Brain (2000), 123, 2531–2541
Neural consequences of acting in near versus far
space: a physiological basis for clinical
dissociations
Peter H. Weiss,1,2 John C. Marshall,4 Gilbert Wunderlich,1,2 Lutz Tellmann,1 Peter W. Halligan,5
Hans-Joachim Freund,2 Karl Zilles1,3 and Gereon R. Fink1,2
1Institut
für Medizin, Forschungszentrum Jülich, Jülich,
Klinik and 3C. and
O. Vogt-Hirnforschungsinstitut, Heinrich-Heine-Universität,
Düsseldorf, Germany, 4Neuropsychology Unit, University
Department of Clinical Neurology, Radcliffe Infirmary,
Oxford and 5School of Psychology, Cardiff University,
Cardiff, UK
2Neurologische
Correspondence to: Dr Gereon R. Fink, Institut für
Medizin (IME), Forschungszentrum Jülich, 52425 Jülich,
Germany
E-mail: [email protected]
Summary
We used PET to determine which brain regions are
implicated when normal volunteers bisect horizontal lines
and point to dots in near (peripersonal) or far
(extrapersonal) space. Studies of line bisection in patients
with right hemisphere lesions have shown that bisection
performance can be severely impaired in either near or
far space while remaining within normal limits in the other
spatial domain. Likewise, clinical dissociations between
pointing to objects in near and far space have been
reported. The normal functional anatomy of these
dissociations has not been demonstrated convincingly.
Regional cerebral blood flow measurements using PET
were carried out in 12 healthy right-handed male
volunteers who bisected lines or pointed to dots in near
or far space, using a laser pen. Subjects performing either
task in near space showed neural activity in the left dorsal
occipital cortex, left intraparietal cortex, left ventral
premotor cortex and left thalamus. In far space, subjects
performing either task showed activation of the ventral
occipital cortex bilaterally and the right medial temporal
cortex. These data provide physiological support for
the clinically observed dissociations demonstrating that
attending to and acting in near space differentially
employs dorsal visuomotor processing areas, whereas
attending to and acting in far space differentially draws
on ventral visuoperceptual processing areas, even when
the motor components of the tasks are identical when
performed in the two spaces.
Keywords: line bisection; pointing; positron emission tomography; visuomotor processing; visuoperceptual processing
Abbreviations: ANOVA ⫽ analysis of variance; rCBF ⫽ regional cerebral blood flow; SPM ⫽ statistical parametric map
Introduction
It has long been suspected that the ‘visual perception of
space within and without arm’s reach may be subserved by
different central mechanisms liable, perhaps, to selective
impairment’ (Paterson and Zangwill, 1944). One of the first
demonstrations of such a dissociation was reported in a case
of right hemisphere glioblastoma: the patient was impaired
in pointing to objects in near space without comparable
difficulty for objects in far space (Brain, 1941).
Studies of radial line bisection performed within arm’s
reach (peripersonal space) have shown that bilateral temporooccipital lesions can be associated with significant
misbisection towards the body. This pattern of performance
has been interpreted as ‘far’ neglect (Shelton et al., 1990).
© Oxford University Press 2000
By contrast, lesions of bilateral parieto-occipital cortex have
been associated with significant misbisection away from the
body—‘near’ neglect (Mennemeier et al., 1992). In both
cases, however, the lesions are too extensive to support strong
neuroanatomical localization. Other clinical studies have
shown that unilateral lesions of the right parietal lobe may
produce neglect of both left and near peripersonal space.
This conclusion has been drawn from the fact that, in
cancellation tasks, the largest number of omissions is typically
found in the inferior left quadrant of the display (Morris
et al., 1985; Halligan and Marshall, 1989).
Although the clinical evidence is far from conclusive, the
overall pattern of results has suggested that the ventral visual
2532
P. H. Weiss et al.
stream is implicated primarily in attending to far space,
whereas the dorsal visual stream is implicated primarily in
attending to (and acting in) near space (Heilman et al., 1990;
Previc, 1990). Studies of line bisection in normal subjects
are consistent with this line of argument (Geldmacher and
Heilman, 1994). When subjects bisect lines oriented radially
in the sagittal plane, the nature of the error pattern differs
when the lines are presented above eye level (upper visual
field/ventral stream) versus below eye level (lower visual
field/dorsal stream). When they are presented above eye
level, radial lines are misbisected distant to the true centre
(near pseudoneglect). When presented below eye level, radial
lines are misbisected significantly closer to the body (far
pseudoneglect). Nonetheless, it must be stressed that the
studies described above do not contrast performance within
arm’s reach (peripersonal space) with performance outside
arm’s reach (extrapersonal space).
However, behaviour dissociations can be seen that support
the original distinction made between spatial perception
‘within and without arm’s reach’ (Paterson and Zangwill,
1944). Thus, left visuospatial neglect in near (peripersonal)
space has been found in association with normal performance
in far (extrapersonal) space (Halligan and Marshall, 1991).
Contrariwise, severe left neglect in far (extrapersonal) space
with good performance in near (peripersonal) space has been
reported (Vuilleumier et al., 1998). These dissociations in
patients with left neglect after right hemisphere lesions have
been investigated with horizontal line bisection and a laser
pen (Halligan and Marshall, 1991; Cowey et al., 1994;
Vuilleumier et al., 1998), so that both the action required to
bisect and the visual angle subtended by the stimulus line
can be held constant in near and far space. Although
behavioural dissociations between lateral neglect in near
versus far space have been well established in both humans
and monkeys (Rizzolatti et al., 1983), the large
cerebrovascular lesions that typically provoke chronic visual
neglect in humans do not throw much light on the precise
neural mechanisms that represent the two spatial domains.
