderstomovewereexecuted ontherightside,and,whenconfronteddirectly,thepatientdidnot see any absurdity in having only one body side. Zingerle and Lhermittesawinthisprofoundunawarenessforonebodysidethe source of other clinical phenomena such as anosognosia and unilateralneglect.IntheFrenchclinicalliterature,notionssuchas unawareness(me´connaissance)andlackofownership(de´sappar- tenance)werelater oftenusedtodescribehemiasomatognosia’s diverse manifestations. Frederiks [27] attempted to clarify the issuebyproposingadistinctionbetween‘‘conscious’’and‘‘non- conscious’’ hemiasomatognosia.Theformer referred topatients whoperceivedtheirbodyasincompleteoramputated, yetfully realizedtheillusorynatureofthesefeelings.Thelatterreferredto thesubjective‘‘disappearance’’ofonehalfofthebody(mostoften theleftone), withoutthepatientbeingabletonoticethis very disappearance. Today, ‘‘non-conscious hemiasomatognosia’’ is conceivedofaspersonalneglect, motorneglect, oranosognosia for hemiplegia. In each of these cases, there is some kind of

indifferenceforabodypart.Personalneglectreferstotheclassical pictureofapatientwho‘‘forgets’’tocomb,shaveormakeupthe leftsidebecauseofanattentional,perceptiveorrepresentational disorder.Motorneglectreferstopatientswithoutobjectivemotor disorders who underutilize their left members [1,28]. These patients behave as if they were hemiplegic although they are not.Conversely,patientswithanosognosiaforhemiplegiabehave asiftheyarenotparalysed:importantly,theynotonlydenythat they are paralysed, they also tend to ignore their left side in general. Frederiks [27]summarized thesesymptoms as ‘‘atten- tionaldisordersforthehemibody’’. Garc¸inetal.[29]wroteofa particularlystrikingcase:‘‘Theobservergetsthefeelingthatthe subjectbehavesasifheunderwentanamputationoftheleftsideof thebody’’.Thelossoflateralizedbody partrepresentationscan occuratdifferentlevelsofmultisensory,sensorimotororcognitive integration.Forinstance,patientsmayrecognizetheirownbody partswhenpresentedvisuallybutcompletelyforgetaboutthem whentheyareoutofsight.

2.1.2. Feelingsofamputation,hemi-depersonalization

Hereweaddressdisorderslabelled‘‘conscioushemiasomato- gnosia’’byFrederiks[27].Suchpatientshavevividfeelingsthata partoftheirbodyhasdisappearedorfeelitstronglydiminishedor blurred.Conceptually,thesedisordersseemtobethereverseof phantom limbs after amputation (most amputees retain a sensationofcompletenessdespitehavingphysicallylostabody part).Inbothcases,patientsfullyappreciatetheillusorynatureof theirsensations.

Arelateddisorderisthefeelingthatabodypartisnolonger attachedtotherestofthebody,asifitwere‘‘floating’’atsome distance(sensation of disconnectionor splitting). Symptoms of absenceorseparationofbodypartsareusuallyofshortduration and appear mostly as part of seizures or migraine episodes [2,4,27,30].Othercasesoccurduetocorticalorsubcorticalstrokes [31]. Directelectrical stimulation at the right temporo-parietal junctioncanalsoinducethistypeofillusioninthevisualmodality [32]. These symptoms are not necessarily accompanied by hemiplegia, unilateral neglect or anosognosia. Sometimes, a sensationofstrangenessforan‘‘absent’’or‘‘disconnected’’body part,thenfeltas‘‘alien’’,‘‘numb’’or‘‘empty’’,isreported:theterm hemi-depersonalization,ordepersonalizationforabodypart,has beensuggested[1,3].Patientswithsuchsymptomsoftenfeelthe needtocontrolthesebodypartsbysightortouch,withoutsuch strategiesbeingalwaysabletorestorenormalbodilyfeelings[1].

Othersymptomscaninvolvedistortionsintheperceivedsizeof selectedbodypartsorhalfofthebody.Thesesubjectivealterations ofbodilysizearevividlyexperiencedbutareusuallyrecognizedas illusory.Torefertothesesymptoms,Frederiks[33]usedtheterms microsomatognosia and macrosomatognosia. Both terms are reminiscentoftheconceptsofhyposche´matie(i.e.,ashrinkingof thebodyrepresentation)andhypersche´matie(i.e.,anenlargement Table1

Mainmodelsforneuropsychological,neuroscientificandneuroimaginginvestigationsofthebodyschema/imageandthesenseofbodyownershipandself-location/

embodiment.

Bodilyexperience Clinical(neuropsychological)model Experimental(neuroscientific)model Functionalneuroimagingmodel Bodyschema/bodyimage Macro/microsomatognosia[33]

Supernumeraryphantoms[44]

Anorexia,bulimianervosa

Bodyidentityintegritydisorder[108]

Rubberhandillusion[24]

Immersivevirtualreality[22,23,104]

Shrinkingillusion[109]

Bodyownership Somatoparaphrenia[50]

Bodyidentityintegritydisorder[108]

Rubberhandillusion[24]

Virtualarmillusion[81]

Numbnessillusion[84]

Rubberhandillusion[85,87]

Virtualarmillusion[110]

Self-location/embodiment Autoscopichallucinations:

out-of-bodyexperience[67]

Heautoscopy[74]

Full-bodyillusion[92]

Out-of-bodyillusion[25]

Bodyswapillusion[98]

Immersivevirtualreality[103]

Full-bodyillusion[76]

Bodyswapillusion[86]

ofthebody representation)originallycoinedbyBonnier[34]in 1905todescribedistortionsofthebody schema(asche´matie)in patients withsensory and central disorders [35].In neurology, microsomatognosia(hyposche´matie)occurs whena bodypartis experiencedassmallerthanusual(e.g.,somehemiplegicpatients perceive their hand as a child’s hand). In macrosomatognosia (hypersche´matie), a body partis perceivedas largerthan usual, oftentimes muchheavier too.The illusion can expand to such degreesthatthebodypartisfeltasfillingtheroomorhittingthe roof.These disorders occur mostoftenduring migraine[36]or epileptic seizures [2] but also after brain damage leading to sensoryormotorimpairment[33,37].

