The attention network of the human brain: Relating structural damage associated with spatial neglect to functional imaging correlates of spatial attention

Neuropsychologia 49 (2011) 3063–3070

Contents lists available at ScienceDirect

Neuropsychologia journal homepage: www.elsevier.com/locate/neuropsychologia

The attention network of the human brain: Relating structural damage associated with spatial neglect to functional imaging correlates of spatial attention Radek Ptak a,b,∗ , Armin Schnider a,b a b

Division of Neurorehabilitation, Department of Clinical Neurosciences, Geneva University Hospitals, Geneva, Switzerland Department of Clinical Neurosciences, Medical School, University of Geneva, Geneva, Switzerland

a r t i c l e

i n f o

Article history: Received 4 April 2011 Received in revised form 29 June 2011 Accepted 8 July 2011 Available online 19 July 2011 Keywords: Visual attention Spatial hemineglect Attention network Frontal eye fields Posterior parietal cortex Lesion analysis

a b s t r a c t Functional imaging studies of spatial attention regularly report activation of the intraparietal sulcus (IPS) and dorsal premotor cortex including the frontal eye fields (FEF) in tasks requiring overt or covert shifting of attention. In contrast, lesion-overlap studies of patients with spatial neglect – a syndrome characterized by severe impairments of spatial attention – show that the critical damage concerns more ventral regions, comprising the inferior parietal lobule, the temporal–parietal junction (TPJ), and the superior temporal gyrus. We performed voxel-based lesion-symptom mapping of 29 right-hemisphere stroke patients, using several performance indices derived from a cueing task as measures of spatial attention. In contrast to previous studies, we focused our analyses on eight regions of interest defined according to results of previous functional imaging studies. A direct comparison of neglect with control patients revealed that neglect was associated with damage to the TPJ, the middle frontal gyrus, and the posterior IPS. The latter region was also a significant predictor of the degree of contralesional slowing of target detection and the extent to which ipsilesional distracters captured attention of neglect patients. Finally, damage to the FEF and posterior IPS was negatively correlated with the tendency of neglect patients to orient attention toward behaviourally relevant distracters. These findings support the results of functional imaging studies of spatial attention and provide evidence for a network account of neglect, according to which attentional selection of relevant environmental stimuli and the reorienting of attention result from dynamic interactions between the IPS, the dorsal premotor cortex, and the TPJ. © 2011 Elsevier Ltd. All rights reserved.

1. Introduction Functional neuroimaging studies of spatial attention consistently report activations of several regions located in the parietal and frontal association cortex when participants are engaged in tasks probing spatial attention. The prototypical paradigm used to examine regions involved in the shifting of attention is the spatial cueing task (Eriksen & Hoffman, 1972; Posner, 1980): participants are required to detect or discriminate a stimulus presented left or right of fixation following a brief cue that summons attention to the

Abbreviations: ANOVA, analysis of variance; BOLD, blood–oxygen level dependent; FEF, frontal eye field; fMRI, functional magnetic resonance imaging; IPL, inferior parietal lobule; IPS, intraparietal sulcus; TPJ, temporal–parietal junction; MFG, middle frontal gyrus; ROI, region of interest; VLSM, voxel-based lesionsymptom mapping; VOI, volume of interest. ∗ Corresponding author at: Division of Neurorehabilitation, Geneva University Hospitals and, Faculty of Medicine, University of Geneva, 26, Av. de Beau-Séjour, 1211 Geneva 14, Switzerland. Tel.: +41 22 382 35 24; fax: +41 22 382 83 38. E-mail address: [email protected] (R. Ptak). 0028-3932/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuropsychologia.2011.07.008

