Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

download Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

of 17

Transcript of Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    1/17

    Right hemisphere control of visuospatial attention:line-bisection judgments evaluated with high-density electrical

    mapping and source analysis

    John J. Foxe,a,b,c,* Mark E. McCourt,d and Daniel C. Javitta,e

    a The Cognitive Neurophysiology Laboratory, Nathan S. Kline Institute for Psychiatric Research, Program in Cognitive Neuroscience and Schizophrenia,

    140 Old Orangeburg Road, Orangeburg, NY 10962, USAb Department of Neuroscience, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA

    cDepartment of Psychiatry and Behavioural Science, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USAd Department of Psychology, North Dakota State University, Fargo, ND 58105, USA

    e

    Department of Psychiatry, New York University School of Medicine, 550 1st Avenue, New York, NY 10016, USA

    Received 20 December 2001; revised 22 November 2002; accepted 18 December 2002

    Abstract

    The line-bisection task has proven an especially useful clinical tool for assessment of spatial neglect syndrome in neurological patients.

    Here, we investigated the neural processes involved in performing this task by recording high-density event-related potentials from 128 scalp

    electrodes in normal observers. We characterized a robust net negative potential from 170 400 ms poststimulus presentation that correlates

    with line-bisection judgments. Topographic mapping shows three distinct phases to this negativity. The first phase (170190 ms) has a

    scalp distribution exclusively over the right parieto-occipital and lateral occipital scalp, consistent with generators in the region of the right

    temporo-parietal junction and right lateral occipital cortices. The second phase (190240 ms) sees the emergence of a second negative

    focus over the right central parietal scalp, consistent with subsequent involvement of right superior parietal cortices. In the third phase

    (240400 ms), the topography becomes dominated by this right central parietal negativity. Inverse source modeling confirmed that right

    hemisphere lateral occipital, inferior parietal, and superior parietal regions were the likeliest generators of the bulk of the activity associated

    with this effect. The line stimuli were also presented at three contrast levels (3, 25, and 100%) in order to manipulate both the latency of

    stimulus processing and the relative contributions from magnocellular and parvocellular inputs. Through this manipulation, we show that

    the line-bisection effect systematically tracks/follows the latency of the N1 component, which is considered a temporal marker for object

    processing in the ventral visual stream. This pattern of effects suggests that this task invokes an allocentric (object-based) form of

    visuospatial attention. Further, at 3% contrast, the line-bisection effect was equivalent to the effects seen at higher contrast levels, suggesting

    that parvocellular inputs are not necessary for successful performance of this task.

    2003 Elsevier Science (USA). All rights reserved.

    Introduction

    Line-bisection tasks are a commonly used metric for the

    clinical assessment of visuospatial neglect syndrome, a condi-tion that predominantly results from vascular lesions to the

    right inferior parietal or temporoparietal cortex (e.g., Vallar

    and Perani, 1987; Cappa et al., 1991; Mesulam, 2000; Na et al.,

    2000; Kerkhoff, 2001;seeKarnath, 2001for a comprehensive

    treatment). Patients with visuospatial neglect syndrome gener-

    ally bisect lines significantly to the right of veridical center

    (Robertson and Halligan, 1999), due either to a decrement inthe ability to allocate attention to left visual space or to hyper-

    attention to rightward space. Clinical findings from neglect

    patients allied to the scores of neuroimaging studies showing a

    predomination of right parietal activity during tasks requiring

    visuospatial attention (e.g., Corbetta et al., 2000; Coull et al.,

    2001) have led to the formulation that while both left and right

    parietal areas are involved in attention to the right visual field,

    right parietal areas alone may control attention to the left visual

    Supplementary data associated with this article can be found at

    doi: 10.1016/S1053-8119(03)00057-0.

    * Corresponding author. Fax: 1-845-398-6545.

    E-mail address:[email protected] (J.J. Foxe).

    NeuroImage 19 (2003) 710 726 www.elsevier.com/locate/ynimg

    1053-8119/03/$ see front matter 2003 Elsevier Science (USA). All rights reserved.

    doi:10.1016/S1053-8119(03)00057-0

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    2/17

    hemifield (e.g., Heilman and van den Abell, 1980; Weintraub

    and Mesulam, 1987). A measure of support for this conjecture

    is derived from the finding that neurologically normal subjects

    demonstrate a phenomenon known as pseudoneglect (Bow-

    ers and Heilman, 1980), in which a significant and systematic

    misbisection of lines (or space) occurs that is leftward of

    veridical center. This leftward tendency in normal observershas been theorized to result from a profound parietal asymme-

    try in attentional control, which gives rise to some degree of

    hyper-attentiveness to the left visual field (in space-based at-

    tention), or to the left-hand side of individual objects (in object-

    based attention) (see Jewell and McCourt, 2000; McCourt,

    2001; McCourt and Jewell, 1999; McCourt et al., 2000).

    Functional imaging (fMRI) studies have confirmed a

    central role for right parietal cortices in performance of both

    line-bisection tasks (Weiss et al., 2000; Fink et al., 2000a,

    2000b, 2001, 2002; Galati et al., 2000) and judgments of

    object location relative to the midsagittal plane (Vallar et

    al., 1999; Galati et al., 2000). There have, to date, been no

    electrophysiological investigations of line bisection; conse-

    quently the timing of right parietal involvement relative to

    the timing of ongoing stimulus processing is as yet un-

    known. In the current study, we performed high-density

    electrical mapping (from 128 scalp electrodes) of the visual

    event-related potential (ERP) while subjects engaged in

    either a tachistoscopic forced-choice line-bisection task

    (McCourt and Olafson, 1997) or a control task in which they

    simply judged whether or not lines were transected. Our

    main objectives were to both confirm the involvement of

    and define the timecourse of right parietal activity in per-

    formance of the line-bisection task. In particular, we wished

    to assess the relationship of parietal attentional processes inthe dorsal visual processing stream to the processing of the

    line stimuli by object recognition areas of the ventral visual

    stream. An open question is whether line midpoint judg-

    ments can only be made upon completion of object recog-

    nition processes for the object that is to be bisected. In the

    present study, we systematically varied the timing of object-

    recognition processes by varying the contrast level at which

    lines were displayed. We used the well-characterized N1

    component of the visual evoked potential (VEP) as an index

    of the timing of object recognition processes, since this

    component has repeatedly been shown to be correlated with

    the processes involved in object recognition (e.g., Allison etal., 1999; Bentin et al., 1999; Doniger et al., 2000, 2001).

    Systematically varying the latency of this component al-

    lowed us to assess whether bisection processes tracked this

    latency manipulation.

    Materials and methods

    Subjects

    Nine (4 male) neurologically normal, paid volunteers

    (ages 1945 years, mean 29.2 years) participated. All

    subjects provided written informed consent, and the Insti-

    tutional Review Boards of the Nathan Kline Research In-

    stitute and North Dakota State University approved all pro-

    cedures. All subjects possessed normal or corrected-to-

    normal vision and were right-handed, as measured using the

    Oldfield (1971) laterality inventory (mean score 58.3, SD

    22.7).

    Instrumentation and stimuli

    Subject responses were sensed and collected, and stimuli

    were presented as 640 480 pixel images on a 21-in

    flat-screen monitor; frame refresh rate was 60 Hz and mean

    display luminance was 22 cd/m2. The generation and se-

    quencing of stimuli and the collection of subject responses

    were accomplished using the ERTS software package

    (Beringer, 1995).Luminance and contrast calibrations were

    made using a photometer.

    Stimuli were line segments (29 cm wide by 0.5 cm high)presented for 150 ms. At the viewing distance of 108 cm,

    lines subtended 15.3 (width) by 0.27 (height) of visual

    angle. All lines were horizontally and vertically centered

    within the display. Seventy-five percent of the lines were

    transected at 1 of 25 different locations ranging from 0.6

    of visual angle relative to veridical line midpoint. This range

    of transector locations is sufficient to determine perceived

    line midpoint in most observers (McCourt, 2001). The re-

    maining 25% of lines were nontransected. Lines were pre-

    sented at three levels of Michelson contrast: 3, 25, and

    100%. This range of stimulus contrast was employed for

    both methodological and theoretical reasons. Methodologi-cally, the presence of low contrast lines ensured that the

    visual discriminations in the feature detection task

    (transected vs nontransected) would not be trivially easy,

    and that subjects would remain alert and vigilant throughout

    the blocks of trials. Theoretically, it was of interest to

    determine whether there would be any unique behavioral or

    electrophysiological signature regarding the lowest contrast

    stimulus that, at 3% contrast, is well below the operating

    range of the parvocellular division of the geniculostriate

    projection, and thus preferentially engages the magnocellu-

    lar stream(Shapley and Perry, 1986).

    Fig. 1illustrates examples of the line stimuli. Both mem-bers of the upper pair of lines (Figs. 1A, and B) possess

    100% contrast, and are transected to the left of center, by

    0.33 (2.17%) and 0.60 (3.91%), respectively.

    Both members of the lower pair of lines (Figs. 1E and F)

    possess 3% contrast and are transected to the right of center,

    by 0.47 (3.04%) and 0.17 (1.09%), respectively. Lines

    C and D possess 25% contrast; C is veridically transected

    and line D lacks a transector. Pairs (Figs. 1A and B) and

    (Figs. 1E and F) differ in contrast polarity. Lines of varying

    contrast and contrast polarity appeared with equal fre-

    quency. The order of appearance of all line types was

    quasi-randomized within blocks of trials.

    711J.J. Foxe et al. / NeuroImage 19 (2003) 710726

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    3/17

    Procedures and tasks

    Subjects were seated in a comfortable chair with their

    midsagittal plane aligned with the display monitor. Stimuli

    were viewed binocularly through natural pupils.