Accordingly, we report a study of horizontal line bisection
and pointing in near (peripersonal) and far (extrapersonal)
space with healthy volunteers using PET to isolate the brain
regions responsible for the clinical dissociations. The factorial
design of the experiment also allowed us to investigate
whether the neural mechanisms involved in the particular
clinical task performed (line bisection or dot pointing) interact
with the neural bases underlying action in near and far space.
Subjects and methods
Subjects
Twelve healthy, right-handed male volunteers (age
29.7 ⫾ 10 years) with no history of neurological or psychiatric
illness were recruited. We studied only male volunteers in
order to avoid the normal variation in brain size and shape
between the sexes, and hence improve image normalization
(see below). Informed consent was obtained before
participation. The study was approved by the ethics committee
of the University Hospital, Düsseldorf, and permission to
administer radioactivity was obtained from the responsible
federal authorities.
Experimental design
The experimental design was factorial, with the factors ‘task’
(line bisection versus pointing) and ‘space’ (near versus far).
Subjects lay comfortably in the PET scanner holding a laser
pointer in their supported right hand (Fig. 1). An intravenous
cannula was placed in the left cubital vein for injection of the
radioactive tracer. In the ‘line bisection’ task, subjects were
asked to point with the laser to the centres of horizontal lines
of varying length; the lines were presented pseudorandomly in
the four screen quadrants to prevent subjects from holding a
representation of the centre of the lines in mind. In the
‘pointing’ task, the laser pointer was used to point to a dot,
which was displayed pseudorandomly where the centres of the
lines had been in the bisection task. Subjects were always
instructed to try to make a single ballistic movement to the
centre of the line or to the dot. In both tasks, subjects were
asked to return their gaze and the laser pointer to the centre of
the display to wait for the appearance of the next visual
stimulus. Thus, the motor requirements of the two tasks (line
bisection and pointing) were closely similar, whereas the
visuospatial cognitive operations differed. For bisection,
subjects had to perform an implicit line centre judgement
followed by a visuomotor transformation to direct the laser
pointer to the judged centre. In the pointing task, a reactive
movement towards the dot had to be executed.
Videotape recordings were used for monitoring and off-line
performance analyses. For the comparison of near versus far
space, the visual stimuli were displayed either on a monitor
screen within reaching distance (eye-to-monitor distance 0.7 m,
i.e. near space) (Fig. 1) or beyond reaching distance (eye-toscreen distance 1.7 m, i.e. far space) (Fig. 1). The geometrical
configurations of the screens were carefully matched, so that
both screens covered a horizontal visual angle of 27° and a
vertical visual angle of 21°. The horizontal lines corresponded
to a visual angle of 8.5° to 9°. The displayed dots extended
over a visual angle of 0.3°. The width of the laser beam was
4 mm in near space and 5 mm in far space. Each of the resulting
four conditions (line bisection, near space; line bisection, far
space; pointing, near space; pointing, far space) was replicated
three times per subject, giving a total of 144 observations (12
scans, 12 subjects). The order of conditions was
counterbalanced within and across subjects.
Eye movements and behavioural measurements
For the analysis of differential eye movements across
conditions, eye movements were recorded during the rCBF
measurements with an infrared device (iView system;
SensoMotoric Instruments, Teltow, Germany) (Fig. 1).
Action in near versus far space
2533
Fig. 1 Experimental set-up. A computer monitor within reaching range was used for visual stimulation
in near space. A screen beyond reaching range, on which the visual stimuli were projected, was used
for far space presentation.
Because of apparatus failure, data from four of the 12 subjects
could not be used. First, the eye movement data were analysed
for contamination by eye-blinks. This led to the exclusion of
~5% of the total data acquired. Eye position was computed
using the normalized x and y coordinates of the subject’s gaze
on the monitor and screen, respectively. For each of the four
conditions, the data were analysed for the length of the total
scan path (given by the number of pixels traversed) and the
dwell time during which the subjects kept their gaze in the
proximity of the stimuli. The dwell time was computed by
summing the intervals (in ms) during which the subject’s gaze
was within predefined frames encompassing the area in which
the stimuli appeared on the monitor screen. These frames
corresponded to the four quadrants of the monitor, omitting its
central part (4.2° ⫻ 21.0°). Epochs of eye movement
measurement were averaged across all stimuli in each visual
quadrant during each condition period (90 s). The mean values
for each experimental condition were compared using twoway analysis of variance (ANOVA) with the factors space
(near, far) and task (line bisection, dot pointing).
Task performance was recorded with the aid of a video
camera during PET scanning (Fig. 1). Behavioural
measurements were taken off-line by determining for each
stimulus on the video screen the absolute distance (in mm)
from the optimal position of the laser pen’s dot (guided by the
subject) to the displayed dot and the real centre of the line. This
absolute metric distance was converted into a relative value (a
percentage), thus normalizing for the different line lengths.
The mean relative values of each experimental condition were
compared using two-way ANOVA as before.
