Abstract Spatial orientation is based on coordinates re-ferring to the subject’s body. A fundamental principle isthe mid-sagittal plane, which divides the body and spaceinto the left and right sides. Its neural bases were investi-gated by functional magnetic resonance imaging (fMRI).Seven normal subjects pressed a button when a verticalbar, moving horizontally, crossed the subjective mid-sag-ittal plane. In the control condition, the subjects’ task wasto press a button when the direction of the bar movementchanged, at the end of each leftward or rightward move-ment. The task involving the computation of the mid-sag-ittal plane yielded increased signal in posterior parietal andlateral frontal premotor regions, with a more extensiveactivation in the right cerebral hemisphere. This directevidence in normal human subjects that a bilateral, main-ly right hemisphere-based, cortical network is active dur-ing the computation of the egocentric reference is consis-tent with neuropsychological studies in patients with uni-lateral cerebral lesions. Damage to the right hemisphere,more frequently to the posterior-inferior parietal region,may bring about a neglect syndrome of the contralesio-nal, left side of space, including a major rightward dis-placement of the subjective mid-sagittal plane. The exis-tence of a posterior parietal-lateral premotor frontal net-work concerned with egocentric spatial reference framesis also in line with neurophysiological studies in the mon-key.

Key words Egocentric coordinate frames · Mid-sagittal plane · Functional magnetic resonance imaging · Spatial hemineglect · Right hemisphere

Introduction

Biological organisms, including humans, exist in space,surrounded by objects and living entities. Spatial orienta-tion is a fundamental human ability, which is concernedwith the appreciation of the relationships between exter-nal objects and the body of the observer and the organiza-tion of goal-directed motor behavior, such as reaching for(or avoiding) an object through, for instance, arm move-ment or locomotion. Since the beginning of the century,it has been proposed that perception of the spatial as-pects of the body and the organization of movement in-volve a “schema”, a postural model which is continuous-ly modified by new postures and movement (Critchley1953; Head and Holmes 1911; Schilder 1950). The cere-bral basis of this schema was unclear, but a main role ofthe posterior parietal cortex was repeatedly suggested(Critchley 1953; Frederiks 1969; Mountcastle et al. 1975).The spatial features of this representation of the bodyneed, however, to be qualified in terms of the specificframes of reference involved. The basic coordinate systemis the egocentric reference, which allows the computa-tion of the spatial position of visual objects with respectto the observer.

In the visual input-motor output chain, these egocen-tric co-ordinate systems (e.g., head-centred, body-centred)are interposed (intermediate representations) between sen-sory input (encoded in retinal co-ordinates) and motoroutput, such as an arm reaching movement towards a vi-sual object. These egocentric spatial co-ordinates are com-puted through the integration of inputs from multiple sen-sory sources (visual, proprioceptive-somatosensory, ves-tibular), entailing a co-ordinate transformation from sen-sory (retinotopic) to higher-order spatial egocentric andworld-centred representations or frames of reference (An-dersen et al. 1993, 1997).

G. Vallar · G. Galati · L. PizzamiglioIRCCS S. Lucia, Rome, Italy

E. Lobel · G. Galati · D. Le BihanSHFJ, Department of Medical Research, CEA, Orsay, France

G. Vallar (✉) · G. Galati · L. PizzamiglioIRCCS S. Lucia, Via Ardeatina 306, 00179 Roma, Italye-mail: givallar@imiucca.csi.unimi.itTel.: +39-06-51501358, Fax: +39-06-51501366

E. Lobel · A. BerthozLPPA, Collège de France, and CNRS, Paris, France

Exp Brain Res (1999) 124:281–286 © Springer-Verlag 1999

R E S E A R C H A RT I C L E

Giuseppe Vallar · Elie Lobel · Gaspare GalatiAlain Berthoz · Luigi Pizzamiglio · Denis Le Bihan

A fronto-parietal system for computing the egocentric spatial frame of reference in humans

Received: 11 March 1998 / Accepted: 10 September 1998

The main source of information concerning the neuralbasis of spatial egocentric representations in humans hasbeen, so far, the neurological syndrome of spatial hemi-neglect. These patients may fail to explore the side ofspace contralateral to the lesion and to report stimuli pre-sented in that portion of space, as well as show a de-rangement of egocentric coordinates (Vallar 1998). In themajority of such patients, the lesion involves the posteri-or inferior parietal cortex; damage to the dorsolateralpremotor cortex can also bring about spatial hemineglect(Heilman et al. 1994; Vallar 1993).