    Each block of trials consisted of 200 line presentations.

    On alternate blocks subjects performed one of two tasks. In

    the Control Task subjects performed an oddball (feature

    discrimination) task in which they judged whether lines

    were transected (75% of trials) or not (25% of trials). Thenonbisected lines thus served as target stimuli. Use of the

    oddball task as the control condition ensured that subjects

    were actively engaged in an attentionally demanding task in

    both conditions and that equivalent button push responses

    were made during both tasks. If a simple passive viewing

    task were to be used, changes in general arousal level might

    have accounted for any effects seen.

    In the Line-Bisection task subjects made judgments re-

    garding transector location (left vs right) relative to per-

    ceived line midpoint by depressing the appropriate mouse

    button left or right, respectively. Subjects were in-

    structed to respond to nontransected lines by depressing

    either mouse button.

    Subjects used their dominant (right) hand to depress

    mouse buttons. Subjects were instructed to delay their re-

    sponses for at least 1 s following stimulus presentation to

    obviate motor artifacts in the EEG signal. Intertrial intervals

    were approximately 2 s, since subsequent trials were initi-

    ated 750 ms following the previous response. Each subject

    completed three or four blocks of both Line-Bisection and

    Control Task trials. At each level of line contrast subjectsmade either six or eight leftright judgments at each

    transector location, such that estimates of perceived line

    midpoint in each line contrast condition were determined

    based on 150200 bisection trials. Each block of trials was

    completed in approximately 9 min.

    Measurements and analyses

    Behavioral measures

    For behavioral analysis the dependent measure was the

    proportion of trials on which subjects indicated that transec-

    tors appeared to the left of perceived line midpoint. The

    Fig. 1. Line stimuli used in the experiment. Both members of the upper pair of lines (A and B) possess 100% contrast, and are transected to the left of center,

    by 0.33 (2.17%) and 0.60 (3.91%), respectively. Both members of the lower pair of lines (E and F) possess 3% contrast and are transected to the

    right of center, by 0.47 (3.04%) and 0.17 (1.09%), respectively. Lines C and D possess 25% contrast; C is veridically transected and line D lacks a

    transector. Pairs (A and B) and (E and F) differ in contrast polarity. Lines of varying contrast and contrast polarity appeared with equal frequency. The order

    of appearance of all line types was quasirandomized within blocks of trials.

    712 J.J. Foxe et al. / NeuroImage 19 (2003) 710726

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    4/17

    method of constant stimuli was used to derive psychometric

    functions, and nonlinear least-squares regression was used

    to fit a cumulative Gaussian distribution to the psychometric

    functions (see McCourt and Jewell, 1999). Based on these

    fits, transector locations corresponding to a 50% left re-

    sponse rate were determined. The transector location for

    which left and right responses occur with equal fre-quency is called the point of subjective equality (pse) and

    is an objective measure of perceived line midpoint. Infer-

    ential statistical analyses were performed on the pse values.

    Electrophysiological measures

    Continuous EEG was acquired from 128 scalp electrodes

    (impedances 5k), referenced to nose, band-pass-filtered

    from 0.05 to 100 Hz, and digitized at 500 Hz. Data were

    epoched off-line from 100 prestimulus to 500 ms post-

    stimulus and baseline-corrected from 100 to 0 ms. There-

    after, trials with blinks and eye movements were rejected on

    the basis of horizontal and vertical electrooculogram. An

    artifact rejection criterion of 60V was used at all other

    scalp sites to reject trials with excessive EMG or other noise

    transients. The average number of accepted trials per con-

    dition across subjects was 476 (SD 84.2). Responses to

    the nontransected lines, which served as target stimuli in

    the control task, were not included in the averages nor

    analyzed further. This latter point is important, as during the

    control task the N2/P3 component complex associated with

    infrequent target stimuli was generated for these stimuli

    (see, e.g., Ritter and Vaughan, 1969; He et al., 2001). Thus,

    use of these responses as constituents of the control av-

    eraged response would have been problematic. Also, false

    alarms (which were very few) were also excluded from thisanalysis.

    The latencies and scalp topographies of the standard ERP

    componentry over posterior scalp were identified from

    group-averaged waveforms collapsed across the Control

    and Line-Bisection conditions. Repeated measures analyses

    of variance (ANOVA) was used to test for significant dif-

    ferences between conditions for four preselected time win-

    dows, centered at the peak amplitude of the P1 (92100 ms),

    the N1 (174182 ms), the P2 (230240 ms), and the fol-

    lowing negative deflection over occipitotemporal regions,

    N280ot (270280 ms). For these time windows, a measure

    of integrated amplitude was derived between the responseand the 0-V baseline. ANOVAs possessed a repeated-

    measures 2 2 4 design with the following factors:

    condition (control, line-bisection), hemisphere (left, right),

    and electrode (4 electrode pairs over lateral occipitoparietal

    scalp).

    Topographic mapping

    Scalp topographic maps in the present study represent

    interpolated potential distributions, derived from the 128

    scalp measurements and based on the computation of a

    common average reference. These interpolated potential

    maps are displayed on the 3-D reconstruction of an average

    rendered scalp surface (derived from anatomical MRIs),

    using the boundary element method (BEM; e.g., Fuchs et

    al., 1998) and as implemented in the CURRY multimodal

    neuroimaging analysis software package (Version 4.0,

    Philips Research, Hamburg, Germany).

    Exact electrode locations were assessed for each subject

    on the day of testing by 3-D digitization of the locations ofthe scalp electrodes with respect to fiduciary landmarks (i.e.,

    the nasion and preauricular notches) using a magnetic digi-

    tization device (Polhemus Fastrak and 3DspaceDX soft-

    ware, Neuroscan, Inc.). Electrode placement was highly

    consistent across subjects due to the use of a custom-de-

    signed electrode cap that constrained interelectrode spacing

    and placement. An averaged version of these electrode lo-

    cations was projected onto the averaged rendered head for

    computation of the group topographic data.

    Last, one obvious constraint of the printed page is that

    only a limited number of discrete maps can be shown to

    represent a given topographic distribution, and such static

    maps fail to depict the full spatiotemporal dimensionality of

    the data. This can make it particularly difficult for the reader

    to assess the extent of contribution to the maps of back-

    ground noise. In determining the display gain to be used for

    the maps in the current study, we followed the topography

    over its entire timecourse (through observing animated time

    series). Observation of these maps in the baseline period

    (from 100 ms up to the onset of significant activity at

    about 50 ms) allowed us to determine the relative contribu-

    tion of noise to the topographic maps. The gain was then set

    so that background noise during this baseline period ac-

    counted for less than 12 topographic lines of potential in

    the maps.This is more readily seen in the animation appendix to

    this paper, which can be viewed or downloaded on Science-

    Direct.

    Dipole source analyses

    Information about the intracranial generators contribut-

    ing to effects seen in the data was also obtained through

    dipole source analysis using electromagnetic source estima-

    tion as implemented in CURRY (Version 4.0) software.

    This method assumes that there are a limited and distinct

    number of active brain regions over the evoked potential

    epoch, each of which can be approximated by an equivalentdipole. Dipole generators are placed within a three-shell

    spherical volume conductor model and overlaid on and

    adjusted to a BEM-segmented structural MRI (in this case,

    an averaged brain). The forward solution to this dipole

    configuration is tested against the observed experimental

    data. When not fixed, the positions and orientations of the

    dipoles are iteratively adjusted to minimize the residual

    variance between the forward solution and the observed

    data. The upper bound of the number of modeled dipole

    sources is determined using a test dipole (Scherg and Picton,

    1991). If the number of modeled sources, m, is adequate,

    then addition of another source (test dipole) and solving for

    713J.J. Foxe et al. / NeuroImage 19 (2003) 710726

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    5/17

    m 1 sources would not be expected to further reduce the

    residual variance, above that attributable to noise. The

    reader is referred to the following papers for comprehensive

    treatments of source localization procedures (Simpson et al.,

    1995a,b; Scherg and Berg, 1996; Michel et al., 2001).

    Results

    Behavioural results

    Fig. 2 presents group-averaged psychometric functions

    obtained in the Line-Bisection condition at each level of line

    contrast. Open symbols plot the mean percentage left

    responses ( 1 SEM) against transector location (in degrees

    relative to veridical line midpoint). The data are well be-

    haved, and the range of transector locations sampled is

    observed to encompass perceived line midpoint in this sam-

    ple of subjects. Solid lines depict cumulative Gaussian dis-tributions fitted to the data by nonlinear least-squares re-

    gression. The cumulative Gaussian function is described by

    the equation

    fx,,, 50 50 erf((x / 20.5 )) ,

    where xis transector location (in degrees relative to veridi-

    cal line midpoint), is an overall gain parameter, is the

    x-axis location corresponding to the mean of the underlying

    Gaussian density function (i.e., pse, the transector location

    at which leftright responses occur with equal frequency),

    andis its standard deviation. The error function (erf) is an

    approximation to the cumulative Gaussian distribution, forwhich there is no closed-form analytical expression. The

    horizontal dashed line in each panel indicates the 50% left

    response rate; the transector location for which the cumu-

    lative Gaussian intersects the dashed line is one measure of

    perceived line midpoint (pse).

    The solid symbols (and vertical dotted lines) in each

    panel plot mean pse (1 SEM) based on the analysis of

    psychometric functions from individual observers. The

    agreement with pse estimates from the group-averaged fits

    is excellent, and reveals the systematic leftward error

    (pseudoneglect) in perceived line midpoint that typifies the

    performance of neurologically normal right-handed observ-ers (Jewell and McCourt, 2000; McCourt, 2001). A one-way

    repeated-measures ANOVA conducted on the pse values in

    each line contrast condition revealed no significant effect of

    line contrast, F(2, 16) 0.74, P 0.49. While single-

    sample t tests comparing mean pse values in each line

    contrast condition against veridical fail to achieve signifi-

    cance (100%,t(8) 1.63,P 0.14; 25%,t(8) 1.73,

    P 0.12; 3%, t(8) 1.68, P 0.13), increasing statis-

    tical power by collapsing across levels of line contrast

    revealed, however, that grand mean pse is shifted signifi-

    cantly leftward of veridical by approximately 0.2, or 1.4%,

    t(26) 2.98, P 0.006.