PET scanning
Regional cerebral blood flow (rCBF) was measured by
recording the regional distribution of cerebral radioactivity
after the intravenous injection of [15O]butanol, which is a
positron emitter with a half-life of 2.05 min. The PET
measurements were made with a Siemens ECAT EXACT HR⫹
PET scanner (CTI, Knoxville, Tenn., USA), with a total axial
field of view of 155 mm covering the whole brain. Data were
acquired in 3D mode with interdetector collimating septa
removed and a Neuro-Insert installed to limit the acceptance
of events originating from out-of-field-of-view activity (i.e.
the whole body). The transaxial resolution was determined
with a point source to be 4.4 mm at the centre, increasing to
4.8 mm (tangential) and 8.0 mm (radial) at a distance of 20 cm
for the 3D acquisition mode. The average axial resolution
varied between 4.1 mm at the centre to 7.8 mm at a radius
of 20 cm.
For each measurement of relative rCBF, 555 MBq of
[15O]butanol was given intravenously as a bolus injection. The
whole study involved the administration of 5 mSv effective
dose equivalent of radioactivity per subject. Twelve
consecutive PET scans were collected, each beginning when
an activity threshold of 5% above the background level was
exceeded. Emission data were thereafter collected sequentially
2534
P. H. Weiss et al.
over 40 s and corrected for background. This process was
repeated for each emission scan, with 9 min between scans to
allow adequate decay of radioactivity. All emission scan data
were corrected for scattered events and for the effects of
radiation attenuation (e.g. by the skull) by means of a
transmission scan taken before the first relative rCBF
measurement. The corrected data were reconstructed into 63
transverse planes (separation 2.4 mm) and into 128 ⫻ 128
pixels (size 2.1 mm) by 3D filtered back-projection using a
Hanning filter with a cut-off frequency of 0.5 cycles per pixel.
MRI
On a separate occasion, a high-resolution anatomical MRI of
each subject’s brain was obtained (i) to exclude the possibility
of morphological or pathological abnormalities, and (ii) for
stereotaxic normalization into the standard anatomical space
(Talairach and Tournoux, 1988). This imaging was performed
with a Vision 1.5 T whole-body scanner (Siemens, Erlangen,
Germany) using the 3D magnetization-prepared, rapid gradient
echo (MPRAGE) sequence with the following parameters:
echo time ⫽ 4.4 ms, repetition time ⫽ 11.4 ms, flip angle ⫽
15°, inversion time ⫽ 300 ms, matrix ⫽ 200 ⫻ 256, field of
view ⫽ 20 cm, 128 sagittal slices, slice thickness ⫽ 1.33 mm.
factors task and space and their interactions were estimated on
a pixel-by-pixel basis using SPM97d. Task-related differences
in global CBF, within and between subjects were removed by
treating global activity as a covariate (Friston et al., 1995b).
This removed systematic state-dependent differences in global
blood flow associated with the different conditions, which can
obscure task-related regional alterations in activity. For each
pixel in stereotaxic space, the ANCOVA (analysis of
covariance) generated a condition-specific adjusted mean
rCBF value (arbitrarily normalized to 50 ml/min) and an
associated adjusted error variance (Friston et al., 1995b). This
allowed the planned comparisons of the mean blood flow
distributions across all sets of conditions. For each pixel, across
all subjects and all scans, the mean relative rCBF values were
calculated separately for each of the main effects. The means
were compared with the t statistic and thereafter transformed
into normally distributed Z statistics. The resulting set of Z
values constituted a statistical parametric map (SPM{Z} map;
Friston et al., 1995b). For the contrasts of interest, the
significances of these statistical parametric maps were assessed
by comparing the expected and observed distribution of the t
statistic under the null hypothesis of no differential activation
effect on rCBF. Only activations that were significant at
P ⬍ 0.001 (uncorrected for multiple comparisons) or better
are reported.
Image processing
All calculations and image manipulations were performed with
an ULTRA 10 workstation (Sun Microsystems, Palo Alto,
Calif., USA) using MATLAB version 4 (Mathworks, Natick,
Mass., USA). Statistical parametric mapping software
(SPM97d, Wellcome Department of Cognitive Neurology,
London, UK; www.fil.ion.ucl.ac.uk/spm) was used for image
realignment, image normalization and smoothing, and to create
statistical maps of significant relative changes in rCBF (Friston
et al., 1995a, b).
To correct for head movement, all PET scans were realigned
to the first emission scan using SPM97d software. A mean
relative rCBF image was created for each subject. Each
individual’s MRI and PET mean image (serving as a template
for the individual PET images) were co-registered and then
transformed into a standard stereotaxic anatomical space using
linear proportions and a non-linear sampling algorithm. The
PET images were thereafter filtered using a low-pass Gaussian
filter (resulting in an image resolution of 10 mm) to reduce
the variance due to individual anatomical variability and to
improve the signal-to-noise ratio (Friston et al., 1995a). The
resulting pixel size in stereotaxic space was 2 ⫻ 2 mm with an
interplane distance of 4 mm. Data were then expressed in terms
of standard stereotaxic coordinates (x, y, z) as defined in Table 1
(Talairach and Tournoux, 1988).
Statistical analysis
After stereotaxic normalization and image smoothing,
statistical analysis was performed. The main effects of the
Localization of activations
The stereotaxic coordinates of the pixels of local maximum
significant changes in relative rCBF within areas of significant
relative rCBF change associated with the different factors were
determined. These local maxima were localized anatomically
by reference to a standard stereotaxic atlas (Talairach and
Tournoux, 1988). Additional validation of this method of
localization was obtained after superimposition of the SPM{Z}
maps on the group mean MRI image calculated after each
individual’s MRI had been stereotaxically transformed into the
same standard stereotaxic space.