2.1.3. Supernumeraryphantomlimbs

‘‘Supernumerary phantom limbs’’ refer to the subjective experienceofhavinganadditionalbodypart,usuallyalimb,felt asanentitysharingpropertieswithaphysicalbodycounterpart andoccupyingadifferentplaceinspace.Somepatientsexperience thepresenceofa‘‘thirdlimb’’andclearlyidentifythisexperience as an illusion, whereas others report multiple arm or leg reduplicationsaspartofadelusionandseeminglybelieveintheir physicalexistence[38–40].Inthelattercase,theterm‘‘delusional reduplicationofpartsofthebody’’hasbeenapplied[39].Super- numerary phantom limbs have been scarcely reported in the literature [40–42], yet they display striking diversity in their manifestations. In most cases, the‘‘additional limb’’ is a static somestheticperceptlocatedseparately,butonthesameside,ofa plegiclimb.Itissometimesfeltassmallerandindifferent,oreven awkward,positions,thanthephysicalcounterpartlimb.

Movementsofthesupernumeraryphantomareratherrareand mostoftenautomaticorinvoluntary(seeFig.1and[42]).Awoman with right fronto-mesial damage involving the supplementary motorarea andthecingulargyrusandno hemiplegiaormotor disordersreportedthefollowingpeculiarity:whenevershemoved her physical limb, a phantom seemed to occupy, after a few seconds,theplaceleftbythereallimb[43].Thus,theexistenceof thissupernumeraryphantomspecificallyreliedonmovementsof therealcounterpartlimb.Stillanotherpatientreportednotonly beingable to ‘‘trigger’’ and moveher supernumerary phantom voluntarilybutalsoclaimedtoseeit(assomewhatwhitishand transparent)andsometimesevenfeltittouchherface.Inaddition, this complex phantom, as reported by the patient, could not coexistinthesameplacewithotherobjectsorbodyparts(inwhich casesit‘‘disappeared’’instantly)[44].Suchaclinicalpresentation is rare but neatly illustrates the diversity of supernumerary phantomlimbsintermsoftheinvolvedmodalities,motoraspects andsensitivitytofeedback.Thisdiversityisalsoreflectedinthe involved brain areas: most often, the right hemisphere is implicated,butthebasalganglia,parietallobe,thalamus,medial prefrontal cortex or supplementary motor area can also be involved[15].

2.1.4. Somatoparaphrenia

Gerstmann[45]coinedtheterm‘‘somatoparaphrenia’’torefer toclearlydelusionaldisordersofthebodilyself.Accordingtothis author,somerelatedcasesdeservedtobedistinguishedaccording to the mental frame of the patients. Most patients with hemiasomatognosia and anosognosia for hemiplegia remain indifferent or make rather limited claims regarding their impairment,whereasothers seemto‘‘standapart’’byvirtueof their sheer bizarreness and exuberance. Here, is Gerstmann’s original definition of somatopraphrenia: ‘‘[A] specific psychic elaboration(markedbyformationofillusions,confabulationsand delusions)withrespecttotheaffected membersorside of the body, believed or experienced as absent’’ (p. 912). Somewhat confusingly,theterm‘‘verbalasomatognosia’’hasalsobeenused

torefertocasesinwhichpatientsmisidentifytheirownlefthand, presentedvisuallybytheexperimenterforsomeone else’shand [46].Somatoparaphreniais mostlyusedtorefer tofalse beliefs regarding one’s body part or half body, the most frequently reportedbeingthebeliefthattheybelongtosomeoneelse.Tothat extent, current and classical definitions of delusional disorders directlyapplytosomatoparaphrenia[47–49].However, aswith supernumerary phantom limbs, the clinical presentation of somatoparaphrenia differs considerably among patients, which suggestsdifferenttypesofsomatoparaphrenia.

Themostcommontypeofsomatoparaphreniaisthemisattri- butionofone’sbodypart(usuallythelefthand,armorleg)tothe doctor,anurse,a neighbouringpatient,orsomeoneunspecified [26,45,50].Somepatientsalsomisattributetheirparalysedlimbto someone absent or even long dead [51]. A sophisticated and delusionalaccountoftheeventcanbeprovided,oftentimeswitha Fig.1.Supernumeraryphantoms.A37-year-oldwomanreportedthesensationof havingathirdleftarmandlessfrequentlyreportedthesensationofhavingathird leftleg.Thepatienthadnovoluntarycontrolofthephantomlimbmovement,but thephantomfollowedthemotionofthereallefthand.Thepatientwasfullyaware oftheillusorynatureofherexperience,althoughthephantomlimbsfeltsovivid thatshehaddifficultiesdistinguishingthemfromthephysicallimbs.MRIrevealed aninfarctionoftherightdorsomedialfrontallobeaswellasprenatallesionofthe corpuscallosum.Notethatthepatient’sdrawingismirror-reversedsothatshe representedherleftbodyontherightsideoftheimage.

