Tactile Awareness and Limb
Position in Neglect:
Functional Magnetic
Resonance Imaging
Nathalie Valenza, BA,1 Mohamed L. Seghier, PhD,1,2
Sophie Schwartz, PhD,1,3 François Lazeyras, PhD,2 and
Patrik Vuilleumier, MD1,4

We studied a patient with right parietal damage for
whom tactile stimuli on the right/ipsilesional hand (projecting to the intact left hemisphere) were extinguished
from awareness during double simultaneous stimulation,
when his right hand was positioned in the left/contralesional space. This demonstrates the role of an egocentric
spatial reference frame in attention that can determine
awareness of stimuli despite intact sensory pathways. Using functional magnetic resonance imaging to elucidate
the neural correlates of such perceptual extinction, we
found that limb position modulated neural responses to
tactile stimuli at early cortical stages (SI) in the intact
hemisphere. Activity in bilateral middle frontal gyri also
was modulated by limb position and may contribute to
integrate sensory inputs into a supramodal, egocentric
representation of space.
Ann Neurol 2004;55:139 –143

Unilateral parietal lesions in humans often cause spatial
neglect and perceptual extinction, in which patients
may remain unaware of stimuli contralateral to the lesion, in the absence of major sensory deficits. Extinction may concern tactile inputs, characterized by a failure to report a contralesional touch when stimuli are
simultaneously delivered to both hands, although single
contralesional touches are correctly detected.1,2 Several
explanations have been proposed for this loss of tactile
awareness on double simultaneous stimulation (DSS).
Recent theories argue that extinction may result from a

From the 1Laboratory of Neurology and Imaging of Cognition, Department of Neurology and Neurosciences and 2Department of Radiology, University of Geneva, Switzerland; 3Institute of Cognitive
Neuroscience, University College London, United Kindgom; and
4
School of Educational Sciences and Psychology, University of Geneva, Switzerland.
Received April 14, 2003, and in revised form July 28. Accepted for
publication Oct 15, 2003.

Address correspondence to Dr Vuilleumier, Laboratory NIC, Department of Neurosciences and Physiology, University Medical Center, 1 Rue Michel-Servet, 1211 Geneva 4, Switzerland.
E-mail: patrik.vuilleumier@medecine.unige.ch

spatial bias in attention affecting the processing of inputs on the contralesional side, according to various
frames of spatial representation.1–3 However, the role
of subtle sensory dysfunction is still unclear. Other accounts suggest that extinction might reflect a degradation or slowing in processing of sensory stimuli.4,5 In
support of spatial attentional mechanisms, patients
have been reported in whom tactile extinction was
modulated by position of the affected hand relative
trunk6 or to the location of the other hand.3 Thus,
tactile extinction may improve when the left hand is
positioned in the right rather than left hemispace6 or
when hands are crossed, in either hemispace.3 Tactile
extinction also may arise on an intact/ipsilesional limb,
under conditions in which awareness of tactile stimuli
is determined by their relative spatial position on the
limb.7 Other findings also support a role of hemispatial
factors in neglect and extinction.8
Here, we describe a patient who presented with a
striking deficit in tactile perception on his right/ipsilesional hand when it was placed in the left side of space,
after a focal right parietal lesion. Using functional magnetic resonance imaging (fMRI) in this patient, we
were able to determine the neural correlates of tactile
extinction for stimuli delivered on an intact limb, without any concomitant sensory deficit potentially associated with contralesional tactile extinction. Critically, we
sought to determine brain areas showing differential responses on DSS as a function of the spatial position of
the right hand, paralleling the patient’s changes in perceptual awareness.
Case Report
A 68-year-old right-handed man, suffered two hemorrhagic strokes in right superior parietal lobe in 1998
and 2000. MRI scan at the time of our study in 2002
showed residual right parietal damage (Fig 1A, B). The
patient showed marked signs of left visual neglect in
the acute stage (eg, 14 –20% left omissions in cancellation tests; 7–9% rightward displacement in line bisection) but only mild signs at the time of our study
(eg, slower detection of visual targets in left versus right
hemifield [mean, 429 vs 272 milliseconds], but no
omission in cancellation tests, and no visual extinction). A detailed neuropsychological examination
showed that other cognitive functions were normal. He
had minimal weakness and decreased position sense in
the left arm but no deficit in superficial touch, pain,
and temperature. During clinical testing of tactile extinction, he reported 20 of 20 single touches on the
left (contralesional) hand, but only 1 of 20 during DSS
of both hands. Right-hand touches were always correctly perceived in these conditions.
A modulation of tactile extinction by hand position
in egocentric space was demonstrated by delivering single or DSS stimuli to the right hand and to the right