Results
PET activations
Greater neural activity associated with near space (relative to
far space) was observed in areas within the dorsal visuomotor
stream: the dorsal occipital cortex and the parietal cortex along
the intraparietal sulcus. In addition, the thalamus and ventral
premotor cortex were activated (Table 1A and Fig. 2). By
contrast, the reverse comparison of far versus near space
revealed differential neural activity in areas within the ventral
visuoperceptual stream: the ventral occipital cortex bilaterally
and the right medial temporal cortex (Table 1B and Fig. 3).
Action in near space showed relatively greater activation of
left hemisphere sites despite the fact that all responses in all
conditions were made with the right hand. The activation focus
in the left dorsal occipital cortex during performance in near
Action in near versus far space
2535
Table 1 Relative increases in brain activity during performance of the line bisection or the pointing task and increases
associated with action in near and far space
Region
Side
Coordinates
Z score
x
y
z
(A) Main effect of near space: (NB ⫹ NP) ⬎ (FB ⫹ FP)
Dorsal occipital cortex (1)
Intraparietal cortex (2)†
Intraparietal cortex (3)†
Ventral premotor cortex (4)
Thalamus (5)
L
L
L
L
L
–18
–46
–40
–60
–10
–86
–36
–28
⫹4
–18
⫹20
⫹52
⫹34
⫹34
⫹8
5.1*
4.1**
4.4**
3.4**
3.7**
(B) Main effect of far space: (FB ⫹ FP) ⬎ (NB ⫹ NP)
Ventral occipital cortex (6)
Ventral occipital cortex (7)
Medial temporal cortex (8)
R
L
R
⫹16
–24
⫹20
–92
–94
–40
–16
–14
–8
5.0*
4.0**
3.9**
Inferior parietal cortex
(Dorsal) premotor cortex
R
L
R
L
R
R
⫹26
–12
⫹22
–22
⫹34
⫹26
–92
–88
–64
–60
–48
–6
⫹14
–6
⫹64
⫹64
⫹40
⫹56
5.6*
6.0*
5.0*
4.9*
3.9**
4.5*
(D) Main effect of pointing task: (NP ⫹ FP) ⬎ (NB ⫹ FB)
Medial extrastriate cortex
L
–8
–68
⫹32
3.7**
(C) Main effect of line bisection task: (NB ⫹ FB) ⬎ (NP ⫹ FP)
Occipital cortex
Superior parietal cortex
The table shows brain regions where relative increases in rCBF were associated with each comparison of interest. For each region of
activation, the coordinates in standard stereotaxic space (Talairach and Tournoux, 1988) are given; they refer to the maximally activated
focus within an area of activation, as indicated by the highest Z score: x ⫽ distance (mm) to right (⫹) or left (–) of the midsagittal
plane; y ⫽ distance anterior (⫹) or posterior (–) to vertical plane through the anterior commissure; z ⫽ distance above (⫹) or below (–)
the intercommissural (AC–PC) plane. All k values (extent of activated area) of regions reported are significant at P ⬍ 0.05. NB ⫽ line
bisection task in near space; FB ⫽ line bisection task in far space; NP ⫽ pointing task in near space; FP ⫽ pointing task in far space.
†These coordinates refer to the two local maxima within an area of activation extending along the intraparietal sulcus. *P ⬍ 0.05
(corrected for multiple comparisons); **P ⬍ 0.001 (uncorrected for multiple comparisons).
space lay in the vicinity of left upper V3. The activation foci
in the ventral occipital cortex bilaterally during performance
in far space were located closely to the lower portions of right
and left V3 (Shipp et al., 1995).
For the line bisection conditions, we replicated the activation
of the right parietal cortex observed previously during line
bisection judgements (Fink et al., 2000). Line bisection relative
to pointing activated the extrastriate cortex bilaterally, the
superior parietal cortex bilaterally and the right inferior parietal
cortex (Table 1C). An additional area of activation was
observed in the dorsal premotor cortex in the vicinity of the
right frontal eye field (Paus, 1996). In the comparison of
pointing versus line bisection, increases in rCBF were found
only in the left medial extrastriate cortex (Table 1D).
No significant interaction of task (line bisection, pointing)
and spatial domain (near, far) was observed.
[F(1,28) ⫽ 4.2; P ⬍ 0.05], although there was a trend with dot
pointing for longer dwell times in the upper field for attending
to and acting in near space and in the lower field for attending
to and acting in far space. The pattern was different with line
bisection. Here dwell times were slightly higher in the lower
hemifield during near space performance, and slightly higher
in the upper hemifield during far space performance.
All responses in both the pointing and the line bisection
condition were accurate to within 0.2°, which corresponds
to the off-line measurement accuracy on the video screen.
Additional analyses of the behavioural data showed no
significant difference in performance accuracy between tasks
(line bisection versus dot pointing) and between spatial
domains (near versus far).
Discussion
Eye movements and behavioural measures
The scan path and the dwell time around the stimuli in each
visual quadrant did not differ significantly across conditions
(Table 2). Neither the main effects of space or task nor the
interaction terms reached the predefined level of significance
The differential neural bases for attending to and subsequently
acting in near and far space in the intact human brain
irrespective of the specific demands of the task constitute the
main findings of our study. These distinct neural activations
when an identical action is performed in response to a
stimulus in near versus far space provide strong physiological
2536
P. H. Weiss et al.
Fig. 2 Relative rCBF increases associated with action in near space (P ⬍ 0.001, uncorrected, or better).