    Electrophysiological results

    The analysis of behavioral results revealed that line con-

    trast had no significant effect on bisection performance, and

    we therefore collapsed across the three levels of line con-

    trast in the initial statistical analysis of the electrophysio-

    logical data. Inspection of group-averaged visual evoked

    potentials for both the Line-Bisection and Control Task

    conditions revealed the traditional series of ERP compo-

    Fig. 2. Behavioral data. Group-averaged psychometric functions obtained

    in the Line-Bisection condition at each level of line contrast. Open symbols

    plot the mean percentage left responses (1 SEM) against transector

    location (in degrees relative to veridical line midpoint). Solid lines depict

    cumulative Gaussian distributions fitted to the data by nonlinear least-

    squares regression. The horizontal dashed line in each panel indicates the

    50% left response rate; the transector location for which the cumulative

    Gaussian intersects the dashed line is one measure of perceived line

    midpoint (pse). The solid symbols (and vertical dotted lines) in each panel

    plot the mean pse (1 SEM) based on the analysis of psychometric

    functions from individual observers. The agreement with pse estimates

    from the group-averaged fits is excellent, and reveals the systematic left-ward error (pseudoneglect) in perceived line midpoint that typifies the

    performance of neurologically normal right-handed observers.

    714 J.J. Foxe et al. / NeuroImage 19 (2003) 710726

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    6/17

    nents, including P1, N1, and P2 (Fig. 3.). These componentswere maximal over visual cortices.

    The earliest robust component, the P1, showed no

    amplitude or latency difference between conditions.

    However, over right hemisphere occipitoparietal sites,

    responses in the Line-Bisection condition showed a

    markedly increased negativity relative to those in the

    Control Task; this difference began during the peak of the

    N1 component and then continued for some 300 ms. At

    left hemisphere electrode sites, this negative shift was

    seen to onset later during the N1/P2 transition. This

    negativity can be seen in Fig. 3, where the right hemi-

    sphere electrode site of maximal difference is plottedalong with the equivalently positioned left hemisphere

    site.

    As expected, the ANOVA for the P1 latency bin (92

    100 ms) yielded no significant main effect of condition (P

    0.62). The ANOVA for the N1 latency range yielded a

    significant interaction of ConditionXHemisphere (F(1,8)

    5.19, P 0.05). Follow-up planned comparisons

    showed that this was due to a robust main effect of

    condition over the right hemisphere electrode sites

    (F(1,8) 8.8, P 0.02), whereas no main effect of

    condition was seen over the equivalent left hemisphere

    sites in this latency range (P 0.44). The ANOVA for

    the P2 latency range revealed a significant main effect ofcondition (F(1,8) 31.33, P 0.001) but no interaction

    of Condition X Hemisphere (P 0.68). Similarly, the

    ANOVA for the N280ot latency range revealed only a

    significant main effect of condition (F(1,8) 29.27, P

    0.001).

    As the ANOVA for N1 showed that the effect of Con-

    dition onset earlier over the right hemisphere, we performed

    a post hoc analysis to determine rough onset times for the

    effect in both hemispheres. We used a series of paired

    two-tailed t tests between the Control and Line-Bisection

    conditions at the four representative pairs of electrode sites

    used in the above analyses. Tests were conducted at laten-cies preceding the P2 peak to mark the earliest timepoint

    that conformed to a 0.05 criterion. Onset latencies across the

    four left and across the four right hemisphere electrode sites

    were averaged to provide a best estimate of onset in a given

    hemisphere. A point was only accepted as the earliest di-

    vergence if at least 11 subsequent consecutive time points

    (20 ms at 500 Hz digitization rate) met the 0.05 criterion

    (see also Guthrie and Buchwald, 1991; Foxe and Simpson,

    2002). The criterion was met at 172 ms for the right hemi-

    sphere and 208 ms for the left, indicating that bisection-

    related attentional processes onset in the right hemisphere

    some 3040 ms earlier than they do in the left hemisphere.

    Fig. 3. Sustained negative modulation during line-bisection task. The 128-channel electrode montage is shown here projected onto a three-dimensional scalp

    reconstruction (from structural MRI data) of one of the subjects (B.H.). Group-averaged (n 9) ERP voltage waveforms are plotted for two electrode sites

    (darkened in Montage display), showing the marked negative deflection when subjects performed the line-bisection task (solid trace) relative to the control

    task (dashed trace). The bottom two panels show the P values (pointwise two-tailed t test comparisons between conditions), illustrating the timecourse of

    significant differences at these electrode sites.

    715J.J. Foxe et al. / NeuroImage 19 (2003) 710726

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    7/17

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    8/17

    (1) A dipole was fixed in the coordinates of RLOC and

    allowed to fit the data in the initial 10-ms period of

    the effect (172182 ms). This dipole accounted for

    57.9% of the variance in the data for this epoch. Note

    that allowing the dipole to freely fit the data in this

    period resulted in a dipole that went more medially

    and superiorly into the medial occipital cortex (7,

    87, 24). This freely fit dipole resulted in no real

    improvement in fit with essentially the identical per-

    centage of the variance explained (58.1%). The di-

    pole was therefore fixed in the RLOC and its orien-

    tation was also fixed (see Table 1).

    (2) A second dipole was fixed in the RIPP location. The

    fitting period was extended for another 10 ms (172

    192) and the dipole orientation was allowed to freely

    fit the data. This resulted in an explained variance of63.1% over the 20-ms epoch. Again, we allowed this

    dipole to freely fit as a test. This caused the dipole to

    move to a more inferior and medial location in the

    region of the right superior occipital gyrus (32, 84,

    29) with an improvement in explained variance to

    69.1%. This was a clear improvement in the propor-

    tion of explained variance, suggesting that there was

    significant activity in the region of the right superior

    occipital/occipitotemporal region that needed to be

    explained in this early period. We therefore fixed this

    second dipole in the freely fit location in the right

    superior occipital gyrus (RSOG).

    (3) We added a third dipole, again fixing its location in

    the RIPP, and allowed it to fit for orientation over the

    172 to 192-ms epoch. Explained variance improved

    to 73.8%. Allowing this dipole to freely fit for loca-

    tion and orientation resulted in only a very slight

    change in location (1 cm) and no improvement in

    explained variance. We therefore fixed the dipole in

    the RIPP cords defined above. We tested the stability

    of the fit so far by allowing dipole 2 to freely fit this

    epoch again. This resulted in no change in the di-

    poles position or orientation, indicative of a rela-

    tively stable fit.

    (4) We added a fourth freely fitting dipole and extendedthe analysis window out to 202 ms (a 30-ms epoch).

    This resulted in a second dipole in the region of the

    superior occipital gyrus/middle temporal gyrus (49,

    82, 22) that was more lateral than dipole 2 above.

    Explained variance for the 30-ms epoch was 71.0%.

    (5) The proximity of this dipole to dipole 2 suggested

    that these two dipoles might be trying to explain the

    same data. We tested this by turning off dipole 2 and

    allowing 4 to freely fit over this 30-ms epoch. This

    resulted in a dipole that moved slightly more medi-

    ally between the two previous dipole locations (45,

    85, 25). The explained variance was 71.5%.Clearly, only a single dipole in the RSOG was re-

    quired. Therefore, we fixed the location and orienta-

    tion of this latter dipole.

    (6) We then added a dipole in the RSPP and again

    extended the epoch by 10 ms to 212 ms. The reader

    will note that during this latter 10-ms period, the

    focus over superior parietal scalp has begun to

    emerge strongly, as detailed in the topographic anal-

    ysis above. Addition of this dipole resulted in an

    explained variance of 76.8% over the 40-ms epoch

    (172212). It is noteworthy that at the end of the

    epoch, explained variance was 83.5% (212 ms). Al-

    Fig. 4. Topographic mapping of the line-bisection effect. Group-averaged

    (n 9) potential maps are shown for three timepoints during the three

    phases of the line-bisection effect. These topographies are derived from the

    difference waves, obtained by subtracting the Control condition from the

    Line-Bisection condition. Surface negativity is depicted in blue/cyan and

    positivity in red/yellow. The first phase of the effect (top maps 176 ms)

    shows a highly lateralized negative focus over right temporoparietal scalp.

    During the second phase (middle maps 220 ms), at least two distinct

    negative foci are evident with the emergence of a strong central parietal

    negative focus. During the third phase, the time period of the largest effect(bottom maps 310 ms), the topography is dominated by a right central

    parietal negativity.

    717J.J. Foxe et al. / NeuroImage 19 (2003) 710726

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    9/17

    lowing this dipole to freely fit resulted in only a very

    slight change of location, indicating that its location

    in RSPP was a relatively stable fit, so the dipole was

    fixed in the coordinates defined above.

    (7) We next added both a medial striate and left cere-

    bellar dipole and extended the fitting window to 50

    ms (172222 ms), allowing the orientations of thesetwo dipoles to fit the data. These six dipoles resulted

    in an explained variance of 81.7% over the epoch

    with a best fit at the last point (222 ms) of 93.0%. We

    fixed these latter two dipoles orientations.