left or right hemifield. At the behavioural level, this task reveals that detection of the target is faster when the cue indicates its correct position (valid cue) than when it orients attention to the hemifield opposite the target (invalid cue, Posner, 1980; Posner & Petersen, 1990). At the anatomical level, this pattern results in strong activation increases in dorsal fronto-parietal cortex including the intraparietal sulcus (IPS) and the frontal eye fields (FEF), a cortical area with a decisive role in saccade programming (Bruce & Goldberg, 1985; Paus, 1996; Pierrot-Deseilligny, Ploner, Müri, Gaymard, & Rivaud-Péchaux, 2002). The IPS and FEF exhibit strong activity related to the cue – independently of whether subjects move their attention covertly or by shifting their gaze – though both regions also show target-evoked activity (Corbetta, Kincade, Ollinger, McAvoy, & Shulman, 2000; Hopfinger, Buonocure, & Mangun, 2000; Kastner, Pinsk, De Weerd, Desimone, & Ungerleider, 1999; Mort et al., 2003b; Perry & Zeki, 2000; Yantis et al., 2002). In contrast to these dorsal regions the activation of more ventral areas – the inferior parietal lobule (IPL), the temporal–parietal junction (TPJ), and the lateral prefrontal cortex – is more variable, as suggested by the failure of some studies to observe significant activity (Corbetta

3064

R. Ptak, A. Schnider / Neuropsychologia 49 (2011) 3063–3070

et al., 1998; Gitelman et al., 1999). However, studies attempting to distinguish activity related to the cue from activity evoked by the target reported that TPJ activity is mainly associated with the target (Corbetta et al., 2000; Kincade, Abrams, Astafiev, Shulman, & Corbetta, 2005). In addition, while the IPS and FEF increase their activity whether subjects shift attention voluntarily or reflexively, the TPJ only becomes active when a stimulus of high behavioural relevance (e.g. a distracter that possesses some target-defining properties) appears at an unexpected position (Corbetta et al., 2000; Indovina & Macaluso, 2007; Natale, Marzi, & Macaluso, 2010). Compared to these findings, lesion studies of spatial neglect show a different picture. Patients with left neglect exhibit striking deficits of spatial attention: they may fail to find an object left of their body, particularly when it is presented together with other, distracting objects on their right; they may bump into objects on their left, or fail to react to a person addressing them from their left side (Halligan, Fink, Marshall, & Vallar, 2003; Kerkhoff, 2001; Milner & McIntosh, 2005). Neglect patients also show a supramodal deficit of attentional disengagement from ipsilesional cues (Bartolomeo, Siéroff, Decaix, & Chokron, 2001; Golay, Hauert, Greber, Schnider, & Ptak, 2005; Posner, Walker, Friedrich, & Rafal, 1987). Several lesion studies have attempted to specify the region critical for spatial neglect by analysing the maximal overlap of multiple lesions or by comparing patients with neglect to right-hemisphere damaged patients without neglect. These studies identified the inferior parietal lobule (Golay, Schnider, & Ptak, 2008; Mort et al., 2003a; Vallar & Perani, 1986), the white matter beneath the central sulcus (Doricchi & Tomaiuolo, 2003), and the superior temporal gyrus (Karnath, Fruhmann Berger, Küker, & Rorden, 2004) as the brain regions whose damage is most predictive of spatial neglect. Importantly, none of the studies that used a lesion-averaging approach reported an association of spatial neglect with damage to the IPS or FEF. Thus, whereas functional imaging studies of spatial attention regularly report activation of a dorsal fronto-parietal network, lesion studies frequently report damage to more ventral regions and fail to present evidence that would support involvement of the dorsal network. A tempting solution to this paradox is to ascribe such discrepancies to the methodological limits of functional imaging and lesion studies. For example, activation of the dorsal fronto-parietal network might be considered as epiphenomenal and thus not necessary for spatial attention. However, this explanation fails to consider two important findings: First, neurons in the posterior parietal cortex (Bisley & Goldberg, 2003; Bushnell, Goldberg, & Robinson, 1981; Cohen, Cohen, & Gifford, 2004; Constantinidis & Steinmetz, 2001; Gottlieb, Kusunoki, & Goldberg, 1998) and the FEF (Bichot & Schall, 1999; Fecteau & Munoz, 2006; Thompson, Hanes, Bichot, & Schall, 1996) are critically involved in biasing attention in favour of particular stimuli or locations in space. Second, though structurally intact, the dorsal fronto-parietal network is nevertheless functionally impaired in the acute phase of neglect (Corbetta, Kincade, Lewis, Snyder, & Sapir, 2005), and damage or functional inhibition of the main fibre tract interconnecting the intact IPS and FEF results in signs of neglect (Shinoura et al., 2009; Thiebaut de Schotten et al., 2005). Here, we explore the possibility that the dorsal regions apparently lying outside the typical lesion area of neglect patients are not systematically damaged and thus escape the lesion averaging approach. Indeed, previous lesion studies were mostly interested in the ‘critical’ damage predicting spatial neglect relative to control patients, and thus may have failed to highlight rarely damaged, yet important regions lying in the dorsal network. Neglect is a heterogeneous disorder, and may affect attentional, intentional, or representational processes to different degrees, as a function of the extent to which damage extends into parietal (Binder, Marshall, Lazar, Benhamin, & Mohr, 1992; Golay et al., 2008), temporal