(Upper row) Sagittal and transverse views, shown as through-projections onto representations of
standard stereotaxic space (Talairach and Tournoux, 1988). (Lower rows) Transverse SPM{Z} maps
superimposed upon the group mean MRI, which had been spatially normalized into the same stereotaxic
space. The height of the transverse section (the z coordinate is given above each slice) was selected to
show the local maximum within the activated brain area (activations are numbered according to Table
1A). For action in near space, areas within the dorsal stream of visuomotor processing are activated: the
left dorsal occipital cortex (1), the left intraparietal cortex (2 and 3) and the ventral premotor cortex (4).
The left thalamus was also activated (5). P ⫽ posterior; A ⫽ anterior; R ⫽ right; L ⫽ left.
support for clinical evidence of dissociations between these
spatial domains in pointing and line bisection tasks (although
in this experiment we cannot assess the respective
contributions of acting in near and far space versus the
processing of the stimuli to which action is directed). That
we found no significant interaction between task and spatial
domain further suggests that patients who show a dissociation
between performance in near and far space on one task
should show the same dissociation on other tasks, although
the dissociation may be more pronounced in visuomotor
tasks (Vuilleumier et al., 1998). The patient of Vuilleumier
and colleagues showed very mild left neglect in far, but not
near, space in perceptual tasks that did not require a manual
response (for example, reading words, and judging whether
Action in near versus far space
2537
Fig. 3 Relative rCBF increases associated with action in far space (P ⬍ 0.001 uncorrected, or better).
The activations are numbered according to Table 1B. For explanation and abbreviations, see Fig. 2.
Action in far space resulted in activations located in the ventral visuoperceptual processing stream: the
right medial temporal cortex (8) and bilateral ventral occipital cortex (6 and 7).
Table 2 Eye movement data
Condition 1
Condition 2
Condition 3
Condition 4
Level of significance
Near space,
line bisection
Near space,
dot pointing
Far space,
line bisection
Far space,
dot pointing
Near space
vs. far space
Line bisection
vs. dot pointing
Scan path (pixels)
41 030
(2508)
42 529
(2508)
40 542
(4095)
43 500
(2681)
F(1,28) ⫽ 0.01
n.s.
F(1,28) ⫽ 0.54
n.s.
Right, upper
quadrant (ms)
12 841
(2662)
15 120
(2662)
14 426
(4347)
6600
(2846)
F(1,28) ⫽ 1.17
n.s.
F(1,28) ⫽ 0.75
n.s.
Right, lower
quadrant (ms)
16 007
(3438)
13 714
(3438)
18 009
(5615)
20 814
(3676)
F(1,28) ⫽ 1.21
n.s.
F(1,28) ⫽ 0.01
n.s.
Left, upper
quadrant (ms)
19 497
(3457)
22 097
(3457)
26 045
(5645)
11 620
(3696)
F(1,28) ⫽ 0.22
n.s.
F(1,28) ⫽ 2.01
n.s.
Left, lower
quadrant (ms)
20 957
(5169)
18 931
(5169)
15 641
(8441)
28 334
(5526)
F(1,28) ⫽ 0.11
n.s.
F(1,28) ⫽ 0.73
n.s.
Upper hemifield
(ms)
32 338
(5436)
37 216
(5436)
40 471
(8878)
18 219
(5812)
F(1,28) ⫽ 0.69
n.s.
F(1,28) ⫽ 1.76
n.s.
Lower hemifield
(ms)
36 964
(7213)
32 644
(7213)
33 650
(11779)
49 147
(7711)
F(1,28) ⫽ 0.58
n.s.
F(1,28) ⫽ 0.41
n.s.
The table shows the scan path (number of pixels traversed by the subject’s gaze on the monitor or screen) and the dwell time (ms) for
different areas of the monitor or screen. Data are means with standard error of the means in parenthesis.
2538
P. H. Weiss et al.
squares are complete or have a gap in them) (Vuilleumier
et al., 1998). Nonetheless, left neglect in far space was much
more severe in tasks that did require a manual response, such
as cancellation and line bisection.
It also follows that models of line bisection per se do not
need to postulate distinct mechanisms for the performance
of this task in near versus far space (Anderson, 1996). The
task itself draws primarily upon circuits that include the
superior and inferior parietal cortex. Performing the task in
near or far space then implicates areas in the dorsal or
ventral occipital cortex, respectively, that are more generally
concerned with the visuomotor tasks which are undertaken
in one of these two spatial domains. Dissociations between
line bisection performance in near and far space after brain
damage thus suggest that sufficient cortex may remain
functional in the right parietal cortex to calculate the midpoint
of a line, but that the responsible regions cannot communicate
appropriately with areas that are more generally concerned
with near or far space. A similar ‘disconnection’ argument
could apply to dissociations between pointing in near and
far space. This conjecture may be applicable to the patient
of Halligan and Marshall, who showed left neglect for near
but not far space (Halligan and Marshall, 1991): structural
MRI showed some sparing of the right superior and medial
parietal cortex, the latter region being undercut by deep white
matter lesions. In cases in which destruction of the right
parietal cortex is more complete, the extent to which left
parietal regions can substitute for the spatial functions of the
right requires further investigation. That many patients with
severe left neglect make a good recovery suggests that some
rebalancing of the contralateral attentional vector of each
hemisphere can take place after right parietal lesion
(Kinsbourne, 1993).