ReproducedfromHarietal.[42]Threehands:fragmentationofhumanbodily awareness.NeurosciLett1998;240(3):131–4,withpermissionfromElsevier.

persecutoryflavour. Forinstance,a patient reportedby He´caen etal.[52],wenttothelengthofcomplainingbylettertoanurseshe accusedofhavingtakenawayherarmandthreateninglyaskedher tobring it back. The intensityof the delusion can vary: some patients acknowledge that their claims are bizarre, and others steadfastlyholdontotheirbeliefs.Bycloselyinspectingthearm from the shoulder down, patients can come to realize that it belongstothem,butnotalways[51].Onepatientthusdeclared:

‘‘my eyes and my feelings don’t agree, and I must believe my feelings.Iknowtheylooklikemine[theaffectedlimbs],butIcan feeltheyarenot,andIcan’tbelievemyeyes’’[5].

OnepatientreportedbyHe´caenanddeAjuriaguerraprovidesa goodillustrationofhowdifficultitissometimestocategorizethese disorders[2].Thispatientspontaneouslydeniedthathislefthand belongedtohimbutdidnotattributeittosomeoneelse(when shownhishand,hesimplysaid‘‘it’snotmine’’).However,upon seeing the doctor’s hand, he claimed he recognized his hand (‘‘thereis mine’’). The patientheld tothis idea even when the doctorproducedmovementswithhishandandtookafewsteps back.Hebegantodoubthisassertiononlywhenthedoctorwastoo faraway(‘‘I’mstartingtobelievethatitisnotmine’’).Thus,there are2 typesof ownershipmisattributionin somatoparaphrenia:

somepatientsidentifytheirlimbasfromanotherperson(self-as- othererror),whileothersidentifyanotherperson’slimbastheir own (other-as-self error, this type being much more rarely reported).

Brain damageleadingtosomatoparaphreniainvolvesa wide fronto-temporo-parietal network in the right hemisphere and morespecificallytheinsula,theprefrontalandorbitarycortex,the underlying white matter and subcortical structures (thalamus, basalgangliaandamygdala)[50,53–55]aswellasthehippocam- pus[15].Theimplicationofmultisensoryregionsintheabnormal sense of owning a body suggests that coherent multisensory integrationisrequiredforelaboratingbodilyself-consciousness.

Somatoparaphrenia, like many other alterations of bodily self- consciousness[9],mayinpartberelatedtomisintegrationofor conflicts between proprioceptive, interoceptive and vestibular signalsaboutone’sbodypositionandmotionwithvisualsignals fromthebody.Interestingly,sensorysignalsmodulationbycaloric vestibular stimulation [56–58] and visual inspection of the disowned hand in a mirror [59] can significantly decrease somatoparaphrenicdelusions.Therefore,bodyownershipisunder thecontrolofperipheralsensorysignals.

2.2. Non-lateralizedandbilateraldisordersofthebodilyself The previous section provided an overview of lateralized (mostlyon theleft side)disorders of the bodilyself. Here, we brieflyaddressnon-specificallylateralizeddisordersanddisorders extendingtotheentirebody.Forinstance,macrosomatognosiacan sometimesinvolvetheheadortheentirebody,inducingfeelingsof enormityorof‘‘fillingtheroom’’. Thiskindof symptomoccurs mostlyduringmigraineaurasandarecalledAliceinWonderland syndrome[60].Otherpatientscanfeeltheirentirebodyasabsent or unreal, typically during depersonalization [61] or, in more extreme cases, describe their body as dead or non-existent, a conditioncalledCotardsyndrome(whichLhermittetermed‘‘total asomatognosia’’[3]). Moregenerally,suchsymptomsand those that follow can be conceptualized as ‘‘complete’’ forms of the unilateraldisordersreviewedintheprevioussections[62–64].

Themoststrikingdisordersinvolvingthemisrepresentationof theentirebodyarethosesometimeslabelled‘‘illusorydoubles’’, reminiscent of the ‘‘delusional reduplication of body parts’’

describedabove,and ofthe‘‘double’’and doppelga¨ngermotives inromanticandgothicliterature[65].Suchstrikingalterationsof the global bodily self occur under varied circumstances, not

necessarilypathological,andaremostoftentransitory.Theyare currently described and studied under the term ‘‘autoscopic phenomena’’ [66–68]. This area includes multimodal illusions producingmoreorlesscompletedoublesofthebody.Subjectscan thusperceiveavisualprojectionoftheirownbodyinfrontofthem (autoscopichallucination),inwhichcasethe‘‘double’’appearsasa mirror reflection in external space,while the real selfremains firmlytighttoitsphysicalbodyandlocation.Therefore,autoscopy assuchismainlyavisualphenomenon,althoughitcansometimes involvesomemotorresonancewhenthedoublemovesaccording tothesubject’sownmovements.

Inothercases,thesubjectreports feelingthelocationofthe

‘‘self’’atthesametimeintheprojecteddoubleandinthephysical body (heautoscopy)oralternatingbetween them.The ‘‘double’’

can then acquire some limited sensorimotor and psychological autonomyandisthusclosetotheliterarydoppelga¨ngerdescribed by Hoffmann, Dostoyevsky, Poe and Hogg. Heautoscopy has a visual component but also involves unstable vestibular and sensorimotoraspects[67,69]andastrongemotionalaffinity(or repulsion)towardsthe‘‘double’’.Whenthedoubleisnotactually seen but rather feltin close spatial proximity,thecondition is called ‘‘feeling of a presence.’’ The subject has a more or less fleeting sensationof someone’spresencenearby, withoutbeing abletoclearlyperceiveoridentifyitbutalsowithoutexperiencing itasadoubleofoneself[70].