    (8) We then opened up the epoch across the duration of

    the entire effect (172400 ms) and allowed these six

    dipoles to fit the data. This resulted in an average

    explained variance of 92.0% over this 228-ms epoch

    (see Fig. 6B). It is of note that during the period

    when the effect was of largest amplitude, 220320

    ms (i.e., highest signal-to-noise in our data), the

    averaged explained variance was 97.5%.(9) Finally, it is clear from the topographic data that

    there is activity over the left hemisphere that is

    unlikely to be simply a result of volume conduction

    from right hemisphere activations. Left hemisphere

    activations are also seen in the fMRI data although

    the exact loci across studies are not consistent (ex-

    cept for the left cerebellum). We added a left hemi-

    sphere inferior posterior parietal dipole (38, 42,

    40) and fit it to a window from 200 to 260 ms. These

    coordinates were taken from Fink et al. (2001). It

    was during this epoch that the strongest activity over

    left scalp was observed electrophysiologically. We

    then fixed its orientation and opened the epoch up to

    the entire window (172400 ms) and refit the data

    with these seven dipoles. Explained variance across

    the entire epoch was only marginally improved to

    92.4%.

    It is noteworthy that Fink and colleagues also find con-

    sistent activation of right frontal regions including orbito-

    frontal cortex and the right dorsolateral prefrontal cortex. In

    our study, we also see foci of electrical activity over frontal

    regions but these activations occur relatively late in time,

    mostly after the posterior parietooccipital line-bisection ef-

    fect is over. These frontal effects are not treated further inthe present data analysis.

    It is important to emphasize that the fMRI activations

    seen in previous studies, and which we have used to guide

    this source analysis, cannot be thought of as discrete acti-

    vations of individual brain regions. Rather, as can be seen in

    Fig. 7, many of these activation clusters are much larger

    than a single cortical area and are likely to represent the

    activity of a cluster of functional areas. The same is cer-

    tainly the case for the electrophysiological results reported

    in the present paper, where a given scalp topography almost

    certainly represents coordinated activity within a cluster of

    functionally related areas. For instance the lateral occipital

    cortex contains a cluster of subregions, and it would be

    incorrect to think of the initial phase of the line-bisection

    effect as representing activity in just a single one of these.

    As such, fitting each cluster or time epoch with a single

    equivalent current dipole clearly represents a highly over-

    simplified model of activity within a given cluster of areas.

    As such, it is important that the reader should consider thesedipoles to represent a center-of gravity rather than a

    discrete neural locus or a single neural event.

    Discussion

    The current findings explicate the brain mechanisms un-

    derlying performance of the visual line-bisection task, a

    perceptual version of the perceptual-motor task that is fre-

    quently employed in the clinic to disclose the presence and

    severity of visuospatial neglect syndrome. We define an

    electrophysiological correlate of line-bisection judgments,

    which manifests as a right parieto-occipital negative poten-

    tial that is significantly earlier in latency and larger in

    magnitude than that over left scalp. This net negative po-

    tential change is seen to persist over a latency window from

    approximately 170 to 400 ms poststimulus, and consists of

    three distinct phases that can be discerned through topo-

    graphic mapping. Dominating the earliest phase of the re-

    sponse (165190 ms) is a negative focus concentrated

    over the right lateral parieto-occipital scalp. In this early

    phase, the effect appears to be largely if not entirely unilat-

    eral with no differences seen over the left hemisphere. In the

    second phase (190240 ms), a distinct additional negative

    focus develops over the more superior right central parietalscalp, during which the right parieto-occipital focus persists.

    In the final phase, as the point of maximal amplitude effect

    is approached (310 ms), the topographic pattern becomes

    dominated by this second focus over the right central pari-

    etal scalp. We refer to this net negative difference as the

    line-bisection effect.

    The intracranial generators of the line-bisection effect

    Recent fMRI investigations using very similar experi-

    mental paradigms to ours have detailed a network of brain

    areas that are activated during line-bisection judgments(see, e.g., Fink et al., 2000a, 2000b, 2001). Fig. 7, which is

    adapted from work by Fink and colleagues (Fink et al.,

    2001), illustrates the regional distribution of the activated

    regions during a line-bisection task, and clearly shows the

    involvement of right inferior and superior parietal cortices,

    in addition to earlier visual areas, frontal regions, and re-

    gions in the cerebellum. Our topographic results (Fig. 3)

    appear to be in good agreement with these findings and

    provide the critical temporal activation pattern across these

    regions. Although topographic mapping of the ERP alone

    permits only relatively crude spatial localization in terms of

    the intracranial generators responsible for a given scalp

    718 J.J. Foxe et al. / NeuroImage 19 (2003) 710726

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    10/17

    topography, observation of the correspondence between the

    fMRI results of Fink et al. (2001) and the present data

    allows for a degree of added confidence in such interpreta-

    tions. Clearly, the scalp topography of the earliest phase of

    the line-bisection effect suggests that regions in the right

    lateral occipital cortex and the right temporoparietal occip-

    ital junction (TPJ) are the likeliest generators, rather than

    parietal areas. Indeed, previous fMRI results show an ex-tended region of robust activation in these more inferior

    right occipital regions (e.g., Fink et al., 2001). In turn, the

    second phase of the line-bisection effect, during which a

    strong focus develops over the right central parietal scalp, is

    likely generated in regions of the right superior parietal

    cortex, where again, fMRI shows a very strong regional

    activation.

    To further assess the putative intracranial sources of the

    effect, we conducted a source analysis to augment our

    topographic mapping data. A stable fit that accounted for

    more than 90% of the variance contained in the data was

    found, which included generators in right lateral occipital

    cortex, right superior occipital gyrus, right lateral inferior

    posterior parietal cortex, and right lateral superior posterior

    parietal cortex. Additional generators were modeled in me-

    dial occipital cortex, the left cerebellum, and left lateral

    inferior parietal cortex. The latter dipole in left inferior

    parietal cortex contributed very little to the source solution.

    To arrive at this solution, we took advantage of the previousfMRI studies of line bisection to guide our source investi-

    gation (e.g., Fink et al., 2001). The model shows that di-

    poles seeded in the major activation centers previously

    defined provide a very stable fit of our data.

    It is noteworthy that the earliest phase of the line-bisec-

    tion effect we describe here has a somewhat more inferior

    distribution than might be predicted for a generator in the

    inferior parietal lobule and may be more consistent with a

    primary generator in superior occipitotemporal or lateral

    occipital regions of cortex. Source analysis suggested that

    two dipoles were needed to account for his early phase, one

    in right lateral occipital cortex and a second in right lateral

    superior occipital gyrus, directly adjacent to the middle

    temporal gyrus. Karnath (2001) has recently suggested that

    the rostral portion of the superior temporal cortex, just

    inferior to the temporoparietal junction, is the principal

    region involved in spatial awareness and the primary site of

    injury in cases of neglect (see also Karnath et al., 2001).

    Karnath argues that this region represents the interface be-

    tween the ventral (object: what) and dorsal (spatial: where)

    visual processing streams. Our data appear to accord well

    with this interpretation, particularly when the temporal se-

    quence of activations is taken into account. First, we find

    that there is extensive visual processing prior to the occur-

    rence of any spatial attention effects, as indexed by the N1component of the VEP which has been repeatedly impli-

    cated as a marker for ventral stream object processing (e.g.,

    Allison et al., 1999; Doniger et al., 2000, 2001; Murray et

    al., 2002). The onset of the initial phase of the line-bisection

    effect is seen to follow the peak of the N1 and appears to be

    generated in and around the right lateral occipital region and

    the right superior occipital gyrus, which is relatively close to

    the TPJ. As Karnath notes, the TPJ is ideally situated at the

    junction between the two visual processing streams. Subse-

    quently, the effect appears to be transmitted into inferior

    parietal and then more superior parietal areas of the dorsal

    stream, potentially for further spatial processing. A strongprediction based on the present results is that object recog-

    nition processes, as indexed by the N1 component, should

    be relatively intact in neglect patients but that the essential

    link or relay between initial object-based processing (ac-

    complished within the ventral stream) and subsequent

    visuospatial processing (subserved by the dorsal stream) is

    selectively impaired.

    The observation that lesions to the superior temporal

    cortices appear to be the primary culprits in spatial neglect

    syndrome (Karnath, 2001) raises questions regarding the

    functional role played by superior parietal regions during

    performance of the line-bisection task. Clearly superior pa-

    Fig. 5. Three levels of contrast. Group-averaged (n 9) voltage wave-

    forms are shown for the three levels of display contrast used (sameparietooccipital scalp site as shown in Fig. 3). The line-bisection effect can

    be seen at each level. As N1 latency increases with decreasing contrast, the

    line-bisection effect moves systematically with this N1 latency shift.

    719J.J. Foxe et al. / NeuroImage 19 (2003) 710726

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    11/17

    rietal areas are involved in the processing demanded for

    line-bisection judgments as represented by the late phase of

    the line-bisection effect over superior parietal scalp seen

    here, and based on results from the many hemodynamic

    imaging studies of line bisection (e.g., Fink et al., 2000a.,

    2001). One possible explanation might be that the initial

    phase of the line-bisection effect over TPJ represents a

    necessarybutinsufficientcomponent of the task whereas the

    subsequent processing in more superior parietal regions

    represents more refined processes critical to the accuracy of

    judgments but not necessaryfor the performance of the task

    per se. Such a sequence might help to explain the sometimes

    confusing and contradictory findings regarding the multiple

    areas of right parietal and/or temporal cortex which appear

    to result in neglect when lesioned (e.g., Vallar and Perani,

    1986; Vallar, 2001).

    Attempts to dissociate the relative contributions of inferior

    and superior parietal regions to the processes of visuospatial

    Fig. 6. Source analysis of the line-bisection effect. (A) The locations of the seven dipoles used to model the line-bisection effect are shown projected into

    a 3-D rendering of an averaged human brain. Each dipole location is color-coded. (B) The amount of variance explained in the data by these seven dipoles

    across the 172 to 400-ms epoch of the effect. (C) Waveforms depicting the source strength of each dipole are plotted for the 172 to 400-ms epoch (same color

    convention as used in A).