The dorsal occipital cortex represents the lower visual field
(Previc, 1990), and thus (typically) near space. Seemingly
consistent with this topographical organization is our
observation of activation of the dorsal occipital cortex during
action in near space. However, there were no visual field
differences in stimulus presentation between the near and
far space conditions. Also, no significant differential eye
movements between near and far conditions were observed
that could explain the differential activations in extrastriate
areas. Thus, the activation of the dorsal occipital cortex
during near conditions is likely to reflect enhanced neural
processing in the dorsal visual stream during action in near
space. Such enhancement would be equivalent to the neural
gating effects described previously in studies of visual
attention in human and non-human primates (Moran and
Desimone, 1985; Corbetta et al., 1990; Fink et al., 1997). It is
now well established that attentional processes can influence
neural activity in early visual areas by ‘top-down’ modulation
from the temporoparietal and frontal cortex (Desimone, 1998;
Kastner et al., 1998).
The differential activation of the left (posterior) parietal
cortex during action in near space is likewise consistent with
this gating hypothesis. The role of the posterior parietal
cortex in space representation and movement guidance is well
established (Andersen et al., 1997). Single-cell recordings in
non-human primates have revealed neurones in area 7b with
visual receptive fields that respond to movements of stimuli
near the face or arm, but not to stimuli in far space more
than 1 m away (Leinonen et al., 1979). In accordance with
these electrophysiological findings, we observed differential
activation of the posterior parietal cortex during near but not
far conditions. Taken together, the data strongly support the
view that the posterior parietal cortex is a control region for
the operation of the limbs, hands, and eyes within near space
(Mountcastle et al., 1975).
The differential activation of the premotor cortex during
action in near space follows the functional anatomy of the
parietal–premotor connections known to be involved in the
control of visually guided arm movements (Geyer et al.,
2000). The premotor region activated in the near space
conditions in our experiment may be functionally equivalent
to the inferior frontal area 6 in non-human primates. The
inferior premotor cortex of the monkey (inferior area 6aα)
encodes visual stimuli adjacent to the hand or arm in armcentred coordinates (Graziano et al., 1994). This
representation of the space near the body is used for the
visual control of reaching (Graziano and Gross, 1998). In
the monkey, inferior frontal area 6 receives input from area
7b (rostral part of the inferior parietal lobules), forming part
of the circuit that guides movements in near space, and which
when lesioned produces near visual neglect (Rizzolatti et al.,
1983). That performance in near space selectively activates
left-hemisphere regions is initially puzzling in an experiment
in which the stimuli, tasks and form of response are identical
across all conditions (Table 1A). One plausible explanation
involves the fact that the left hemisphere ‘leads’ the planning
and execution of skilled movements of the dominant right
hand in near space. There may accordingly be a closer
coupling between performance in near space and the
involvement of the left hemisphere even when the response
is a simple pointing movement (Barrett et al., 1998). That
is, although similar movements were made by the right hand
in both near and far conditions, the fact that objects can be
touched and manipulated in near (but not far) space
additionally boosts the activity of the hemisphere that controls
the hand of response.
Nonetheless, there is an apparent discrepancy here with
the results of lesion studies: if it is primarily left hemisphere
sites that are activated by performing tasks in near space,
why should right hemisphere lesions provoke neglect of
near (left) space (Halligan and Marshall, 1991; Berti and
Frassinetti, 2000)? One might rather have expected from
the functional neuroimaging results that patients with right
parietal damage would be biased to near space because of
preserved left hemisphere functions. There is some indication
that this may be so, in that patients with neglect for far space
with relative preservation of performance in near space seem
to be more common than patients who show the opposite
dissociation (Cowey et al., 1994). An explanation would still
Action in near versus far space
be required for the ‘anomalous’ patients of Halligan and
Marshall (Halligan and Marshall, 1991) and Berti and
Frassinetti (Berti and Frassinetti, 2000), who show neglect
in near space with relative preservation of performance in
far space. Berti and Frassinetti have suggested that the
anteromedial aspect of the right frontal lobe (preserved in
both their patient and the patient of Halligan and Marshall)
may suffice to code for far space (Berti and Frassinetti, 2000).
Activation of the ventral occipital cortex during action in
far space is consistent with the fact that stimuli in far space
typically project to the upper visual fields (Previc, 1990).
However, as already argued, these activations in the ventral
extrastriate cortex cannot result directly from upper visual
field stimulation. Rather, the results suggest that the ventral
occipital cortex is differentially implicated in visually guided
action in far space irrespective of purely retinotopic
considerations. The right medial ventral temporal cortex has
previously been associated with visual feature extraction and
shape analysis (Dolan et al., 1997; Kanwisher et al., 1997;
George et al., 1999). The coordinates we report for the
maximally activated focus (x ⫽ ⫹20, y ⫽ –40, z ⫽ –8 mm)
are strikingly close to those of the ‘parahippocampal place
area’ (x ⫽ ⫹18, y ⫽ –39, z ⫽ –6 mm), which are activated
in processing scenes and landmarks (Epstein and Kanwisher,
1998). Although our experiment was conducted in a darkened
room, both the monitor and the screen reflected sufficient
light for the subject to see the surrounding environment.
More of this environment was visible when the screen was
in far space than when the monitor was in near space. This
difference may explain why the medial ventral temporal
cortex is activated more strongly when the stimulus is
presented in far space. There is accordingly a possible
confound in our experiment: it may be that the increased
activity in ventral areas in the far condition resulted, in part,
from the subject seeing more of the environment around the
scanner and possibly the objects therein (although the scanner
environment was darkened). The number of objects in the
scene would be greater in the tasks performed in far space,
with possible attentional competition between those objects.
If ventral object recognition processes were indeed more
engaged by the far tasks, this confound would, of course, be
typically found in the real world. Looking into far space
almost invariably means seeing (even if not foveating) many
objects between the viewer and the focus of interest in
extrapersonal space.