Whentheperspectiveofthesubjectisentirelyrelocalizedinthe projecteddoubleandthusthesubjecthasthevividsensationof being‘‘out-of-the-body’’andcan‘‘see’’thephysicalbodyfroma distantandelevatedperspective,theconditioniscalled‘‘out-of- body experience’’ [32,67,68]. During such episodes, widely popularized by their inclusion in the hallucinatory phenomena called ‘‘near-death experience’’ [71], there is a very strong vestibularinvolvement(feelingsoflightnessandfloating,reversal of the visuo-spatial perspective), and the purely autoscopic component(‘‘seeingoneself’’)canbelessprominentoraltogether absent[69,72,73].

In neurologicalcases, autoscopichallucination, beingmostly visualdisturbances,involvesdamagetotheoccipitalcortex,often unilaterally [74]. Otherdisorders, being morecomplex, involve varied disturbances of multimodal and vestibular integration, notablyattheleftposteriorinsularegardingheautoscopy[74],the insulaandafronto-temporo-parietalnetworkforthefeelingofa presence[75],andtherighttemporo-parietaljunctionforout-of- bodyexperiences[76].

In the next section, we describe experimental procedures developed to study mechanisms underlying the bodily self in healthy subjects, to better understand the disorders we have described.

3. Neuroscientificinvestigationsinhealthyparticipants

Inthissection,wesummarizerecentworkfromneuroscience andexperimentalpsychologythathasendeavouredtoinvestigate howthebrainrepresentssomeaspectsofthebodyandtheself.We focusonresearchofhealthyparticipantsrelatedtothesenseof owningthebody(alteredinsomatoparaphrenicpatients)andself- location/embodiment (altered during out-of-body experiences) (see Table 1), 2 experiences deemed crucial for establishing a minimalsenseofselfhood[77].

3.1. Investigatingbodypartrepresentations

Healthyparticipantsresearchhasextensivelyusedthe‘‘rubber hand illusion’’ (RHI) [24,78] to investigate the multisensory foundationsofbodyownershipaswellasitsneuralunderpinnings

(Fig.2A).The RHIisevoked whena rubberhandplaced inthe participant’s field of view is touched in synchrony with the participant’shand(hiddentotheparticipant).Aftera minuteof synchronous stimulation of the fake and real hands, some participantsreportthattherubberhandfeelsasifitweretheir own hand. Subjective reports measured by questionnaires (i.e., visual analog scales) indicate that illusory ownership for the rubberhandissignificantlylargerforsynchronousthanasynchro- nousvisuo-tactilestroking.Inaddition,theRHIischaracterizedby amislocalizationoftheparticipant’shandinspace.Whenaskedto locate theirhandinthehorizontalplane(e.g.,bypointingwith their right hand toward the tip of their left index finger), participantstendtolocatetheirhandclosertotherubberhand thanitactuallyis.Thiserrorinself-handlocalizationtowardsthe rubber hand has been termed ‘‘proprioceptive drift’’ and is classically interpreted as a consequence of a ‘‘visual capture’’

[79,80].

SeveralvariantsoftheRHIdevelopedincludepresentationofa virtualhandinimmersivevirtualenvironments(i.e.,the‘‘virtual arm illusion’’[81]; see Fig.2B),or thepresentation ofmultiple hands[82].Ehrssonetal.[83]developedavariantoftheRHIduring whichtheexperimenterusestheparticipant’srightindexfingerto touch a left rubber hand while the experimenter touches the participant’slefthandwiththeexperimenter’sownfinger(Fig.2C).

Thisprocedurecreatesthesensationthatparticipantsaretouching theirownhand(insteadofarubberhand)withtheirrightindex finger.Anotherparadigm,the‘‘numbnessillusion’’[84],replaces the rubber hand by a confederate’s hand. In this illusion, the participant and the confederate have the palm of their hands pressedagainsteachother.Participantsstroketheirownleftindex fingerusingtheirrightthumbandatthesametimestrokewith their rightindexfingertheconfederate’sindex finger(Fig.2D).

Participantsmayexperiencetheirleftindexasnumb(hencethe term‘‘numbnessillusion’’)orbigger,asifitwereencompassingthe confederate’sindexfinger.Thesesensationsoccuronlywhenthe participant’s and confederate’s indexfingers are synchronously stroked.

Several neuroimagingstudiesinvestigated theneuralunder- pinningsof body partownershipby using theRHI.Brain areas significantlymoreactivatedduringtheRHI(i.e.,duringsynchro- nousvisuo-tactilestimulation)weremostlylocatedintheinsula, cingulatecortex,premotorcortexandextrastriatecortex(extras- triate body area, EBA) [83,85,86].Moreover, an increase in the BOLD signal in the insula was positively correlated with the magnitudeoftheproprioceptivedrift[87].Finally,onestudy[84]

recordedsomatosensoryevokedpotentialsduringthenumbness illusionwhile participantsreceived electricalstimulation ofthe median nerve. The numbness illusion was associated with increased amplitude of the N20 component of somatosensory evokedpotentials. Thisresultsuggestssomatosensoryenhance- ment in the primary somatosensory cortex, in keeping with increased evoked potentials over the somatosensory cortex reportedafterrealanaesthesiaoftheparticipant’shand[88].

Fig.2.Paradigmstoinvestigateownershipforbodyparts.A.Inthevisualvariantof the‘‘rubberhandillusion,’’theexperimenter uses2identicalpaintbrushesto synchronouslyorasynchronouslystroketheparticipant’shand(notvisibletothe

participant)andarealisticrubberhand(visibletotheparticipant).B.Inthe‘‘virtual armillusion,’’spatialandtemporalsynchronyiscreatedbetweenthetouchapplied totheparticipant’shand(toppartofthefigure)andthetouchobservedonthe virtualarm(bottompartofthefigure:sceneshowninthehead-mounteddisplay).