    720 J.J. Foxe et al. / NeuroImage 19 (2003) 710726

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    12/17

    attention have been made (Corbetta et al., 2000). Using fMRI,

    Corbetta and colleagues conducted a typical visuospatial cue-

    ing experiment, whereby centrally presented arrows cued

    subjects to the likeliest location for a subsequent target stim-

    ulus, which was then presented to either the left or the right of

    fixation. They found that the right TPJ was primarily activatedwhen visual attention was reoriented to the occurrence of a

    target stimulus and that this activation was greater to targets

    that occurred in the uncued region of space. Thus, they inter-

    preted this finding as evidence that TPJ was involved in the

    reorienting of spatial attention when attention was captured

    commonly known as exogenous attention in the literature.

    Importantly, the TPJ was not found to be active in response to

    the cue stimulus that oriented subjects attention to the subse-

    quent targets, arguing that TPJ was not involved in the volun-

    tary orienting of attention (so-called endogenous attention). On

    the other hand, they found that the intraparietal sulcus and

    superior parietal regions were involved in the voluntary direct-ing and maintenance of attention, showing sustained activation

    to the endogenous cues used to direct attention. This latter

    finding accords well with recent electrophysiological studies of

    endogenous attention mechanisms, which showed involvement

    of right parietal regions in the anticipatory biasing of visuo-

    spatial attention in somewhat similar cue-target designs (Foxe

    et al., 1998; Worden et al., 2000; Fu et al., 2001).

    However, on initial consideration, the present findings do

    not appear to accord well with the findings of Corbetta and

    colleagues. Here we find that the first and putatively most

    critical modulation of the ERP, due to the spatial attention

    required for the line-bisection task, appears to occur in the

    region of the right lateral occipital cortex and the right TPJ.

    Clearly, in this task, attention is not being captured by thestimulus. Rather, this task involves the voluntary/endoge-

    nous orienting of spatial attention to an object (the line).

    However, the apparent discrepancy between these results

    may have to do with the large differences in experimental

    paradigm. One clear difference between the current task and

    the kinds of tasks that are typically used to investigate

    visuospatial attention is that here spatial attention is being

    directed to an object, whereas in the Corbetta study, spatial

    attention was being directed to a portion of space. Thus,

    regions of TPJ may become involved when spatial attention

    is voluntarily oriented to an object. In support, Karnath

    (2001) argues that the rostral superior temporal gyrus, byvirtue of inputs from both the dorsal and the ventral stream,

    may be involved in both spatial orienting and the analysis of

    objects in space. By this reasoning, both the findings of

    Corbetta et al. (2000) and the present findings are consonant

    with activity in the TPJ. Of course, given the spatial uncer-

    tainty regarding the exact generator locus of the early line-

    bisection effect from ERP topographic mapping alone, an

    alternate explanation may be that subregions in and around

    Table 1

    Results from source analysis

    Brain Region Talairach Coordinates Dipole Orientation

    X Y Z nx ny nz

    R. Lat Occip 34 89 3 0.09 0.75 0.65

    R. Lat Sup Post Parietal 23 58 61 0.24 0.72 0.65

    R. Lat Inf Post Parietal 41 40 50 0.18 0.83 0.52

    R. Sup Occip Gyrus 44 84 25 0.13 0.87 0.48

    Med. Striate/Extrastriate 0 80 14 0.20 0.89 0.40

    L. Cerebellum 30 60 30 0.33 0.9 0.27

    L. Lat Inf Post Parietal 38 42 40 0.57 0.82 0.05

    Note. Talairach coordinates and orientations are given for each dipole.

    Fig. 7. fMRI activation during the line-bisection task. Comparison of the fMRI activations seen here with the scalp topography of the line-bisection effect

    shown in Fig. 4 shows strong correspondence between the electrophysiological and hemodynamic data sets. (adapted with permission from Fink et al.,

    NeuroImage 14 (2001) S59.) Areas of significant activation are shown projected into standard stereotactic space in a sagittal, coronal, and axial view (from

    left to right).

    721J.J. Foxe et al. / NeuroImage 19 (2003) 710726

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    13/17

    the temporoparietal junction serve different roles and that

    the region in TPJ that is active during exogenous attention

    in the Corbetta et al. (2000) study is not the same region that

    is activated in the present study.

    Magnocellular and parvocellular involvement in spatial

    attention

    Arising in the retina, and projecting centrally into tem-

    poral and parietal cortical regions, the primate visual system

    is composed of two principal neural pathways, the parvo-

    cellular (P) and magnocellular (M) processing streams

    (Shapley and Perry, 1986). The M stream has a predominant

    dorsal projection to areas V2, V3, V4, MT, MST, and 7a in

    posterior parietal cortex. The P stream courses ventrally to

    areas V2, VP, and V4, and to regions of inferotemporal

    cortex (DeYoe and Van Essen, 1988; Webster and Unger-

    leider, 1998). Cells in the M pathway possess high-contrast

    sensitivity, and demonstrate early response saturation; cells

    of the P pathway are far less sensitive, and respond in a

    graded fashion even to high-contrast stimuli (Kaplan and

    Shapley, 1982, 1986). Neuronal contrast response functions

    are well-described by the MichaelisMenten equation,

    where C50represents the stimulus contrast producing half-

    maximal response (the semisaturation constant). Primate P

    and M LGN neurons possess median values of C50of 0.50

    and 0.11, respectively; for cells in primary visual cortex

    (V1), which receive both M and P input, this value is 0.33,

    and for area MT it is 0.07 (Sclar et al., 1990).

    It is currently unknown to what extent visuospatial at-

    tentional mechanisms depend differentially upon M and P

    pathway input. Relevant is evidence that VEPs to luminancecontrast in the neglected hemifield of neglect patients are

    delayed compared to the nonneglected hemifield. The delay

    is exacerbated at lower luminance contrasts, and no latency

    differences are observed for isoluminant chromatic stimuli

    (Spinelli et al., 1994, 1996). These findings have been

    interpreted to suggest that neglect may, at least in part,

    result from a selective impairment of M stream input to

    parietal cortical regions. Anatomical evidence also indicates

    that the parietal cortex receives rich input from the M

    stream, supported by recent functional neuroimaging evi-

    dence in humans (Tootell et al., 1995a, 1995b) that neurons

    in dorsal V3 are activated at low stimulus contrasts. AreaV3, in turn, receives M stream input from layer 4B of V1

    and from the thick stripes of V2, and has reciprocal

    connections with other regions innervated by the M stream,

    such as parietal cortex and V5 (Shipp et al., 1994; Zeki and

    Shipp, 1988; Webster and Ungerleider, 1998).

    The present behavioral results, as well as those of a more

    extensive study involving 59 observers (McCourt, unpublished

    results), reveal that leftward error (pseudoneglect) on the line-

    bisection task is remarkably stable as a function of stimulus

    contrast over the range of 1.5100%, which readily encom-

    passes the transition from M to P stream function. It is clear

    from these data that parvocellular inputs do not appear to be a

    necessary component for successful performance of the line-

    bisection task, as at contrast levels below about 8% there is

    little or no parvocellular input (Tootell et al., 1988). The cur-

    rent electrophysiological data provide similar evidence against

    a major role for parvocellular inputs as at the 3% contrast level,

    a robust line-bisection effect is observed and this effect has

    highly similar morphology to the effects seen for the 25 and100% contrast conditions. The line-bisection effect at 3% con-

    trast persists despite large effects upon P1 amplitude and N1

    latency due to the contrast manipulation. Rather, the primary

    effect of the contrast manipulation upon the line-bisection

    effect was simply to delay its onset, which will be discussed in

    the next section.

    Confluence of object-based and space-based attentional

    processing

    While the most prominent omnibus symptom of neglect

    syndrome is simply left inattention, the existence of numer-

    ous subtypes of neglect has become increasingly clear, dis-

    sociated along the dimensions of near (peripersonal) vs far

    (extrapersonal) space (Halligan and Marshall, 1991; Vuil-

    leumier et al., 1998; Cowey et al., 1994; Tegner and

    Levander, 1991b), perceptual vs motor origin (Bisiach et al.,

    1990; Tegner and Levander, 1991a), and referenced to ego-

    centric (self-centered) vs allocentric (object-based) spatial

    coordinate systems (Behrmann, 1999; Driver and Pouget,

    2000; Walker, 1995). Similar distinctions have been made

    with respect to pseudoneglect, where the degree of leftward

    error on line-bisection tasks is influenced by motor activity

    (McCourt et al., 2001), viewing distance (McCourt and

    Garlinghouse, 2000a), and object geometry (McCourt andJewell, 1999; McCourt and Garlinghouse, 2000b). Our re-

    sults are particularly interesting with respect to the distinc-

    tion between egocentric and allocentric attention. Note that

    spatial judgements of line midpoint in a bisection task

    necessarily involve computing the spatial location of a spe-

    cific feature (the transector) with respect to features of an

    object (the line itself), and thus explicitly invoke object-

    based attention. Our electrophysiological results speak to

    this issue.