(Hillis et al., 2005; Ptak & Valenza, 2005), or prefrontal cortex (Husain & Kennard, 1997; Verdon, Schwartz, Lovblad, Hauert, & Vuilleumier, 2010). If regions such as the FEF, prefrontal cortex, or cortex around the IPS are only involved in a relatively small number of neglect patients (or subgroups characterized by specific patterns of impairment) this comparison will fail to confirm their importance. This problem has much to do with statistical power. Voxel-based lesion-symptom mapping (VLSM) analyses examine the implication of anatomical regions in the manifestation of a behavioural symptom by performing a statistical test on each damaged voxel (Bates et al., 2003; Kimberg, Coslett, & Schwartz, 2007; Rorden, Karnath, & Bonilha, 2007). Given that across several patients many thousands of voxels may be damaged this procedure amounts to a vast number of statistical tests being performed. Consequently, in order to decrease the risk of alpha (false positive) error it is important to perform adequate corrections of the level of statistical significance. However, given the high number of statistical comparisons the correction of the alpha-level often leads to the statistical analysis being overly conservative, seriously limiting the number of significant results. In order to reduce the number of statistical tests to be performed, functional neuroimaging studies often confine statistical analysis to specific regions of interest (ROI). The methodological rationale of this approach relies on the assumption that the choice of ROIs is independent of the statistical significance observed in the examined data set: independence would be violated if only regions that provide the most significant results in a preliminary analysis were selected as ROIs for further statistical treatment (Vul, Harris, Winkielman, & Pashler, 2009). The aim of this study was to verify the hypothesis that the lesion-overlap method underestimates the contribution of key regions of the fronto-parietal attention network to spatial attention deficits in neglect. We therefore used the ROI approach to analyse a data set comprising 20 neglect patients whose maximal overlap of lesions is localized in the right inferior parietal lobule (Ptak & Schnider, 2010). We avoided the independence error by selecting ROIs based on regions identified by previous functional imaging studies of spatial attention. In addition, we tested whether damage to specific ROIs is predictive of performance in parameters of attentional orienting derived from a spatial cueing task.

2. Materials and methods 2.1. Participants Twenty patients with left spatial neglect (13 females), 10 right-hemisphere (RH) damaged control patients without neglect (4 females), and 10 healthy controls (6 females) were involved in this study. Since the focus of the study was on impairments of attentional orienting that characterize neglect, we did not seek to equalize group sizes. Approval was obtained from the ethical committee of the University Hospitals Geneva, and all participants gave written consent. Patients were examined within a week while hospitalized for neurorehabilitation following a first-ever ischemic or haemorrhagic stroke. Table 1 shows demographic data and the results of clinical testing of neglect and control patients. All patients had preserved visual fields for the central ∼20◦ as assessed with computerized perimetry testing and/or clinical confrontation. All neglect patients manifested behavioural signs of left unawareness (e.g. failure to notice objects or persons placed on their left, difficulties with dressing or grooming) and lateralized failures in the following neglect tests: ‘Bells’ cancellation (Gauthier, Dehaut, & Joanette, 1989), cancellation of inverted among upright Ts, line bisection (Schenkenberg, Bradford, & Ajax, 1980), sentence copying (Wilson, Cockburn, & Halligan, 1987), and copying a landscape. The three groups had similar age (ANOVA: F2,37 = .13), and the elapsed time between lesion onset and testing was comparable between the two patient groups (T-test: t28 = .88). The neglect group scored significantly worse compared to RH-controls on ‘Bells’ cancellation (Mann–Whitney z = 3.67, p < .001), ‘T’ cancellation (z = 4.41, p < .0001), line bisection (z = 2.46, p < .05), sentence copying (z = 2.30, p < .05) and the landscape-copying test (z = 3.42, p < .001).