In this experiment, we investigated the neural basis ofa fundamental egocentric spatial principle, the median ormid-sagittal plane, which is the plane of bilateral bodysymmetry and divides the subject’s body and extra-per-sonal space into two (left and right) halves. In humans,there is at present no direct evidence about the neuralstructures involved in the computation of this basic frame.We used fMRI to detect activation in the two cerebralhemispheres while subjects executed a spatial-matchingtask, reporting the time at which a vertical bar, movinghorizontally, crossed the subjective mid-sagittal plane.

Materials and methods

Subjects

Seven right-handed male subjects, 21–24 years-old, were exam-ined in the Department of Medical Research, Orsay, France, aftergiving their informed consent. The experimental procedureswere approved by the local ethical committee.

Imaging procedure

All experiments were performed on a 3-Tesla whole-body system(Bruker, Germany) equipped with a quadrature birdcage RF coiland a head-gradient coil insert designed for echo-planar imaging(EPI). Sets of high-resolution images (gradient-echo inversion-re-covery sequence, voxel size=0.86×0.86×2.5 mm) were acquiredfor anatomical identification. Functional images with blood oxy-genation level dependent (BOLD) contrast (Kwong et al. 1992;Ogawa et al. 1990) were obtained with a T2*-weighted single-shotgradient-echo EPI sequence (TE=40 ms, TR=5000 ms, band-width=100 kHz, voxel size=4×4×5 mm). Twenty contiguous 5-mm-thick slices were acquired during each 5-s interval. Visual stimuliwere generated by a personal computer and back-projected througha video-beam onto a screen behind the magnet. Subjects lay su-pine and viewed the screen through a mirror.

Experimental paradigm

During scanning, subjects, in darkness, saw a luminous vertical barmoving horizontally on a black background, alternating betweenleftward and rightward directions of movement. Subjects receivedinstructions to pursue the movement of the bar. In the control con-dition, the subjects’ task was to press a button with their right in-dex finger when the direction of the bar movement changed,namely, at the end of each leftward or rightward movement. In theexperimental condition, the subjects’ task was to press a buttonwhen the bar was “straight ahead”, crossing the subjective mid-sag-ittal plane of the body. The visual stimulation, the motor response,and oculomotor behavior were identical in the two conditions,which differed only in the type of mental computation involved.The sequence of events occurring throughout the experimental ses-sion is shown in Fig. 1.

A measure of the subjects’ accuracy (in degrees) in setting thesubjective midline was computed by averaging the subjects’ dis-placements (the distance of the position of the bar when the sub-ject pressed the response key from the objective mid-sagittal plane).A positive score denoted a rightward displacement, a negative scorea leftward displacement. A measure of the general difficulty of thetwo tasks was computed by averaging the absolute scores of thedisplacements from the objective mid-sagittal plane (in the “mid-sagittal plane” task) and from the position where the bar changeddirection (in the control task).

Psychophysical validation

The tasks were validated in a preliminary psychophysical experi-ment. A separate group of ten right-handed male subjects performedthe experimental tasks outside the fMRI equipment, to rule out dif-ferences in eye movements between the experimental and the con-trol tasks. Visual stimuli were presented on a computer screen indarkness. Eye movements were recorded by means of an infrared-based system (Permobil Meditech, Sweden). The mean ocular pur-suit gain was 0.89 (s.d.: 0.29) during the “mid-sagittal plane” taskand 0.91 (s.d.: 0.22) during the control task [t(9)=0.16, ns]. Sac-cades were less than one per minute in both conditions.