© 2003 American Neurological Association
Published by Wiley-Liss, Inc., through Wiley Subscription Services

139

Fig 1. (A) T1-weighted magnetic resonance imaging (MRI) scan of the brain in our patient showing a chronic lesion in right parietal lobe (white arrows), extending in white matter. Only slight white matter disease was visible in the posterior left periventricular
region, and the patient had no clinical left hemispheric symptoms. (B) Three-dimensional reconstruction of MRI showing the damaged cortical area. (C) Experimental testing of tactile extinction on the intact right hand was performed with the hand placed to
the right or left of the trunk’s midline (stimulation on the ulnar edge of the hand, elbow, both, or none), with eye closed, in a similar way both outside (percentage correct detection in boldface) and inside (percentage correct detection in italics) the scanner.

elbow of the patient (ie, the ipsilesional limb), in two
different spatial conditions (see Fig 1C). Tactile stimuli
were applied by a conventional procedure2 using a rapid
flexion extension of the examiner’s finger (40 trials in
each condition). DSS on right hand and right elbow
produced a marked extinction of hand stimuli when the
hand was on the left of the patient’s trunk, with the
patient reporting only the right elbow touch on 31 of 40
trials. In contrast, detection of hand stimuli on DSS trials was much better with the hand placed to the right of
his trunk (only 8/40 misses; ␹2(1) ⫽ 26.5, p ⬍ 0.001).
Single stimuli on both hand and elbow were always accurately reported (0/40 missed). These data demonstrate
significant extinction on the intact right hand, dependent on this hand being in left hemispace. Incidentally,
we also note that some extinction occurred for righthand touches even in right space during DSS to hand
and elbow, consistent with other clinical observations of
frequent extinction for the more distal of two tactile
stimuli delivered on the same limb.9
Functional Imaging Methods
During fMRI scanning, tactile extinction on the right
hand was tested in a similar way as outside the scanner,
either on the right or left side of the trunk, with eyes
closed (see Fig 1C). A MRI-compatible device (attached
to the patient’s arm by a splint) was used to deliver light
taps with constant intensity and constant intervals, time-

140

Annals of Neurology

Vol 55

No 1

January 2004

locked by a computer (45 trials in each condition). Four
stimulus types (rest; single hand; single elbow; hand ⫹
elbow DSS) were applied in short blocks (24 seconds)
consisting of five stimuli of a particular type (SOA approximately 4 seconds), plus one stimulus from another
type to limit expectations biases. Each condition was repeated three times in random order during a session,
with four sessions performed in the two different hand
positions (ABBA). The patient reported the number of
felt stimuli (one or two) by left-hand key presses.
In addition, a separate hand-localizer scan was performed, during which four blocks of passive right-hand
stimulation (24 seconds) and four blocks of rest (24
seconds) were given in alternation, with the right hand
in right space.
Whole-brain fMRI was performed on a 1.5T MR system, using an echo planar imaging GRE sequence (TR/
TE/flip, 2 seconds/40 milliseconds/80 degrees; FOV,
250mm; matrix, 128 ⫻ 128; 19 contiguous 5mm axial
slices, resolution ⫽ 1.95 ⫻ 1.95mm2). Data were processed and analyzed on a voxel-by-voxel basis using
SPM99,10 including eight different conditions (four
stimulation types ⫻ two limb positions) in the extinction scan session, and two conditions in the handlocalizer session (hand-stimulation and rest). Only activations surviving a threshold at p value 0.001 voxelwise
are reported.