    Perhaps the best-known effect of selective visuospatial

    attention upon the ERP is the oft-reported modulation of the

    P1 component, which is observed when attention is directedto a specific portion of the visual field while other parts of

    visual space are ignored (e.g., Van Hoorhis and Hillyard,

    1977; Martinez et al., 1999, 2001). This spatial attention

    effect occurs relatively early in visual processing (typically

    in the 70- to 130-ms latency range) and is an example of the

    case where attention is being apportioned in an egocentric

    spatial coordinate system. Clearly, the present findings are

    wholly different in that the latency of the line-bisection

    effect is late (post N1 at 170 ms). Moreover, its latency

    tracks with that of N1 (not P1), as a function of decreasing

    line contrast. The timecourse of N1 has been shown to be a

    marker for ventral stream processing (e.g., Allison et al.,

    722 J.J. Foxe et al. / NeuroImage 19 (2003) 710726

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    14/17

    1999; Doniger et al., 2001), and as such, the relationship of

    N1 peak latency and the onset of the line-bisection effect

    suggests that the effect is linked to the timing of object

    processing in the visual system. Object-based attention must

    have a substrate upon which to act, that is, an object, and the

    synthesis of coherent objects from visual primitives such as

    blobs, lines, and edges is associated with processing in theventral pathway (e.g., Gross et al., 1972; Sary et al., 1993;

    Janssen et al., 2001). Attention that is referenced to partic-

    ular objects must therefore receive input from the ventral

    stream and this is what is meant by allocentric spatial

    attention. Thus, in the present study, the line itself (espe-

    cially its endpoints) must first be encoded; spatial attention

    is only subsequently engaged upon the object to determine

    the location of the transector in relation to the objects

    midpoint. Compellingly, several recent studies have re-

    vealed that, in the absence of an object upon which to

    operate, the severity of both neglect (Vuilleumier and Lan-

    dis, 1998) and pseudoneglect are reduced (Post et al., 2001).

    Thus, it would appear that neglect might be more severe

    during allocentric spatial attention. These findings are con-

    sistent with the notion of the superior temporal gyrus as an

    interface between object processing and spatial processing

    streams, as posited by Karnath (2001).

    While our contrast manipulation succeeded in its goal of

    systematically lengthening peak N1 latency, especially at

    the lowest contrast level used (3%), it is worth emphasizing

    that the amplitude of N1 appears to behave quite differently

    from that of P1 in response to this manipulation. The P1 gets

    consistently smaller as the contrast is dropped, especially

    between the 25 and 3% levels, whereas the N1 maintains its

    amplitude across the three contrast levels. It is perhapssurprising that the N1 shows such preserved amplitude at

    the 3% contrast level given that N1 has been often taken to

    predominantly represent processing within ventral visual

    stream structures (e.g., Doniger et al., 2000, 2001; Murray

    et al., 2002). Clearly, these results show that while the

    timing of N1 is affected, the amplitude of N1 is largely

    unaffected when stimulation is biased toward magnocellular

    input. The relative independence of P1 and N1 amplitudes

    suggests that N1 generation does not depend causally on

    inputs from the generators of the earlier P1 component and

    that the processes represented by the N1 component can be

    successfully initiated through magnocellular inputs. Furtherinvestigation is clearly necessary to delineate the relative

    contributions of magnocellular and parvocellular inputs to

    the componentry of the VEP (see also Foxe et al., 2001;

    Butler et al., 2001).

    Relation of ERP timecourse to TMS-derived estimates

    While fMRI studies have provided strong evidence that

    activation of right parietal regions accompanies tasks re-

    quiring spatial attention (e.g., line-bisection tasks), and

    whereas the present study using high-density topographical

    ERP methods has verified the involvement of right parietal

    and temporoparietal occipital regions, including the tempo-

    ral dynamics of their activation, neither fMRI nor ERP

    techniques actually establish a causal relationship between

    neural activity and behavioral performance. In this regard a

    useful adjunct technique is transcranial magnetic stimula-

    tion (TMS), in which briefly generated magnetic fields tran-

    siently disrupt neural activity in target cortical regions. Thedeficits in behavior produced by TMS can be causally re-

    lated to the disruption of normal activity within the net-

    works of stimulated brain regions. Strong evidence that

    right parietal regions are essential for spatial attention

    comes from the finding that rapid-rate TMS over these

    regions attenuates the allocation of attention into left hemi-

    space, producing visual extinction similar to that in neglect

    patients (Pascual-Leone et al., 1994; Fierro et al., 2000).

    More recently a study using single-pulse TMS over right

    posterior parietal cortex (electrode site P6) reported signif-

    icant rightward deviation of line midpoint judgments in a

    line-bisection task (i.e., neglect-like behavior) when pulses

    were delivered 150 ms poststimulus (Fierro et al., 2001).

    This value agrees very well with the observed timecourse of

    initial right parietal activation as revealed in the present

    ERP study, and the optimal stimulation location over site P6

    is highly similar to the observed center of the initial nega-

    tive focus of the line-bisection effect.

    One potentially informative direction for a future TMS

    study would be to selectively interrupt the early (parietooc-

    cipital) and later (superior parietal) phases of the line-bisec-

    tion effect to attempt to dissociate and identify the separate

    contributions of these two distinct regions to line-bisection

    judgments. Mapping both the exact timing and the topog-

    raphy of these information processing phases on a subject-by-subject basis could further enhance the specificity of this

    method, as some degree of temporal and topographic vari-

    ation was observed between subjects in the current sample.

    Clinical utility of ERP measures of the line-bisection

    effect

    One problematic issue surrounding the use of standard

    line-bisection tasks to define spatial neglect syndrome is

    that motor responses are often required from the patient,

    such as asking them to draw bisectors with a pencil through

    lines on a page, or to point to the longer/shorter end of adisplayed line. Critically, there are often significant motor

    deficits in patients with lesions that involve frontal cortex

    and also visual cortices, and this can be a confounding

    factor in results of the line-bisection task that involve such

    overt responses (see, e.g., Darling et al., 2001). That is,

    systematic bisection errors by a motor-compromised patient

    could reflect an inability or reluctance to move the drawing

    or pointing hand in the necessary direction for correct judg-

    ments. The presence of a highly robust ERP effect over right

    occipitoparietal areas may provide a simple means for as-

    sessing integrity of function in patients without the neces-

    sity for overt motor or even verbal responses.

    723J.J. Foxe et al. / NeuroImage 19 (2003) 710726

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    15/17

    Summary

    In conclusion, the current study reveals a robust negative

    potential over the right lateral occipitoparietal and right

    central parietal scalp that indexes the neural processing

    involved in performance of the line-bisection task.

    Through high-density topographic mapping, we detail thespatiotemporal dynamics of this processing on the scalp

    surface. The observed topographies are consistent with re-

    cent hemodynamic imaging studies that have suggested a

    prominent role for regions in and around the right tem-

    poroparietal junction and regions of the right superior pari-

    etal cortices in this task. Source analysis confirmed the

    involvement of these areas as well as early involvement of

    right lateral occipital cortices. The present data provide the

    critical temporal activation pattern across these regions and

    underscore the importance of defining the temporal patterns

    of attentional modulation, which will be critical to our

    understanding of the mechanisms of attentional control in

    the human brain (e.g., Schroeder et al., 2001).

    Acknowledgments

    We express our sincere appreciation to Deirdre Foxe,

    Beth Higgins, and Dr. Micah Murray for their technical help

    with this study. We are most grateful to Dr. Gereon Fink

    and his colleagues for generously allowing us to reproduce

    their fMRI data in the current report. Dr. Antigona Martinez

    provided valuable comments on an earlier version of the

    manuscript for which we are indebted to her. Our thanksalso go to two anonymous reviewers for their careful and

    constructive comments. This work was supported in part by

    grants from the National Institute of Mental Health

    (MH63434 to J.J.F.; MH49334 to D.C.J.), the National Eye

    Institute (EY12267 to M.E.M.), North Dakota EPSCoR

    (M.E.M.), the Neuropsychiatric Research Institute, Fargo,

    ND (M.E.M.), and the North Dakota State University De-

    velopment Foundation (M.E.M.), and by generous support

    from the Burroughs Wellcome Fund.

    Appendix

    Readers are encouraged to visit the online version of this

    paper at ScienceDirect to download the animation file which

    accompanies the present report.

    This animation shows the evolution of the line-bisection

    effect at a rate approximately 100 times slower than real

    time. It begins with a display of the 128-channel montage as

    the 3-D rendered head rotates before settling with a view of

    the scalp from directly behind the head. The electrodes then

    disappear before the topographic maps begin. The counter at

    the bottom right corner of the panel shows time in ms and

    the animation of topography begins at 0 ms, which is when

    the stimulus appeared. The group-averaged data have been

    projected onto a 3-D-rendered scalp surface. Recall that the

    line-bisection effect is derived by subtracting the control

    condition from the experimental condition and therefore all

    activity associated with the common sensory processing of

    these stimuli has been eliminated from the data. Therefore,

    it can be seen that for the first 150 ms, there is no activitythat exceeds noise.

    The following is a description of the evolution of the

    effect, which may help to orient the reader as they observe

    the animation.

    At about 160 ms, a negative focus over right pari-

    etooccipital scalp is seen to emerge and build rapidly

    in amplitude. By 172 ms (the point at which the effect

    reaches statistical significance) the negativity is

    clearly discernable. This right parietooccipital nega-

    tivity continues to build in amplitude up until 200

    210 ms. At approximately 188 ms, the initial right parietooc-

    cipital negativity appears to extend dorsally and cen-

    trally over midparietal scalp and this extension of the

    negativity quickly develops into a second distinct fo-

    cus by approximately 206212 ms.

    By 210 ms, a third negative limb begins to be apparent

    over the left parietooccipital scalp, although this left

    scalp negativity never becomes fully distinct from the

    two right foci.

    At 224 ms, the right parietooccipital focus begins to

    decrease in amplitude and the topography becomes

    dominated by the central parietal negativity. This con-tinues until approximately 260 ms when the right

    parietooccipital focus begins to increase its amplitude

    for a second time.