282

Fig. 1 Visual stimulation during the experimental session. Eachexperiment consisted of four consecutive 60-s acquisition ep-ochs, in which 30 s of the control task were followed by 30 s ofthe experimental task. The figure represents the visual stimula-tion over a single epoch. Either task started with the short pre-sentation of a specific symbolic visual cue (green arrows), fol-lowed by a green vertical bar. The bar started from left or right,alternatively, in each epoch (from left in the figure) and movedat a constant speed (2°·s–1) back and forth along an horizontal14° path, centered on the subject’s objective mid-sagittal plane.Each subject had received instructions to press the response keywhen the bar changed direction (control task: small white ar-rows) or crossed the subjective mid-sagittal plane (“mid-sagittalplane” task: small white arrows). For each task, four responses wererecorded during each epoch

Group analysis

Image analyses were performed with statistical parametric map-ping (Friston et al. 1991, 1994; Worsley et al. 1992), usingSPM96b software. The scans from each subject were first correct-ed for head movement, resampled, and transformed into standardstereotaxic space, based upon data provided by the MNI (ICBM,NIH P-20 project), with a final voxel size of 4×4×4 mm. Imageswere then spatially and temporally smoothed with a gaussian filter(Friston et al. 1995a). The effects of the experimental condition oneach and every voxel was then estimated with an analysis of cova-riance (ANCOVA), including global signal changes as a confound-ing covariate (Friston et al. 1990, 1995b). The resulting statisticalmap was thresholded at z=3.09 (or P=0.001, uncorrected).

We defined a volume of interest (VOI), comprising four brainregions hypothesized to be involved in the computation of themid-sagittal plane: the posterior parietal lobe and the dorsolateralpremotor cortex in the left and in the right hemisphere (see Intro-duction). These four brain regions were each composed of 400 vox-els. Activated clusters belonging to the VOI were considered sig-nificant at P<0.05, corrected for multiple comparisons over the VOI.Clusters outside the VOI were considered significant at P<0.05,corrected for multiple comparisons over the entire brain volume(Worsley et al. 1995).

Individual analysis

Individual functional maps were obtained by computation of theautocorrelation peak of the MRI-signal time-course on a pixel-by-pixel basis. This approach enables phase-independent and robust de-

283

Fig. 2 Regions of the brainactivated during the “mid-sagit-tal plane” task. Activated regions are superimposed on athree-dimensional view of astandard brain. Lateral views ofthe two hemispheres and aview from the top (in the mid-dle) are shown. Parietal regionsare shown in red and frontal re-gions in green

Table 1 “Mid-sagittal plane” task. Peaks of increased activity in-side the activated regions. For each peak, the anatomical locationsare given in accordance to a standard coordinate system (Talairachand Tournoux 1988) and as a Brodmann’s area (BA). Peak Z-val-

ues, regional spatial extents and corrected P-values were derivedfrom the statistical parametric map. Number of subjects showing asignificant activation of each brain region (results of the individualanalysis)

Regions Extents Peaks BA x y z Z scores Corrected No. ofP-values subjects

Left 14 voxels Superior occipital gyrus 19 –28 –80 36 4.14 0.002 5/7parieto-occipital (896 mm3)Right 84 voxels Intraparietal sulcus 7 28 –84 44 4,87parieto-occipital (5376 mm3) Angular gyrus 39 52 –64 36 3.38 0.0003 5/7Left frontal 4 voxels Precentral sulcus 44 –48 8 28 3.23 0.04 3/7

(256 mm3)Right frontal 69 voxels Precentral sulcus 6 48 0 40 4.24

(4416 mm3) Inferior frontal gyrus 44 52 8 20 3.95 0.001 6/7Inferior frontal gyrus 44 44 12 20 3.72

tection of a periodic temporal response (Paradis et al. 1997). Three-dimensional clusters of more than two voxels showing an autocor-relation peak greater than 0.5 were retained as activated. This thresh-old was determined to correspond to an estimated uncorrected P-value of 2.10–4 based on an analysis of 6000 images acquired inthe “resting” condition (null hypothesis), in the same subjects, dur-ing the same acquisition session, and with the same acquisition se-quences and parameters as the experimental images. Activation fo-ci were defined in relation to the individual anatomy, superimpos-ing activated clusters on the individual anatomical images, and werethen compared across subjects.

Results

Subjects were very accurate in computing the mid-sagittalplane (mean error: +0.06°, s.d.: 0.27). The mean absolutedisplacement was 0.49° (s.d.: 0.17) in the “mid-sagittalplane” task and 0.59° (s.d.: 0.13) during the control task,suggesting that the general levels of difficulty of the twotasks were comparable (paired t-test, df=5, t=1.82, ns).