Results
During fMRI scanning, as observed in behavioral testing outside the scanner, DSS on the right (ipsilesional)
hand and elbow resulted in a severe extinction of hand
stimuli when this hand was in left space (DSS-L: 66%
missed), compared with when the hand was in right

space (DSS-R: 26% missed; ␹2(1) ⫽ 23.73, p ⬍
0.001). Hand position did not affect detection of single stimuli on either the hand (11% vs 6% missed;
␹2(1) ⫽ 0.39, NS) or elbow (6% missed in each condition). Elbow stimuli were correctly detected during
DSS regardless of limb position.

Fig 2. (A) Activation by single-hand more than single-elbow stimuli were found in left postcentral (SI) and opercular (SII) somatosensory cortex. (B) Activation modulated by hand position during double simultaneous stimulation (DSS-R ⬎ DSS-L) occurred in
similar left postcentral regions and bilateral middle frontal gyri. (C) Region of interest in postcentral somatosensory cortex (SI) defined by an independent localizer-scan (hand dorsum ⬎ rest), in which average activity across each block condition showed a consistent increase during single-hand and DSS stimuli only when the hand was in right space, but not in left space.

Valenza et al: fMRI of Tactile Extinction

141

Our fMRI analysis first determined neural responses
elicited by single right-hand stimuli, irrespective of hand
position in space (hand ⬎ rest). Significant activation
was found in postcentral gyrus and parietal operculum
of the left hemisphere (Fig 2A), corresponding to human primary (SI) and secondary (SII) somatosensory areas, respectively.11,12 Bilateral parietal regions, precentral
frontal cortex, and left superior occipital gyrus also were
activated (see the Table for complete list of activations
and normalized coordinates). Greater responses to hand
than to elbow stimulation were specifically found in a
lateral part of the left post-central gyrus, parietal operculum, and superior parietal cortex. Similar results are typically found during somatosensory stimulation of the
hand in normal subjects.12
The next critical analysis determined brain regions
where somatosensory responses were modulated by
hand position during DSS condition, that is, the neural correlates of tactile extinction when the intact/right
hand was placed in contralesional/left space. Comparing
activation produced by DSS with the right hand on the
right versus left side (DSS-R ⬎ DSS-L) showed greater
responses in the left postcentral and precentral gyri, as
well as left superior medial parietal cortex and bilateral
middle frontal gyri (see Fig 2B). When we restricted

our analysis to areas specifically activated by hand stimulation (DSS-R ⬎ DSS-L, masked inclusively by
hand ⬎ elbow contrast), activation again was found
specifically in left postcentral gyrus and bilateral superior parietal cortex. These results indicate that placing
the right hand in left space reduced responses to tactile
stimuli in somatosensory areas of the intact hemisphere, in parallel with the striking loss of awareness
for hand stimuli in these conditions.
To confirm our findings, we also defined an a priori
region of interest in the postcentral somatosensory cortex based on a separate hand-localizer scan (passive
right-hand stimulation vs rest, in right space). Figure
2C shows the time course of activity in this SI region
across the different conditions of the extinction scan.
Greater responses to single-hand stimulation and DSS
were observed when the right hand was in right space.
In contrast, positioning the right hand in the left space
suppressed responses in SI to the same tactile stimuli.
For completeness, we also determined areas whose
response was independent of hand position (hand ⬎
rest, masked exclusively by DSS-R ⬎ DSS-L). This
showed activation in left parietal operculum (SII), left
inferior frontal sulcus, and right posterior parietal cortex (see Table).