It is unlikely that the association of dorsal stream
activations with performance in near space and ventral stream
activations during performance in far space (Table 1) is due
to retinotopic factors. The eye movement data (Table 2) do not
reveal any reliable overall association between the distance of
the stimuli (near/far) and dwell time in the lower and upper
visual fields. Indeed, the (non-significant) association between
stimulus presentation in near and far space and dwell time
in the lower and upper fields reverses in direction between
the two tasks of dot pointing and line bisection (see Results).
The relative increases in brain activation during performance
2539
in near versus far space (the main effects of distance are
shown in Table 1) do not, therefore, seem to be due to the
difference in dwell time in the upper and lower visual fields.
Finally, we note that the boundary between near
(peripersonal) and far (extrapersonal) space is not fixed
(Cowey et al., 1999). In the monkey, the receptive fields that
encode near space enlarge when the animal uses a stick to
reach for a food pellet (Iriki et al., 1996). A result consistent
with this extension of near space has been found in a patient
with left visual neglect in near but not far space, as assessed
by line bisection with a laser pen. When the patient was
given a stick with which to bisect in far space, left neglect
was again apparent (Berti and Frassinetti, 2000). These
findings suggest that relevant tool use modulates access to
the neural representations of near and far space. More
generally, the observations reported here strongly indicate
that the visual world surrounding us is not only represented
as a visual space but also as a motor space: that is, a space
for actions within and beyond reaching distance (Rizzolatti
et al., 1997; Previc, 1998). Nevertheless, we do not wish to
suggest that the dorsal and ventral streams have exclusive
responsibility for near and far space, respectively. Rather, it
is likely that both visual streams simultaneously encode
stimuli in all distance conditions, albeit with a greater
emphasis upon either the dorsal or the ventral stream
according to the distance of the objects of interest. The
factorial design and statistical contrasts in our experiment
allow us to measure the relative additional demands placed
on one or the other visual stream during each task, not their
absolute degree of participation.
Whether comparable dissociations can be found between
performance in near and far space when the task demands
perceptual and attentional functions but no explicit motor
(manual) response remains to be investigated. In a small
group of patients with visuospatial neglect, Pizzamiglio and
colleagues failed to find evidence of dissociations between
performance in near and far space on tests of purely perceptual
judgement (Pizzamiglio et al., 1989). By contrast, the
previously mentioned patient of Vuilleumier and colleagues
did show slight unilateral neglect on non-manual tasks in far,
but not near, space (Vuilleumier et al., 1998). Yet the
dissociation was far more pronounced when, as in line
bisection, a manual response was required. That the neural
representation of space seems to differ with respect to distance
from the body may thus ‘be related to the independent
perceptual and motor systems that mediate responses to
external stimuli’ (Vuilleumier et al., 1998). The clinical
implication of these results is that patients with a suspected
spatial disorder, such as visuospatial neglect or optic ataxia,
should be examined for performance in near and far space
and, where possible, with both manual and non-manual
responses (e.g. exploratory oculomotor behaviour).
Acknowledgements
The authors wish to thank colleagues at the Institute of
Medicine and A. Kuhlen for their help with data acquisition.
2540
P. H. Weiss et al.
J.C.M. and P.W.H. are supported by the Medical Research
Council; K.Z. (SFB 194, A6) and G.R.F. (SFB 194, A16)
are supported by the Deutsche Forschungsgemeinschaft.
References
Andersen RA, Snyder LH, Bradley DC, Xing J. Multimodal
representation of space in the posterior parietal cortex and its use
in planning movements. [Review]. Annu Rev Neurosci 1997; 20:
300–30.
Anderson B. A mathematical model of line bisection behaviour in
neglect. Brain 1996; 119: 841–50.
Barrett AM, Crosson B, Crucian GP, Heilman KM. Horizontal line
bisections in upper and lower body space [abstract]. Soc Neurosci
Abstr 1998; 24: 1676.
Contrast polarity and face recognition in the human fusiform gyrus.
Nat Neurosci 1999; 2: 574–80.
Geyer S, Matelli M, Luppino G, Zilles K. Functional neuroanatomy
of the primate isocortical motor system. [Review]. Anat Embryol
(Berl). In press 2000.
Graziano MS, Gross CG. Spatial maps for the control of movement.
[Review]. Curr Opin Neurobiol 1998; 8: 195–201.
Graziano MS, Yap GS, Gross CG. Coding of visual space by
premotor neurons. Science 1994; 266: 1054–7.
Halligan PW, Marshall JC. Is neglect (only) lateral? A quadrant
analysis of line cancellation. J Clin Exp Neuropsychol 1989; 11:
793–8.
Halligan PW, Marshall JC. Left neglect for near but not far space
in man. Nature 1991; 350: 498–500.
Berti A, Frassinetti F. When far becomes near: remapping of space
by tool use. J Cogn Neurosci 2000; 12: 415–20.
Heilman KM, Bowers D, Shelton P. Attention to near and far space:
the third dichotomy. Behav Brain Sci 1990; 13: 552–3.
Brain WR. Visual disorientation with special reference to lesions
of the right cerebral hemisphere. Brain 1941; 64: 244–72.
Iriki A, Tanaka M, Iwamura Y. Coding of modified body schema
during tool use by macaque postcentral neurones. Neuroreport 1996;
7: 2325–30.
Corbetta M, Miezin FM, Dobmeyer S, Shulman GL, Petersen SE.