ReproducedfromEvansandBlanke[110].Sharedelectrophysiologymechanismsof bodyownershipandmotorimagery.Neuroimage2013;64:216–28.http://dx.doi.

org/10.1016/j.neuroimage.2012.09.027,withpermissionfromElsevier.C.Inthe non-visualvariantoftherubberhandillusion,spatialandtemporalsynchronyis createdbetweenthetouchappliedtotherubberhandbytheparticipant’sright indexfinger(thatispassivelymovedbytheexperimenter)andthetouchappliedby theexperimentertotheparticipant’slefthand.D.Inthe‘‘numbnessillusion,’’the participant(hereontheleftpartofthepicture)strokeswithhisrightthumbthe dorsalpartofhisleftindexfingerandatthesametimestrokeswithhisrightindex fingerthedorsalpartoftheconfederate’srightindexfinger.

We conclude this section by emphasizing recent findings showingthatillusoryownershipoverafakehandchangesseveral aspectsoftheownbodyphysiologicalstates[89,90].Forexample, Moseley et al. [90] reported that illusory ownership of a right rubberhandsignificantlyreducedthetemperaturerecordedover theparticipant’srighthand.Thistemperaturereductionwasnot found for the participant’s left hand and foot (not stimulated duringtheexperiment),sodisruptingthesenseofownershipfora givenbodypartmodifiestemperatureregulationatthelevelofthis samebodypart.Interestingly,bodyownershipabnormalitiesand changeintemperatureregulationareassociatedinseveralclinical conditionssuchasschizophrenia,neuropathicpain,anorexiaand bulimianervosa[90].Thus,therelationbetweenbodyownership/

disownershipandhomeostaticregulationmayshedlightonthe multisensorymechanismsofseveralconditionscharacterizedby abnormalbodyrepresentations.

3.2. Investigatingwhole-bodyrepresentations

InvestigatingtheRHIandrelatedillusionswouldnotbeenough tounderstandthesensorimotorfoundationsofhumanbodilyself- consciousnessbecauseitaddressesonlybodypartrepresentations.

Yet, in additionto representationsof multiple body parts, and perhapsmoreimportantly,thebodilyself,alsoentailsglobaland coherent whole-bodyrepresentation [9,77]. Accordingly, recent neuroscientific investigations intothe bodily self have endeav- oured toadapt theRHI to the entire body by using the same principles,thatis,visuo-tactileorvisuo-motorsynchronybetween thephysicalbodyandtheseen(fake)body[91].

In the ‘‘full-body illusion’’ [92], participants wore a head- mounted display through which real-time videos of their own bodycanbeseenfromtheback(thevideorecordingsystembeing placedacoupleofmetersbehindtheirownbody),orvideosofa plasticmannequin ora virtualcharacter[92–96]. Thus, partici- pants standing upright observed their own body (or the mannequin’s body) as if it was seen from an external, third- person perspective (i.e., as if participants were located some metersbehindthisbody)(Fig.3A).Whenatactilestimulationwas appliedontheparticipant’sback,insynchronywithtouchapplied to the back of the body depicted through the head-mounted display,participantsself-identifiedwiththisbody.Self-identifica- tion(measuredbyusingquestionnaires)wassignificantlystronger aftersynchronousthanasynchronousvisuo-tactilestroking.Asfor the RHI, self-identification with the avatar decreased skin temperatureattheleveloftheparticipant’sbody[97]andreduced theparticipant’spainperception[95].In addition,thefull-body illusionwascharacterizedbyerrorsinself-localizationinspace.

Fig. 4 illustrates theprocedures for measuringa ‘‘whole-body’’

proprioceptivedriftandshowsthatparticipantslocalizedthem- selvesclosertotheobservedbodyaftertheyreceivedsynchronous visuo-tactilestroking.Acrossseveralvariantsoftheillusion,the meanproprioceptivedriftwas177cmforvisuo-tactilestroking and36cmforasynchronousstroking,whichdifferedsignificantly

Fig.3.Experimentalparadigmstoinvestigatewhole-bodyownershipandself- location.A.Paradigmusedtoevokethe‘‘full-bodyillusion’’:identificationwiththe bodyseeninthehead-mounteddisplayandapparentforwardshiftinself-location.

B. Paradigmusedtoevokethe‘‘out-of-bodyillusion’’:identificationwiththebody seeninthehead-mounteddisplayandapparentbackwardshiftinself-location.

Reproduced from Ehrsson [25]. The experimental induction of out-of-body experiences. Science 2007;317:1048, with permission from The American Association fortheAdvancementof Science. C.Paradigm usedto evokeself- identificationwithaplasticmannequin’sbodywithoutchangeinself-location.

ReproducedfromPetkovaandEhrsson[98].IfIwereyou:perceptualillusionof bodyswapping.PLoSOne2008;3:e3832.http://dx.doi.org/10.1371/journal.pone.

0003832.D.Immersivevirtualenvironmentcoupledwithmotiontrackingsystems toinduceself-identificationwithachildavatarseeninahead-mounteddisplay.

ReproducedfromBanakouetal.[23]Illusoryownershipofavirtualchildbody causesoverestimationofobjectsizesandimplicitattitudechanges.ProcNatlAcad Sciences U S A 2013;110(31):12846–51. http://dx.doi.org/10.1073/pnas.

1306779110.