Table. Brain Areas Activated in Each Condition
Coordinatesa
Side

Brain Areas

x

y

z

Responses to right-hand stimulation (single-hand stimuli ⬎ rest)
⫺68
⫺26
1
L
Parietal operculum (SII)c
L
Posterior inferior frontal gyrus
⫺62
14
25
L
Postcentral gyrus (SI)c
⫺51
⫺40
57
L
Superior occipital gyrus
⫺25
⫺81
48
L
Medial paracentral cortex
⫺9
⫺1
53
L
Posterior parietal cortexc
⫺33
⫺72
57
R
Superior parietal cortex
56
⫺30
49
R
Paracentral
6
⫺5
74
R
Superior parietal (perilesional)
4
⫺37
66
R
Parietal operculum (SII)c
57
27
17
Effects of hand position (DSS in right space ⬎ DSS in left space)
L
Anterior middle frontal gyrus
⫺42
34
19
L
Postcentral gyrus (SI)c
⫺40
⫺32
55
L
Precentral gyrus
⫺44
⫺8
46
L
Superior precentral sulcus
⫺47
18
50
L
Posterior parietal cortexc
⫺61
⫺34
42
R
Anterior middle frontal gyrus
36
41
23
R
Superior frontal sulcus
8
60
33
Responses unaffected by hand position (single-hand stimuli ⬎ rest, masked exclusively by DSS-R ⬎ DSS-L)
L
Parietal operculum (SII)
⫺68
⫺26
1
L
Posterior-inferior frontal gyrus
⫺61
14
32
R
Posterior parietal cortex
56
⫺30
50
a

Coordinates of activation peaks in the patient are converted into the standard normalized MNI space.
All activations p ⬍ 0.001 at the voxel level, p ⬍ 0.05 at the cluster level.
Also hand ⬎ elbow (z ⬎ 3.11, p ⬍ .001).

b
c

DSS ⫽ double simultaneous stimulation.

142

Annals of Neurology

Vol 55

No 1

January 2004

Z-scoreb
5.33
5.28
5.25
4.86
3.89
3.37
5.68
4.58
4.36
3.66
3.84
3.75
3.64
3.62
3.52
4.19
3.50
5.33
5.23
4.96

Discussion
Our study is the first to our knowledge to show neural
correlates for impaired awareness of tactile stimuli on
an intact limb, associated with right parietal damage
and mild left spatial neglect. Salient tactile stimuli on
the right hand were extinguished when this hand was
placed in left space, relative to the patient’s midline.
Such influences of position demonstrate the role of
higher spatial coordinates in attention and extinction,3,6,7 in the absence of primary sensory loss.
Our fMRI results establish that spatial biases can affect early stages of cortical processing, including primary somatosensory areas. With the right hand in left
space, neural responses during DSS decreased in postcentral regions corresponding to SI and in superior parietal cortex of the intact left hemisphere, as well as in
bilateral prefrontal regions. This supports previous
findings in animals13 and healthy humans11,14 that attentional factors can modulate early sensory areas, including SI and SII. Our findings of greater hand position effects on SI than SII appears also consistent with
some suggestions that there might not be a strict serialhierarchical flow from SI to SII.15,16
A recent positron emission tomography study17 similarly showed decreased activation in the right primary
somatosensory area (SI) of patients with left tactile extinction (both during DSS and unilateral stimulation),
raising the possibility of a primary sensory deficit in
tactile extinction. Our findings for an intact/ipsilesional limb strongly suggest that primary somatosensory cortex may be influenced by spatial attention and
be directly involved in awareness of tactile stimuli.18
Note that averaged activity in SI cortex as defined by
our localizer scan showed that, in left space, even single
right-hand stimulation produced smaller fMRI responses, although behaviorally the patient still correctly
reported these single right-hand stimuli. This suggests
that weaker neural responses in SI for stimuli in contralesional space might lead to their erasure from
awareness only in the presence of competing simultaneous stimuli in ipsilesional space, on DSS trials.
Importantly, placing the right hand in left space also
decreased responses in left posterior parietal cortex and
bilateral middle frontal gyri. These frontal regions may
participate to a supramodal representation of space19
that, together with parietal cortex, integrate tactile inputs with proprioceptive information into bodycentered coordinates.19,20 Dysfunction in such a spatial
representation after right parietal lesion might impair
awareness for stimuli processed in intact sensory pathways.