Attentional modulation of neural processing of shape, color, and
velocity in humans. Science 1990; 248: 1556–9.
Cowey A, Small M, Ellis S. Left visuo-spatial neglect can be worse
in far than in near space. Neuropsychologia 1994; 32: 1059–66.
Cowey A, Small M, Ellis S. No abrupt change in visual hemineglect
from near to far space. Neuropsychologia 1999; 37: 1–6.
Kanwisher N, Woods RP, Iacoboni M, Mazziotta JC. A locus in
human extrastriate cortex for visual shape analysis. J Cogn Neurosci
1997; 9: 133–42.
Kastner S, De Weerd P, Desimone R, Ungerleider LG. Mechanisms
of directed attention in the human extrastriate cortex as revealed by
functional MRI. Science 1998; 282: 108–11.
Desimone R. Visual attention mediated by biased competition in
extrastriate visual cortex. Philos Trans R Soc Lond B Biol Sci 1998;
353: 1245–55.
Kinsbourne M. Orientational bias model of unilateral neglect:
evidence from attentional gradients within hemispace. In: Robertson
IH, Marshall JC, editors. Unilateral neglect: clinical and experimental
findings. Hove (UK): Lawrence Erlbaum, 1993. p. 63–86.
Dolan RJ, Fink GR, Rolls E, Booth M, Holmes A, Frackowiak RS,
et al. How the brain learns to see objects and faces in an impoverished
context. Nature 1997; 389: 596–9.
Leinonen L, Hyvarinen J, Nyman G, Linnankoski I. I. Functional
properties of neurons in lateral part of associative area 7 in awake
monkeys. Exp Brain Res 1979; 34: 299–320.
Epstein R, Kanwisher N. A cortical representation of the local
visual environment. Nature 1998; 392: 598–601.
Mennemeier M, Wertman E, Heilman KM. Neglect of near
peripersonal space: evidence for multidirectional attentional systems
in humans. Brain 1992; 115: 37–50.
Fink GR, Halligan PW, Marshall JC, Frith CD, Frackowiak RS,
Dolan RJ. Neural mechanisms involved in the processing of global
and local aspects of hierarchically organized visual stimuli. Brain
1997; 120: 1779–91.
Fink GR, Marshall JC, Shah NJ, Weiss PH, Halligan PW, GrosseRuyken M, et al. Line bisection judgments implicate right parietal
cortex and cerebellum as assessed by fMRI. Neurology 2000; 54:
1324–31.
Friston KJ, Ashburner J, Frith CD, Poline J-B, Heather JD,
Frackowiak RSJ. Spatial registration and normalization of images.
Hum Brain Mapp 1995a; 3: 165–89.
Friston KJ, Holmes AP, Worsley KJ, Poline J-B, Frith CD,
Frackowiak RSJ. Statistical parametric maps in functional imaging:
a general linear approach. Hum Brain Mapp 1995b; 2: 189–210.
Geldmacher DS, Heilman KM. Visual field influence on radial line
bisection. Brain Cogn 1994; 26: 65–72.
George N, Dolan RJ, Fink GR, Baylis GC, Russell C, Driver J.
Moran J, Desimone R. Selective attention gates visual processing
in the extrastriate cortex. Science 1985; 229: 782–4.
Morris R, Mickel S, Brooks M, Swavely S, Heilman K. Recovery
from neglect. J Clin Exp Neuropsychol 1985; 7: 609.
Mountcastle VB, Lynch JC, Georgopoulos A, Sakata H, Acuna C.
Posterior parietal association cortex of the monkey: command
functions for operations within extrapersonal space. J Neurophysiol
1975; 38: 871–908.
Paterson A, Zangwill OL. Disorders of visual space perception
associated with lesions of the right cerebral hemisphere. Brain 1944;
67: 331–58.
Paus T. Location and function of the human frontal eye-field: a
selective review. [Review]. Neuropsychologia 1996; 34: 475–83.
Pizzamiglio L, Cappa S, Vallar G, Zoccolotti P, Bottini G, Ciurli P,
et al. Visual neglect for far and near extra-personal space in humans.
Cortex 1989; 25: 471–7.
Action in near versus far space
2541
Previc FH. Functional specialization in the lower and upper visual
fields in humans: its ecological origins and neurophysiological
implications. Behav Brain Sci 1990; 13: 519–75.
Shipp S, Watson JD, Frackowiak RS, Zeki S. Retinotopic maps in
human prestriate visual cortex: the demarcation of areas V2 and
V3. Neuroimage 1995; 2: 125–32.
Previc FH. The neuropsychology of 3-D space. [Review]. Psychol
Bull 1998; 124: 123–64.
Talairach J, Tournoux P. Co-planar stereotaxic atlas of the human
brain. Stuttgart: Thieme; 1988.
Rizzolatti G, Matelli M, Pavesi G. Deficits in attention and movement
following the removal of postarcuate (area 6) and prearcuate (area
8) cortex in macaque monkeys. Brain 1983; 106: 655–73.
Vuilleumier P, Valenza N, Mayer E, Reverdin A, Landis T. Near
and far visual space in unilateral neglect. Ann Neurol 1998; 43:
406–10.
Rizzolatti G, Fadiga L, Fogassi L, Gallese V. The space around us.
Science 1997; 277: 190–1.
Shelton PA, Bowers D, Heilman KM. Peripersonal and vertical
neglect. Brain 1990; 113: 191–205.
Received May 19, 2000. Revised July 7, 2000.
Accepted July 20, 2000