The cerebral regions activated during the “mid-sagittalplane” task, compared to the control condition, are shownin Fig. 2 superimposed on a three-dimensional view ofthe brain surface, and in Fig. 3 in the form of axial slices.Table 1 reports the anatomical locations and extents ofthe activated regions.

A bilateral pattern of activation in the posterior pari-etal and in the lateral premotor cortices was found, whichwas more extensive in the right hemisphere, where 89%of the activated voxels were located. The parietal regionincluded the posterior section of the intraparietal sulcusand parts of the adjacent cortex in the superior parietallobule and in the superior occipital gyrus, bilaterally. Inthe right hemisphere, it extended ventrally into the angu-lar gyrus of the inferior parietal lobule. The frontal regionwas located inside and around the precentral sulcus. Inthe right hemisphere, it extended ventrally and anteriorlyinto the inferior frontal gyrus.

Table 1 also shows the number of individual subjectsin whom activation of these regions was found. Parietalactivations were located inside the intraparietal sulcusand in the angular gyrus, frontal activations in the inferi-or and middle frontal gyri, inside the precentral sulcus.Other areas of increased signal (at the given uncorrectedthreshold) included: in the right hemisphere, the supple-mentary motor area and the fundus of the central and post-central sulci in two subjects, the paracentral lobule, theinsula and the cingulate cortex in one subject; in the lefthemisphere, the supplementary motor area, the paracen-tral lobule, the insula, the posterior inferior temporal sul-cus, and the area enthorinalis in one subject.

Discussion

This experiment yielded two main results. First, a bilat-eral set of frontal premotor and posterior parietal regionswas active during the task of setting the subjective mid-sagittal plane. Second, a clear hemispheric asymmetrywas found, with a prevailing activation of right hemi-sphere regions.

This fronto-parietal pattern provides precise anatom-ico-functional evidence, which is consistent, in generalterms, with the indirect data resulting from neuropsycho-

logical studies in patients with unilateral cerebral lesionsand spatial hemineglect. This deficit is more frequent andsevere in patients with damage to the right hemisphere(Bisiach and Vallar 1988; Vallar 1993, 1998), who mayshow a major dysfunction of egocentric frames of refer-ence, displacing the subjective mid-sagittal plane right-wards, towards the side of the lesion; patients with dam-age to the left hemisphere, by contrast, make a minor er-ror (Heilman et al. 1983; Karnath 1994; Mark and Heil-man 1990; Vallar et al. 1995). The neural correlates ofleft visuo-spatial hemineglect include damage to the rightinferior parietal lobule (BAs 39, 40) and, although lessfrequently, to the dorso-lateral premotor cortex (BAs 8,6, 44) (Damasio et al. 1980; Heilman, et al. 1994; Vallar1993; Vallar and Perani 1986), with a possible more rele-vant role of lesions of the inferior frontal gyrus (Husainand Kennard 1996). These non-symmetrical pathologicaleffects of unilateral brain damage suggest that a premo-tor frontal and posterior-inferior parietal network, with amajor right hemisphere-based component, provides a rel-evant contribution to the computation of egocentric framesof reference in humans.

Also, functional imaging activation studies in hu-mans, investigating aspects of spatial attention differentfrom the egocentric reference, have provided evidencefor the existence of premotor frontal-posterior parietalnetworks involved in spatial processing. During motor ex-ploration of space (Gitelman et al. 1996) and lateral ori-entation of visuo-spatial attention (Nobre et al. 1997), thepremotor (BA 6) and posterior parietal cortices (Bas 7,39, 40) have been reported to be active, with a prevailingor exclusive lateralization to the right cerebral hemisphere.

Finally, the results of studies in behaving monkeys arein line with the present findings, even though they do notaddress the issue of hemispheric asymmetries. Cells whichencode visual space, combining retinal position of thestimulus with eye position signals (Andersen et al. 1985),head position signals, or both (Brotchie et al. 1995)

284

Fig. 3 Regions of the brain ac-tivated during the “mid-sagittalplane” task, superimposed on aseries of axial slices of a stan-dard brain. Contiguous 4-mmslices are shown, starting from20 mm above the bicommissu-ral line (top left) until 48 mmabove the bicommissural line(bottom right). Activated areasare shown in yellow. The lefthemisphere is on the left side

(= eye and head position-sensitive visual neurons), havebeen found in the posterior parietal cortex (lateral intra-parietal area and area 7a). Cells with retinotopic receptivefields, whose response is affected by eye position havealso been found in the premotor cortex (Boussaoud et al.1993, area 6). Such neurons, whose activity is modulatedby the angle of gaze and the position of the head, maycontribute to a body-centered representation of visualspace (Andersen et al. 1997).