    Between 260 and 340 ms, this pattern is essentially

    maintained. However, the reader will notice that both

    the right parietooccipital and central parietal foci ap-

    pear to wax and wane somewhat independently during

    this period. While it is impossible to determine at this

    stage what these dynamics represent functionally, we

    speculate that these modulations are likely to represent

    successive rounds of activation within functionally

    distinct modules of the spatial attention network, andperhaps communication between two distinct centers

    of the network.

    By 340 ms, the negativity becomes concentrated over

    central parietal scalp and the bilateral parietooccipital

    foci are diminishing in amplitude. By 400 ms, the

    effectis essentially over.

    Note that the reader may find it useful to use the manual

    advance option on their media player to control the progres-

    sion of the animation in order to get a better look at the

    topographic transition points detailed above.

    724 J.J. Foxe et al. / NeuroImage 19 (2003) 710726

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    16/17

    References

    Allison, T., Puce, A., Spencer, D, McCarthy, G., 1999. Electrophysiolog-

    ical studies of human face perception. I. Potentials generated in occipi-

    totemporal cortex by face and non-face stimuli. Cereb. Cortex 9,

    415430.

    Behrmann, M., 1999. Spatial reference frames and hemispatial neglect, in:

    Gazzaniga, M.S. (Ed.), The New Cognitive Neurosciences, 2nd edition.MIT Press, Cambridge, MA, pp. 651666.

    Bentin, S., Mouchetant-Rostaing, Y., Giard, M.H., Echallier, J.F., Pernier,

    J., 1999. ERP manifestations of processing printed words at different

    psycholinguistic levels: time course and scalp distribution. J. Cogn.

    Neurosci. 11, 235260.

    Beringer, J., 1995. Experimental Run Time System (Version 3.13).

    Berisoft Corp., Frankfurt, Germany.

    Bisiach, E., Geminiani, G., Berti, A., Rusconi, M.L., 1990. Perceptual and

    premotor factors of unilateral neglect. Neurology 40, 12781281.

    Bowers, D., Heilman, K.M., 1980. Pseudoneglect: effects of hemispace on

    a tactile line bisection task. Neuropsychologia 18, 491498.

    Butler, P.D., Schechter, I., Zemon, V., Schwartz, S.G., Greenstein, V.C.,

    Gordon, J., Schroeder, C.E., Javitt, D.C., 2001. Dysfunction of early-

    stage visual processing in schizophrenia. Am. J. Psychiatr. 158, 11261133.

    Cappa, S.F., Guariglia, C., Messa, C., Pizzamiglio, L., Zoccolotti, P., 1991.

    Computed tomography correlates of chronic unilateral neglect. Neuro-

    psychology 5, 195204.

    Corbetta, M., Kincade, J.M., Ollinger, J.M., McAvoy, M.P., Shulman,

    G.L., 2000. Voluntary orienting is dissociated from target detection in

    human posterior parietal cortex. Nature Neurosci. 3, 292297.

    Coull, J.T., Nobre, A.C., Frith, C.D., 2001. The noradrenergic 2 agonist

    clonidine modulates behavioural and neuranotomical correlates of hu-

    man attentional orienting and alerting. Cereb. Cortex 11, 7384.

    Cowey, A., Small, M., Ellis, S., 1994. Left visuo-spatial neglect can be

    worse in far than in near space. Neuropsychologia 32, 10591066.

    Darling, W.G., Rizzo, M., Butler, A.J., 2001. Disordered sensorimotor

    transformations for reaching following posterior cortical lesions. Neu-

    ropsychologia 39, 237254.DeYoe, E.A., Van Essen, D.C., 1988. Concurrent processing streams in

    monkey visual cortex. Trends Neurosci. 11, 219226.

    Doniger, G.M., Foxe, J.J., Murray, M.M., Higgins, B.A., Schroeder, C.E.,

    Javitt, D.C., 2001. Visual perceptual learning in human object recog-

    nition areas: a repetition priming study using high-density electrical

    mapping. NeuroImage 13, 305313.

    Doniger, G.M., Foxe, J.J., Murray, M.M., Higgins, B.A., Snodgrass, J.G.,

    Schroeder, C.E., Javitt, D.C., 2000. Activation time-course of ventral

    visual stream object-recognition areas: high density electrical mapping

    of perceptual closure processes. J. Cogn. Neurosci. 12, 615621.

    Driver, J., Pouget, A., 2000. Object centered visual neglect, or relative

    egocentric neglect? J. Cogn. Neurosci. 12, 542545.

    Fierro, B., Brighina, F., Oliveri, M., Piazza, A., La Bua, V., Buffa, D.,

    Bisiach, E., 2000. Contralateral neglect induced by right posteriorparietal rTMS in healthy subjects. Neuroreport 11, 15191521.

    Fierro, B., Brighina, F., Piazza, A., Oliveri, M., Bisiach, E., 2001. Timing

    of right parietal and frontal cortex activity in visuo-spatial perception:

    a TMS study in normal individuals. Neuroreport 12, 26052607.

    Fink, G.R., Marshall, J.C., Shah, N.J., Weiss, P.H., Halligan, P.W., Grosse-

    Ruyken, M., Ziemons, K., Zilles, K., Freund, H.J., 2000a. Line bisec-

    tion judgments implicate right parietal cortex and cerebellum as as-

    sessed by fMRI. Neurology 28, 13241331.

    Fink, G.R., Marshall, J.C., Weiss, P.H., Shah, N.J., Toni, I., Halligan, P.W.,

    Zilles, K., 2000b. Where depends on what: a differential functional

    anatomy for position discrimination in one- versus two-dimensions.

    Neuropsychologia 38, 17411748.

    Fink, G.R., Marshall, J.C., Weiss, P.H., Toni, I., Zilles, K., 2002. Task

    instructions influence the cognitive strategies involved in line bisection

    judgments: evidence from modulated neural mechanisms revealed by

    fMRI. Neuropsychologia 40, 119130.

    Fink, G.R., Marshall, J.C., Weiss, P.H., Zilles, K., 2001. The neural basis

    of vertical and horizontal line bisection judgments: an fMRI study of

    normal volunteers. NeuroImage 14, S59 67.

    Foxe, J.J., Doniger, G.M., Javitt, D.C., 2001. Early visual processing

    deficits in schizophrenia: impaired P1 generation revealed by high-

    density electrical mapping. Neuroreport 12, 38153820.

    Foxe, J.J., Simpson, G.V., Ahlfors, S.P., 1998. Parieto-occipital 10 Hz

    activity reflects anticipatory state of visual attention mechanisms. Neu-

    roreport 9, 39293933.

    Foxe, J.J., Simpson, G.V., 2002. Timecourse of activation flow from V1 to

    frontal cortex in humans: a framework for defining early visual

    processing. Exp. Brain Res. 142, 139150.

    Fu, K.G., Foxe, J.J., Murray, M.M., Higgins, B.A., Javitt, D.C., Schroeder,

    C.E., 2001. Attention-dependent suppression of distracter visual input

    can be cross-modally cued as indexed by anticipatory parieto-occipital

    alpha-band oscillations. Cogn. Brain Res. 12, 145512.

    Fuchs, M., Drenckhahn, R., Wischmann, H.A., Wagner, M., 1998. An

    improved boundary element method for realistic volume-conductor

    modeling. IEEE Trans. Biomed. Eng. 45, 980 997.

    Galati, G., Lobel, E., Vallar, G., Berthoz, A., Pizzamiglio, L., Le Bihan, D.,

    2000. The neural basis of egocentric and allocentric coding of space inhumans: a functional magnetic resonance study. Exp. Brain Res. 133,

    156164.

    Gross, C.G., Rocha-Miranda, C.E., Bender, D.B., 1972. Visual properties

    of neurons in inferotemporal cortex of the Macaque. J. Neurophysiol.

    35, 96111.

    Guthrie, D., Buchwald, J.S., 1991. Significance testing of difference po-

    tentials. Psychophysiology 28, 240244.

    Halligan, P.W., Marshall, J.C., 1991. Left neglect for near but not far space

    in man. Nature 350, 498500.

    He, B., Lian, J., Spencer, K.M., Dien, J., Donchin, E., 2001. A cortical

    potential imaging analysis of the P300 and novelty P3 components.

    Hum. Brain Mapping 12, 120130.

    Heilman, K.M., van den Abell, T., 1980. Right hemisphere dominance for

    attention: the mechanism underlying hemispheric asymmetries of inat-

    tention (neglect). Neurology 30, 327330.Janssen, P., Vogels, R., Liu, Y., Orban, G.A., 2001. Macaque inferior

    temporal neurons are selective for three-dimensional boundaries and

    surfaces. J. Neurosci. 21, 94199429.

    Jewell, G., McCourt, M.E., 2000. Pseudoneglect: a review and meta-

    analysis of performance factors in line bisection tasks. Neuropsycho-

    logia 38, 93110.

    Kaplan, E., Shapley, R., 1982. X and Y cells in the lateral geniculate

    nucleus of macaque monkeys. J. Physiol. 330, 125143.

    Kaplan, E., Shapley, R.M., 1986. The primate retina contains two types of

    ganglion cells, with high and low contrast sensitivity. Proc. Natl. Acad.

    Sci. USA 83, 27552757.

    Karnath, H.O., 2001. New insights into the functions of the superior

    temporal cortex. Nature Rev. Neurosci. 2, 568576.

    Karnath, H.O., Ferber, S., Himmelbach, M., 2001. Spatial awareness is a

    function of the temporal not the posterior parietal lobe. Nature 411,950953.

    Kerkhoff, G., 2001. Spatial hemineglect in humans. Prog. Neurobiol. 63,

    127.

    Martinez, A., Anllo-Vento, L., Sereno, M.I., Frank, L.R., Buxton, R.B.,

    Dubowitz, D.J., Wong, E.C., Hinrichs, H., Heinze, H.J., Hillyard, S.A.,

    1999. Involvement of striate and extrastriate visual cortical areas in

    spatial attention. Nature Neurosci. 2, 364369.