This work was supported by a Swiss National Science Foundation
grant (632-065935, P.V.).

We thank Dr T. Landis and A. Schnider and F. Terrier for their
help.

References
1. Vallar G. Spatial frames of reference and somatosensory
processing: a neuropsychological perspective. Philos Trans R
Soc Lond B Biol Sci 1997;352:1401–1409.
2. Ladavas E, di Pellegrino G, Farne A, Zeloni G. Neuropsychological evidence of an integrated visuotactile representation of
peripersonal space in humans. J Cogn Neurosci 1998;10:
581–589.
3. Aglioti S, Smania N, Peru A. Frames of reference for mapping
tactile stimuli in brain-damaged patients. J Cogn Neurosci
1999;11:67–79.
4. Birch HG, Belmont I, Karp E. Delayed information processing
and extinction following cerebral damage. Brain 1967;90:
113–130.
5. Farah MJ, Monheit MA, Wallace MA. Unconscious perception
of “extinguished” visual stimuli: reassessing the evidence. Neuropsychologia 1991;29:949 –958.
6. Smania N, Aglioti S. Sensory and spatial components of somaesthetic deficits following right brain damage. Neurology 1995;
45:1725–1730.
7. Moscovitch M, Behrmann M. Coding of spatial information in
the somatosensory system: evidence from patients with neglect
following parietal lobe damage. J Cogn Neurosci 1994;6:
151–155.
8. Coslett HB, Schwartz MF, Goldberg G, et al. Multi-modal
hemispatial deficits after left hemisphere stroke. A disorder of
attention? Brain 1993;116:527–554.
9. Critchley M. The phenomenon of tactile inattention with special reference to parietal lesions. Brain 1949;72:538 –561.
10. Friston KJ, Holmes AP, Worsley KJ, et al. Statistical parametric
maps in functional imaging: A general linear approach. Hum
Brain Mapp 1995;2:189 –210.
11. Burton H, Sinclair RJ. Attending to and remembering tactile
stimuli: a review of brain imaging data and single-neuron responses. J Clin Neurophysiol 2000;17:575–591.
12. Paulesu E, Frackowiak RSJ, Bottini G. Maps of somatosensory
systems. In: Frackowiak RSJ, Friston KJ, Frith CD, et al., eds.
Human brain functions. San Diego: Academic Press, 1997:
183–242.
13. Hyvarinen J, Poranen A, Jokinen Y. Influence of attentive behavior on neuronal responses to vibration in primary somatosensory
cortex of the monkey. J Neurophysiol 1980;43:870 – 882.
14. Hamalainen H, Hiltunen J, Titievskaja I. Activation of somatosensory cortical areas varies with attentional state: an fMRI
study. Behav Brain Res 2002;135:159 –165.
15. Iwamura Y. Hierarchical somatosensory processing. Curr Opin
Neurobiol 1998;8:522–528.
16. Zhang HQ, Murray GM, Turman AB, et al. Parallel processing
in cerebral cortex of the marmoset monkey: effect of reversible
SI inactivation on tactile responses in SII. J Neurophysiol 1996;
76:3633–3655.
17. Remy P, Zilbovicius M, Degos JD, et al. Somatosensory cortical activations are suppressed in patients with tactile extinction:
a PET study. Neurology 1999;52:571–577.
18. Meador KJ, Ray PG, Echauz JR, et al. Gamma coherence and
conscious perception. Neurology 2002;59:847– 854.
19. Graziano MS. Where is my arm? The relative role of
vision and proprioception in the neuronal representation
of limb position. Proc Natl Acad Sci USA 1999;96:
10418 –10421.
20. Lloyd DM, Shore DI, Spence C, Calvert GA. Multisensory representation of limb position in human premotor cortex. Nat
Neurosci 2003;6:17–18.

Valenza et al: fMRI of Tactile Extinction

143