There is also some evidence for a more direct coding.Some neurons in the ventral intraparietal area (Duhamelet al. 1997), in the premotor cortex (area 6) (Fogassi etal. 1992, 1996: caudal part of inferior area 6, area F4;Graziano et al. 1994: ventral premotor cortex), and in theparieto-occipital area (Galletti et al. 1993) respond to thevisual stimulation of the same spatial location, regardlessof eye position, encoding visual space in non-retinal (spa-tial) coordinates. The precise nature of these spatial ref-erence frames is a matter of debate. Paradigms in whichthe monkey’s head is fixed and aligned with the trunk donot distinguish between coordinates related to specificbody parts, such as cranio-centered versus trunk-centeredframes of reference (Duhamel et al. 1997; Fogassi et al.1992; Galletti et al. 1993). It also remains possible thatthese neurons code the position of visual objects in allo-centric coordinates, related to stimuli present in the envi-ronment. In a recent study (Fogassi et al. 1996), howev-er, rotation of the monkey, which changed the scene seenby the animal, did not affect the response of premotorneurons with visual receptive fields independent of eyeposition, suggesting, therefore, a body-centered coding.In addition, most premotor neurons were bimodal, visu-al-tactile, with visual receptive fields located around theskin in the peripersonal space. Finally, in the ventral pre-motor cortex, bimodal neurons with a tactile response onthe arm have visual receptive fields in the space adjacentto the arm, which move when the arm moves, but notwhen the eyes move; in other premotor neurons with atactile response on the face, the visual receptive fieldmoves as the head is rotated (Graziano et al. 1997). Thesefindings suggest coding in body-centered (arm, head) co-ordinates. Taken together, these observations indicate thatthe non-retinal coding of visual stimuli by neurons in thepremotor cortex may be qualified in terms of body-part,egocentric coordinates. In the present experiment, theexecution of a task very different from those used in themonkey, which required, however, the computation andutilization of egocentric coordinates, was associated withactivation in the dorsolateral premotor cortex.

The present study indicates that the computation of themid-sagittal egocentric reference is subserved by a bilater-al set of brain regions, which is more extensive on theright side and includes both premotor frontal and posteriorparietal regions. The anatomical connections betweenthese areas (Cavada and Goldman-Rakic 1989a,b; Matelliet al. 1986; Mesulam et al. 1977) suggest a close function-al interaction. Their specific contributions to the develop-ment of nonsensory, spatial coordinates may, however, dif-fer in important respects. A distinction between perceptual

versus premotor/exploratory aspects of spatial representa-tion and attention (Mesulam 1990) captures a number ofneuropsychological dissociations in patients with lefthemineglect (Bisiach et al. 1990, 1995; Vallar 1993). Theposterior parietal cortex may provide non-retinocentric co-ordinates (egocentric and allocentric), but also some high-er-order motor planning (Andersen et al. 1997; Snyder etal. 1997), with the premotor cortex subserving spatial rep-resentations (a “movement-based space”: Rizzolatti et al.1997), specified in terms of body-part-centered coordi-nates (Graziano et al. 1997). Be that as it may, the compu-tation of a basic egocentric reference, such as the subjec-tive mid-sagittal plane, appears to involve the activity of afronto-parietal system, which, as related neuropsycholog-ical and neurophysiological evidence suggests, plays amost relevant role in spatial cognition.

Acknowledgements This study was supported in part by grantsfrom MURST, CNR, Italian Ministry of Health, Institut de Forma-tion Superieure Biomedicale, French Ministry of Research, and byEC BHM1-CT94-1133 from the BIOMED program of the Europe-an community. We are grateful to Jean-Baptiste Poline for his ad-vice on the image analysis.