(P<0.005)[11].Becausethemagnitudeofthisdriftisrathersmall,as fortheproprioceptivedriftobservedduringtheRHI,itcanbeseenasa compromise betweenthe physical self-location and a full-blown disembodied self-location at the position of the avatar the participantsself-identifiedwith.

Avariantofthefull-bodyillusionisreferredtoasthe‘‘out-of- bodyillusion’’[25](Fig.3B).Incontrastwiththefull-bodyillusion, participantswerestrokedontheirchestwhiletheyobservedina head-mounteddisplayareal-timevideooftheir ownbodyseen fromtheback.Self-reportsincludedstatementssuchas‘‘Wow!I

feltasthoughIwasoutsidemybodyandlookingatmyselffrom the back’’ [25]. Self-identification with the seen body was measured byrecordingskinconductanceresponseswhileparti- cipantsobservedathreatdirectedtowardstheseenbody.Theskin conductance response waslargerafter synchronous thanasyn- chronousvisuo-tactilestroking.Inanothervariantoftheillusion, participantsbenttheirheadforwardasiftheywerelookingattheir own stomach and feet [86,98,99] (Fig. 3C). They wore a head- mounted display in which videosof a plasticmannequin were shownfromanelevatedviewpoint.Incontrastwiththefull-body illusion and theout-of-body illusion, the mannequin was seen from a first-person perspective (but seen from a disembodied, third-person viewpoint in the former illusions) and with a descendingviewpoint.Participantsandthemannequin received eithersynchronousorasynchronousvisuo-tactilestroking,andthe degreeofidentificationwiththemannequinwasmeasuredboth subjectively (questionnaires)and objectively(skin conductance response). Synchronous visuo-tactile stroking evoked stronger self-identification with themannequin and stronger emotional responseswhenparticipantsobservedathreatdirectedtowards themannequin’sstomach.

OnlyafewfunctionalMRIstudieshaveanalysedbrainactivity duringthebodyillusionsdescribedabove.The‘‘full-bodyillusion’’

modulates theBOLDsignalatthetemporo-parietal junction, in closevicinityofareasthataremostfrequentlydamagedinpatients reporting out-of-body experiences [76]. In contrast, observing fromafirst-personperspectiveamannequin’sbodybeingtouched insynchronywithone’sbodyactivatedmostlythepremotorcortex butnotthetemporo-parietaljunction[100].

Finally,itisnotablethatimmersivevirtualenvironmentsand motion capturetechnologies havefosteredinvestigationsofthe multisensory foundations ofbody representations becausethey allowpresentingrealistichumanavatarsandmanipulatingtheir shape,size,ageandskincolour[22,101–104].Ageneraloutcomeof thesevirtualrealitystudieswasthatspatio-temporalsynchrony betweentheparticipant’smotionandthatofanavatarembedded inavirtualenvironmentledparticipantstostronglyself-identify with the avatar. Recent studies indicate that sensorimotor synchrony between the physical and virtual body not only modulates the participants’ own body representation but also altersaspectsoftheiraffectiveandmoraljudgments.Forexample, synchronous motion between the participant and an avatar depicting a 4-year-old child modified the participant’s body schema,asmeasuredbytheirabilitytomanipulatevirtualobjects [23](Fig.3D).Inthisstudy,participantsoverestimatedthesizeof objectsaftertheyidentifiedwitha4-year-oldchildavatar.Inthe samelineofresearch,self-identificationwithanoverweightavatar changed the perceived sizeof theparticipant’s body and their potential actionsin theenvironment [104]. Several therapeutic interventions in neurology and psychiatry have recently been derivedfromimmersivevirtualenvironmentsgiventheapparent ease toinduceself-identification withavatars and tomodulate bodily,socialandaffectiverepresentationsbyusingappropriate technologies[101,105].

4. Conclusionsandclinicaloutcomes

Clinicalcasesandrecentneuroscientificstudiessummarizedin thisreviewindicatethatrepresentationsofthebodyandselfare varied,multimodalandplastic.Regardingclinicalcases,onefact mustbehighlighted:ourcurrentunderstandingofthebodilyself historicallymostlydependsoncarefulexaminationofawidearray ofclinicalconditions.Asaprivateandsubjectiveexperience,the bodilyselfanditsdisturbancescanonlybeapproachedthrough systematicandempathicquestioningofpatientswillingtoshare Fig.4.Illusoryself-locationevokedduringthefull-bodyillusion.Afteraperiodof

synchronous or asynchronous visuo-tactile stimulation, participants were displacedbackwardsandaskedtowalkbacktowheretheythoughttheywere locatedduringtheexperiment.Thepositionwhereparticipantsstoppedwastaken asameasureofself-locationinthehorizontalspaceduringthefull-bodyillusion.

Thisprocedurewasconductedtomeasurea‘‘whole-body’’proprioceptivedrift, similartothatmeasuredwiththerubberhandillusion.A.Participantswalked significantly further towards the seen avatar after synchronous visuo-tactile stroking (blue symbols) than after asynchronous visuo-tactile stroking (red symbols). The zero position indicates the location of the participant’s body duringthevisuo-tactilestrokingandpositiveandnegativevaluesindicatedriftsof self-locationtowardsandawayfromtheseenavatar,respectively.Dataareplotted accordingtoref.1aand1b[92],ref.2[94],ref.3[96],ref.4[95],ref.5[111],andref.