    Martinez, A., DiRusso, F., Anllo-Vento, L., Sereno, M.I., Buxton, R.B.,

    Hillyard, S.A., 2001. Putting spatial attention on the map: timing and

    localization of stimulus selection processes in striate and extrastriate

    visual areas. Vision Res. 41, 14371457.

    McCourt, M.E., 2001. Performance consistency of normal observers in

    forced-choice tachistoscopic visual line bisection. Neuropsychologia

    39, 10651076.

    725J.J. Foxe et al. / NeuroImage 19 (2003) 710726

  • 8/10/2019 Foxe Et Al-Right Hemisphere Control of Visuospatial Attention

    17/17

    McCourt, M.E., Freeman, P., Tahmahkera-Stevens, C., Chaussee, M.,

    2001. The influence of unimanual response on pseudoneglect magni-

    tude. Brain Cogn. 45, 5263.

    McCourt, M.E., Garlinghouse, M., 2000a. Asymmetries of visuospatial

    attention are modulated by viewing distance and visual field elevation:

    pseudoneglect in peripersonal and extrapersonal space. Cortex 36,

    715732.

    McCourt, M.E., Garlinghouse, M., 2000b. Stimulus modulation of pseudo-neglect: effect of line geometry. Neuropsychologia 38, 520524.

    McCourt, M.E., Garlinghouse, M., Slater, J., 2000. Centripetal versus

    centrifugal bias in visual line bisection: focusing attention on two

    hypotheses. Front. Biosci. 5, D5871.

    McCourt, M.E., Jewell, G., 1999. Visuospatial attention in line bisection:

    stimulus modulation of pseudoneglect. Neuropsychologia 37, 843

    855.

    McCourt, M.E., Olafson, C., 1997. Cognitive and perceptual influences on

    visual line bisection: psychophysical and chronometric analyses of

    pseudoneglect. Neuropsychologia 35, 369380.

    Mesulam, M.M., 2000. Attentional networks, confusional states, and ne-

    glect syndromes, in: Mesulam, M.M. (Ed.), Principles of Behavioral

    and Cognitive Neurology, 2nd edition. Oxford University Press, Ox-

    ford, pp. 174256.

    Michel, C.M., Thut, G., Morand, S., Khateb, A., Pegna, A.J., Grave dePeralta, R., Gonzalez, S., Seeck, M., Landis, T., 2001. Electric source

    imaging of human brain functions. Brain Res. Brain Res. Rev. 36,

    108118.

    Murray, M.M., Wylie, G.R., Higgins, B.A., Javitt, D.C., Schroeder, C.E.,

    Foxe, J.J., 2002. The spatio-temporal dynamics of illusory contour

    processing: combined high-density electrical mapping, source analysis,

    and functional magnetic resonance imaging. J. Neurosci. 22, 5055

    5073.

    Na, D.L., Adair, J.C., Hye Choi, S., Won Seo, D., Kang, Y., Heilman,

    K.M., 2000. Ipsilesional versus contralesional neglect depends on at-

    tentional demands. Cortex 36, 455467.

    Oldfield, R.C., 1971. The assessment and analysis of handedness: the

    Edinburgh Inventory. Neuropsychologia 9, 97113.

    Pascual-Leone, A., Gomez-Tortosa, E., Grafman, J., Alway, D., Nichelli,

    P., Hallett, M.., 1994. Induction of visual extinction by rapid-rate

    transcranial magnetic stimulation of parietal lobe. Neurology 44, 494

    498.

    Post, R.B., Caufield, K.J., Welch, R.B., 2001. Contributions of object- and

    space-based mechanisms to line bisection errors. Neuropsychologia 39,

    856864.

    Ritter, W., Vaughan Jr., H.G., 1969. Averaged evoked responses in vigi-

    lance and discrimination: a reassessment. Science 164, 326328.

    Robertson I.H., Halligan P.W., 1999. Spatial Neglect: A Clinical Hand-

    book for Diagnosis and Treatment. Psychology Press.

    Sary, G., Vogels, R., Orban, G.A., 1993. Cue-invariant shape selectivity of

    macaque inferior temporal neurons. Science 260, 995997.

    Scherg, M., Picton, T.W., 1991. Separation and identification of event-

    related potential components by brain electric source analysis. Electro-

    encephalogr. Clin. Neurophysiol. Suppl. 42, 2437.Scherg, M., Berg, P., 1996. New concepts of brain source imaging and

    localization. Electroencephalogr. Clin. Neurophysiol. Suppl. 46, 127

    137.

    Schroeder, C.E., Mehta, A.D., Foxe, J.J., 2001. Determinants and mecha-

    nisms of attentional modulation of neural processing. Front. Biosci. 6,

    D672684.

    Sclar, G., Maunsell, J.H.R., Lennie, P., 1990. Coding of image contrast in

    central visual pathways of the macaque monkey. Vision Res. 30, 110.

    Shapley, R., Perry, V.H., 1986. Cat and monkey retinal ganglion cells and

    their functional roles. Trends Neurosci. 9, 229235.

    Shipp, S., de Jong, B.M., Zihl, J., Frackowiak, R.S.J., Zeki, S., 1994. The

    brain activity related to residual motion vision in a patient with a

    bilateral lesion of V5. Brain 117, 10231038.

    Simpson, G.V., Foxe, J.J., Vaughan Jr., H.G., Mehta, A.D., Schroeder,

    C.E., 1995b. Integration of electrophysiological source analyses, MRI

    and animal models in the study of visual processing and attention.

    Electroencephalogr. Clin. Neurophysiol. Suppl. 44, 7692.

    Simpson, G.V., Pfleiger, M.E., Foxe, J.J., Ahlfors, S.P., Vaughan, Jr., H.G.,

    Hrabe, J., Ilmoniemi, R.J., Lantos, G., 1995a. Dynamic neuroimaging

    of brain function. J. Clin. Neurophysiol. 12, 118.

    Spinelli, D., Angelina, P., De Luca, M., Burr, D.C., 1996. VEP in neglect

    patients have longer latencies for luminance but not for chromatic

    patterns. Neuroreport 7, 815819.

    Spinelli, D., Burr, D.C., Morrone, M.C., 1994. Spatial neglect is associated

    with increased latencies of visual evoked potentials. Vis. Neurosci. 11,

    909918.

    Talairach, J., Tournoux, P., 1988. Co-Planar Stereotaxic Atlas of the

    Human Brain. Thieme, New York.

    Tegner, R., Levander, M., 1991a. Through a looking glass. A new tech-

    nique to demonstrate directional hypokinesia in unilateral neglect.

    Brain 114, 19431951.

    Tegner, R., Levander, M., 1991b. The influence of stimulus properties on

    visual neglect. J. Neurol. Neurosurg. Psychiatry. 54, 882887.

    Tootell, R.B., Hamilton, S.L., Switkes, E., 1988. Functional anatomy of

    macaque striate cortex. IV. Contrast and magno-parvo streams. J. Neu-

    rosci. 8, 1594 1609.Tootell, R.B., Reppas, J.B., Dale, A.M., Look, R.B., Sereno, M.I., Malach,

    R., Brady, T.J., Rosen, B.R., 1995a. Visual motion aftereffect in human

    cortical area MT revealed by functional magnetic resonance imaging.

    Nature 375, 139141.

    Tootell, R.B.H., Reppas, J.B., Kwong, K.K., Malach, R., Born, R.T.,

    Brady, T.J., Rosen, B.R., Belliveau, J.W., 1995b. Functional analysis of

    human MT and related visual cortical areas using functional magnetic

    resonance imaging. J. Neurosci. 15, 32153230.

    Vallar, G., 2001. Extrapersonal visual unilateral spatial neglect and its

    neuroanatomy. NeuroImage 14, S5258.

    Vallar, G., Lobel, E., Galati, G., Berthoz, A., Pizzamiglio, L., Le Bihan, D.,

    1999. A fronto-parietal system for computing the egocentric spatial

    frame of reference in humans. Exp. Brain Res. 124, 281286.

    Vallar, G., Perani, D., 1986. The anatomy of unilateral neglect after

    right-hemisphere stroke lesions. A clinical/CT-scan correlation study inman. Neuropsychologia 24, 609 622.

    Vallar G., Perani D., 1987. The anatomy of spatial neglect in humans, in:

    Jennerod M. (Ed.), Neurophysiological and Neuropsychological As-

    pects of Spatial Neglect, North Holland, Amsterdam, pp. 235258.

    Van Hoorhis, S.T., Hillyard, S.A., 1977. Visual evoked potentials and

    selective attention to points in space. Percept. Psychophys. 22, 5462.

    Vuilleumier, P., Landis, T., 1998. Illusory contours and spatial neglect.

    Neuroreport 9, 24812484.

    Vuilleumier, P., Valenza, N., Mayer, E., Reverdin, A., Landis, T., 1998.

    Near and far visual space in unilateral neglect. Ann. Neurol. 43,

    406410.

    Walker, R., 1995. Spatial and object-based neglect. Neurocase 1, 371383.

    Webster, M.J., Ungerleider, L.G., 1998. Neuroanatomy of visual attention,

    in: Parasuraman, R. (Ed.), The Attentive Brain. MIT Press, Cambridge,

    MA, pp. 1934.Weintraub, S., Mesulam, M.M., 1987. Right cerebral dominance in spatial

    attention. Further evidence based on ipsilateral neglect. Arch. Neurol.

    44, 621625.

    Weiss, P.H., Marshall, J.C., Wunderlich, G., Tellmann, L., Halligan, P.W.,

    Freund, H.J., Zilles, K., Fink, G.R., 2000. Neural consequences of

    acting in near versus far sp