References

Andersen RA, Essick GK, Siegel RM (1985) Encoding of spatiallocation by posterior parietal neurons. Science 230:456–458

Andersen RA, Snyder LH, Li C-S, Stricanne B (1993) Coordinatetransformations in the representation of spatial information. CurrOpin Neurobiol 3:171–176

Andersen RA, Snyder LH, Bradley DC, Xing J (1997) Multimodalrepresentation of space in the posterior parietal cortex and itsuse in planning movements. Annu Rev Neurosci 20:303–330

Bisiach E, Vallar G (1988) Hemineglect in humans. In: Boller F,Grafman J (eds) Handbook of neuropsychology, vol I. Else-vier, Amsterdam, pp 195–222

Bisiach E, Geminiani G, Berti A, Rusconi ML (1990) Perceptual andpremotor factors of unilateral neglect. Neurology 40:1278–1281

Bisiach E, Tegnèr R, Làdavas E, Rusconi ML, Mijovic D, Hjalta-son H (1995) Dissociation of ophthalmokinetic and melokinet-ic attention in unilateral neglect. Cereb Cortex 5:439–447

Boussaoud D, Barth TM, Wise SP (1993) Effects of gaze on ap-parent visual responses of frontal cortex neurons. Exp Brain Res93:423–434

Brotchie PR, Andersen RA, Snyder LH, Goodman SJ (1995) Headposition signals used by parietal neurons to encode locationsof visual stimuli. Nature 375:232–235

Cavada C, Goldman-Rakic PS (1989a) Posterior parietal cortex inrhesus monkey. I. Parcellation of areas based on distinctive lim-bic and sensory corticocortical connections. J Comp Neurol 287:393–421

Cavada C, Goldman-Rakic PS (1989b) Posterior parietal cortex inrhesus monkey. II. Evidence for segregated corticocortical net-works linking sensory and limbic areas with the frontal lobe. JComp Neurol 287:422–445

Critchley M (1953) The parietal lobes. Hafner, New YorkDamasio AR, Damasio H, Chang Chui H (1980) Neglect follow-

ing damage to frontal lobe or basal ganglia. Neuropsychologia18:123–132

Duhamel J-R, Bremmer F, BehHomed S, Graf W (1997) Spatialinvariance of visual receptive fields in parietal cortex neurons.Nature 389:845–848

Fogassi L, Gallese V, Pellegrino G di, Fadiga L, Gentilucci M,Luppino G, Matelli M, Pedotti A, Rizzolatti G (1992) Spacecoding by premotor cortex. Exp Brain Res 89:686–690

285

Fogassi L, Gallese V, Fadiga L, Luppino G, Matelli M, RizzolattiG (1996) Coding of peripersonal space in inferior premotorcortex (area F4). J Neurophysiol 76:141–157

Frederiks JAM (1969) Disorders of body schema. In: Vinken PJ,Bruyn GW (eds) Handbook of clinical neurology. Disorders ofspeech, perception and symbolic behaviour. North Holland,Amsterdam, pp 207–240

Friston KJ, Frith CD, Liddle PF, Dolan RJ, Lammertsma AA,Frackowiak RSJ (1990) The relationship between global andlocal changes in PET scans. J Cereb Blood Flow Metab 10:458–466

Friston KJ, Frith CD, Liddle PF, Frackowiak RSJ (1991) Compar-ing functional (PET) images: the assessment of significantchange. J Cereb Blood Flow Metab 11:690–699

Friston KJ, Worsley KJ, Frackowiak RSJ, Mazziotta JC, EvansAC (1994) Assessing the significance of focal activations us-ing their spatial extent. Hum Brain Map 1:214–220

Friston KJ, Ashburner J, Poline JB, Frith CD, Heather JD, Fracko-wiak RSJ (1995a) Spatial registration and normalization ofimages. Hum Brain Map 2:165–189

Friston KJ, Holmes AP, Worsley KJ, Poline JB, Frith CD, Fracko-wiak RSJ (1995b) Statistical parametric maps in functionalimaging: a general approach. Hum Brain Map 2:189–210

Galletti C, Battaglini PP, Fattori P (1993) Parietal neurons encod-ing spatial locations in craniotopic coordinates. Exp Brain Res96:221–229

Gitelman DR, Alpert NM, Kosslyn S, Daffner K, Scinto L, Thomp-son W, Mesulam M-M (1996) Functional imaging of humanright hemispheric activation for exploratory movements. AnnNeurol 39:174–179

Graziano MSA, Yap GS, Gross CG (1994) Coding of visual spaceby premotor neurons. Science 266:1054–1057