6[112].ReproducedfromLopezetal.[11]Inthepresenceofothers:self-location, balancecontrolandvestibularprocessing.NeurophysiolClin2015;45(4–5):241–54.

http://dx.doi.org/10.1016/j.neucli.2015.09.001. Copyright 2015 ElsevierMasson SAS.Allrightsreserved.B.DataaremeanSDproprioceptivedriftcalculatedacross severalvariantsofthefull-bodyillusion.

their feelings and sensations. To further investigate this topic, investigatorsandcliniciansmustknowtheirsemiologyandpursue thispurelyclinicalendeavour.Eveninthisdayand ageof new neuroimaging technologies, data analysis methods and virtual realitysettings,patientsandtheirstoriesremainthemostprecious avenuetodiscovermoreaboutoursenseofbodilyawarenessand ownership.Whenpossible,theclinicalexamshouldinvolveafull neurological,neuropsychological, psychiatric and neuroimaging investigation. The interview should let patients speak by themselves,describe their bodily feelings in their own words, andthenfocusonmoredetailedaspects,suchasthosedescribedin thesection‘‘Disordersofthebodilyself’’ofthisarticle[6].Some questionnairesandguidelineshavebeenpublished[6,106,107].It ishighlyunlikelythatalldisordersofthebodilyselfhavebeen discovered,described and labelled: we thus insist that clinical investigation,casereportsandcaseseriesarestillanabsolutely necessaryapproachforthestudyofthebodilyself.

However,aswehaveseen,experimentalresearchcanexploit multiplemethods and paradigms to further investigate neural, cognitive, affective, spatial and social correlates of bodily awareness.Virtualreality,robotics,neuroprostheticsandincreas- inglyothermethodswillcertainlyhelpdelineatethemechanisms formultisensoryandcerebralbodilyrepresentations.Suchstudies havealreadyrevealedthehighlyplasticnatureofthebodilyself, sensitiveas itis torathersimple visuo-tactile andvisuo-motor conflicts.From these results, these experimental methods now seem to be very promising non-invasive approaches toward rehabilitation of patients with neurologic and psychiatric dis- orders.

Disclosureofinterest

Theauthorsdeclarethattheyhavenocompetinginterest.

Acknowledgements

Someoftheresearchleadingtotheseresultsreceivedfunding fromthePeopleProgramme(MarieCurieActions)oftheEuropean Union’s Seventh Framework Programme (FP7/2007-2013) (REA grant No. 333607) (BODILYSELF, vestibular and multisensory investigationsof bodily self-consciousness). C. Lopez’sresearch is alsosupportedbyagrantfromtheVolkswagenStiftung(grantNo.

89434: Finding Perspective: Determining the embodiment of perspectivalexperience).

References

[1]CritchleyM.Theparietallobes.NewYork:HafnerPress;1953.

[2]He´caenH,deAjuriaguerraJ.Me´connaissancesethallucinationscorporelles.

Inte´grationetde´sinte´grationdelasomatognosie.Paris:Masson;1952.

[3]LhermitteJ.L’imagedenotrecorps.Paris:NouvelleRevueCritique;1939.

[4]Menninger-LerchenthalE.DasTruggebildedereigenenGestalt.Berlin:Kar- ger;1935.

[5]Schilder P. Theimage and appearance ofthe human body. NewYork:

InternationalUniversitiesPress;1935.

[6]DieguezS,StaubF,BogousslavskyJ.Asomatognosia. In:BogousslavskyJ, GodefroyO,editors.Thebehavioralandcognitiveneurologyofstroke.New York:CambridgeUniversityPress;2007.p.215–53.

[7]GallagherS.Howthebodyshapesthemind.NewYork:OxfordUniversity Press;2005.

[8]Thibierge S.Pathologiesde l’imageducorps:e´tudedestroublesdela reconnaissance etde lanomination en psychopathologie. Paris:Presses UniversitairesdeFrance;1999.

[9] BlankeO.Multisensorybrainmechanismsofbodilyself-consciousness.Nat RevNeurosci2012;13:556–71.http://dx.doi.org/10.1038/nrn3292.

[10] BruggerP,LenggenhagerB.Thebodilyselfanditsdisorders:neurological, psychologicalandsocialaspects.CurrOpinNeurol2014;27:644–52.http://

dx.doi.org/10.1097/WCO.0000000000000151.

[11] LopezC,FalconerCJ,DeroualleD,MastFW.Inthepresenceofothers:self- location, balance control and vestibular processing. Neurophysiol Clin 2015;45:241–54.http://dx.doi.org/10.1016/j.neucli.2015.09.001.

[12]GallagherS.Bodyimageandbodyschema:aconceptualclarification.JMind Behav1986;7:541–54.

[13] DeVignemontF.Bodyschemaandbodyimage–prosandcons.Neuropsy- chologia 2010;48:669–80. http://dx.doi.org/10.1016/j.neuropsychologia.

2009.09.022[pii:S0028-3932(09)00378-9].

[14]LopezC,DieguezS.Larepre´sentationducorpsetsestroubles:approchesde laneuropsychologieetdesneurosciencescognitives.In:Pe´rennouD,RodeG, Azouvi P,Brun V,editors. Lacognition spatiale.Montpellier: Sauramps Me´dical;2015.p.63–88.

[15]DieguezS,Annoni JM.Asomatognosia: disordersofthebodilyself. In:

GodefroyO,editor.Vascularcognitiveimpairmentanddementia:thebe- havioralandcognitiveneurologyofstroke.2nded.,Cambridge:Cambridge UniversityPress;2013.p.170–92.

[16]Pick A. Sto¨rrung der Orientierung am eigenen Ko¨rper. Psychol Forsch 1922;1:303–18.

[17]Fe´licianO,CeccaldiM,DidicM,Thinus-BlancC,PoncetM.Pointingtobody parts:adoubledissociationstudy.Neuropsychologia2003;41:1307–16.