Graziano MSA, Tian Hu X, Gross CG (1997) Visuospatial proper-ties of ventral premotor cortex. J Neurophysiol 77:2268–2292

Head H, Holmes G (1911) Sensory disturbances from cerebral le-sions. Brain 34:102–254

Heilman KM, Bowers D, Watson RT (1983) Performance on hemi-spatial pointing task by patients with neglect syndrome. Neu-rology 33:661–664

Heilman KM, Watson RT, Valenstein E (1994) Localization of le-sions in neglect and related disorders. In: Kertesz A (ed) Lo-calization and neuroimaging in neuropsychology. AcademicPress, San Diego, pp 495–524

Husain M, Kennard C (1996) Visual neglect associated with fron-tal lobe infarction. J Neurol 243:652–657

Karnath H-O (1994) Subjective body orientation in neglect andthe interactive contribution of neck muscle proprioception andvestibular stimulation. Brain 117:1001–1012

Kwong KK, Belliveau JW, Chesler DA (1992) Dynamic magneticresonance imaging of human brain activity during primarysensory stimulation. Proc Natl Acad Sci USA 89:5675–5679

Mark VW, Heilman KM (1990) Bodily neglect and orientationalbiases in unilateral neglect syndrome and normal subjects. Neu-rology 40:640–643

Matelli M, Camarda R, Glickstein M, Rizzolatti G (1986) Afferentand efferent projections of the inferior area 6 in the macaquemonkey. J Comp Neurol 255:281–298

Mesulam M-M (1990) Large-scale neurocognitive networks anddistributed processing for attention, language and memory. AnnNeurol 28:597–613

Mesulam M-M, Van Hoesen GW, Pandya DN, Geschwind N(1977) Limbic and sensory connections of the inferior parietallobule (area PG) in the rhesus monkey: a study with a newmethod for horseradish peroxidase histochemistry. Brain Res136:393–414

Mountcastle VB, Lynch JC, Georgopoulos A, Sakata H, Acuna C(1975) Posterior parietal association cortex of the monkey: com-mand functions for operations within extrapersonal space. J Neu-rophysiol 38:871–908

Nobre AC, Sebestyen GN, Gitelman DR, Mesulam MM, Fracko-wiak RSJ, Frith CD (1997) Functional localization of the systemfor visuospatial attention using positron emission tomography.Brain 120:515–533

Ogawa S, Lee T-M, Kay AR, Tank DW (1990) Brain magneticresonance imaging with contrast dependent on blood oxygen-ation. Proc Natl Acad Sci USA 87:9868–9872

Paradis A-L, Mangin J-F, Cornilleau-Pérès V, Bloch I, Frouin V,Le Bihan D (1997) Detection of periodic temporal responsesin fMRI. Neuroimage 5:S469

Rizzolatti G, Fadiga L, Fogassi L, Gallese V (1997) The spacearound us. Science 277:190–191

Schilder P (1950) The image and appearance of the human body.International Universities Press, New York

Snyder LH, Batista AP, Andersen RA (1997) Coding of intentionin the posterior parietal cortex. Nature 386:167–170

Talairach J, Tournoux P (1988) Co-planar stereotaxic atlas of thehuman brain. Thieme, New York

Vallar G (1993) The anatomical basis of spatial hemineglect in hu-mans. In: Robertson IH, Marshall JC (eds) Unilateral neglect:clinical and experimental studies. Lawrence Erlbaum, Hove,pp 27–59

Vallar G (1998) Spatial hemineglect in humans. Trends Cogn Sci2:87–97

Vallar G, Perani D (1986) The anatomy of unilateral neglect afterright hemisphere stroke lesions. A clinical CT/Scan correla-tion study in man. Neuropsychologia 24:609–622

Vallar G, Guariglia C, Nico D, Bisiach E (1995) Spatial hemine-glect in back space. Brain 118:467–472

Worsley KJ, Evans AC, Marrett S, Neelin P (1992) A three-di-mensional statistical analysis for rCBF activation studies inhuman brain. J Cereb Blood Flow Metab 12:900–918

Worsley KJ, Marret S, Neelin P, Friston KJ, Evans A (1995) Aunified statistical approach for determining significant signalsin images of cerebral activation. Academic Press, San Diego

286