[18]GerstmannJ,Fingeragnosie.EineumschriebeneSto¨rungderOrientierungam eigenenKorper.WienKlinWschr1924;37:1010–2.

[19]MayerE,MartoryMD,PegnaAJ,LandisT,DelavelleJ,AnnoniJM.Apurecase ofGerstmannsyndromewithasubangularlesion.Brain1999;122:1107–20.

[20]PetreskaB,AdrianiM,BlankeO,BillardAG.Apraxia:areview.ProgBrainRes 2007;164:61–83.

[21] BerlucchiG,Aglioti SM.Thebodyinthebrainrevisited.Exp BrainRes 2010;200:25–35.http://dx.doi.org/10.1007/s00221-009-1970-7.

[22] LinkenaugerSA,LeyrerM,Bu¨lthoffHH,MohlerBJ.Welcometowonderland:

theinfluenceofthesizeandshapeofavirtualhandontheperceivedsizeand shapeofvirtualobjects.PloSOne2013;8:e68594.http://dx.doi.org/10.1371/

journal.pone.0068594.

[23] BanakouD,GrotenR,SlaterM.Illusoryownershipofavirtualchildbody causesoverestimationofobjectsizesandimplicitattitudechanges.Proc Natl Acad Sci U S A 2013;110:12846–51. http://dx.doi.org/10.1073/

pnas.1306779110.

[24]BotvinickM,CohenJ. Rubberhands‘‘feel’’ touchthateyes see.Nature 1998;391:756.

[25]EhrssonHH.Theexperimentalinductionofout-of-bodyexperiences.Science 2007;317:1048.

[26] ZingerleH.U¨ berSto¨rungen desWahrnehmungdeseigenenKo¨rpersbei organischen Gehirnerkrankungen. Monatsschrift Psychiat Neurol 1913;34:13–36.http://dx.doi.org/10.1159/000203132.

[27]Frederiks JAM. Anosognosie et he´miasomatognosie. Rev Neurol Paris 1963;109:585–97.

[28]CastaigneP,LaplaneD,DegosJD.Troiscasdene´gligencemotriceparle´sion re´tro-rolandique.RevNeurolParis1970;122:234–42.

[29]Garc¸inR,VarayA,DimoH.Documentpourservira` l’e´tudedestroublesdu sche´macorporel.RevNeurolParis1938;69:498–510.

[30] HeydrichL,DieguezS,GrunwaldT,SeeckM,BlankeO.Illusoryownbody perceptions:casereportsandrelevanceforbodilyself-consciousness.Con- scious Cogn 2010;19:702–10. http://dx.doi.org/10.1016/j.concog.2010.

04.010[pii:S1053-8100(10)00130-3].

[31]ArzyS,OverneyLS,LandisT,BlankeO.Neuralmechanismsofembodiment:

asomatognosiaduetopremotorcortexdamage.ArchNeurol2006;63:1022–

5.

[32]BlankeO, Ortigue S, LandisT,SeeckM.Stimulating illusoryown body perceptions.Nature2002;419:269–70.

[33]FrederiksJA.Macrosomatognosiaandmicrosomatognosia.PsychiatrNeurol Neurochir1963;66:531–6.

[34]BonnierP.L’asche´matie.RevNeurolParis1905;12:605–9.

[35] VallarG,RodeG.CommentaryonBonnierP.L’asche´matie.RevNeurol(Paris) 1905;13:605–609. Epilepsy Behav 2009;16:397–400. http://dx.doi.org/

10.1016/j.yebeh.2009.09.001.

[36]RobinsonD,PodollK.Macrosomatognosiaandmicrosomatognosiainmi- graineart.ActaNeurolScand2000;101:413–6.

[37] RodeG,VallarG,RevolP,TiliketeC,Jacquin-CourtoisS,RossettiY,etal.

FacialmacrosomatognosiaandpaininacaseofWallenberg’ssyndrome:

selectiveeffectsofvestibularandtranscutaneousstimulations.Neuro- psychologia2012;50:245–53.http://dx.doi.org/10.1016/j.neuropsycholo- gia.2011.11.018.

[38]EhrenwaldH.Vera¨ndertesErlebendesKo¨rperbildesmitkonsekutiverWah- bildung bei linksseitiger Hemiplegie. Monatsschrift Psychiat Neurol 1930;75:89–97.

[39]WeinsteinEA,KahnRL,MalitzS,RozanskiJ.Delusionalreduplicationofparts ofthebody.Brain1954;77:45–60.

[40] BruggerP.Supernumeraryphantoms:acommentonGrossietal.’s(2002) sparethoughtsonsparelimbs.PerceptMotSkills2003;97:3–10. http://

dx.doi.org/10.2466/pms.2003.97.1.3.

[41] MiyazawaN,HayashiM,KomiyaK,AkiyamaI.Supernumeraryphantomlimbs associatedwithlefthemisphericstroke:casereportandreviewoftheliter- ature. Neurosurgery 2004;54:228–31. http://dx.doi.org/10.1227/01.NEU.

0000097558.01639.F5.

[42]HariR,Ha¨nninenR,Ma¨kinenT,Jousma¨kiV,ForssN,Seppa¨ M,etal.Three hands: fragmentation of human bodily awareness. Neurosci Lett 1998;240:131–4.

[43]McGonigleDJ,Ha¨nninenR,SaleniusS,HariR,FrackowiakRSJ,FrithCD.

Whosearmisitanyway?AnfMRIcasestudyofsupernumeraryphantom limb. BrainJNeurol2002;125:1265–74.