R. Ptak, A. Schnider / Neuropsychologia 49 (2011) 3063–3070

3065

Table 1 Demographic characteristics and performance in clinical neglect tests (mean ± SEM).

Healthy controls RH-controls Neglect

Age

Time post (days)

Bells cancellation (left omissions)

T cancellation (left omissions)

Line bisection (% right bias)

Sentence copying (words missed)

Drawing

(63.6±3.8) 66.8 (±3.8) 65.4 (±2.7)

– 57.8 (±13.4) 46.4 (±6.4)

– 2.1 (±0.8) 10.3 (±1.1)

– 0.6 (±0.3) 17.4 (±2.2)

– 4.0 (±1.1) 10.5 (±1.9)

– 0 (±0) 4.2 (±1.5)

– 2 (±0) 0.8 (±0.2)

Drawing performance was scored as follows: 0 = flagrant omissions on the left side; 1 = distorted left side; 2 = intact.

2.2. Stimuli and procedure The experimental task used to examine deficits of spatial attention and to define several parameters of the orienting of attention required participants to react to coloured target stimuli appearing left or right of fixation, preceded by a lateralized cue (Ptak & Schnider, 2010). The stimulus set consisted of four coloured rectangles (red, green, blue or yellow; not matched for luminance) subtending 4 × 2 degrees, presented on neutral grey background with their inner border 1.5 degrees from fixation. For each participant two of these colours (e.g. red and green) were randomly selected as target colours while the other colours served as distracters; therefore, any bias due to luminance differences between colours was controlled by randomization. On every trial a white fixation cross was presented for 2000 ms, followed by the cue (300 ms), a blank screen (200 ms), and the target display (until the space-key was pressed, max. 3000 ms). Being interested in the deficits of attentional reorienting characterizing spatial neglect, we chose a stimulus-onset asynchrony (500 ms) at which we previously found a reliable disengagement deficit in neglect patients (Ptak & Schnider, 2006). The target display either contained a colour target presented left or right of fixation, paired with a colour distracter on the opposite side (e.g. red–yellow), or two colour distracters (e.g. yellow–blue). The cue was a rectangle presented on the same side as the upcoming target (valid cue) or on the opposite side (invalid cue). It either had a task-relevant and identical colour as the target (e.g. red cue–red target), a task-relevant but different colour than the target (e.g. green cue–red target), or a task-irrelevant colour (e.g. yellow cue–red target). Relevant/irrelevant cues and valid/invalid cues were equally likely, and the occurrence of a target or its position could not be inferred from the presence or position of a particular cue. Target position (left, right), cue relevance (relevant-identical, relevant-different, irrelevant) and cue validity (valid, invalid) were orthogonally varied in blocks of 72 trials containing 48 target-present and 24 target-absent trials. Participants were instructed to depress the space-button if they detected either of the two targets while disregarding the cues and withholding any reaction to them. Fixation was controlled visually during a practice run and checked regularly during the experimental runs. 2.3. Lesion mapping All neglect patients and 9 patients of the RH-control group underwent structural magnetic resonance imaging (MRI) including T1- and T2-weighted acquisitions with a between-slice resolution of 4 mm, on a 1.5 T MRI scanner (Siemens Vision, Munich, Germany). For each patient the area of damage (volume of interest, VOI) was delineated directly on an axial T2-weighted MRI scan using a graphics tablet and MRIcron (Rorden et al., 2007). Each patient MRI including the lesion-VOI was then normalized to standard space using SPM5 (www.fil.ion.ucl.ac.uk/spm). A mask was applied over the lesion in order to minimize normalization artefacts due to contribution of abnormal tissue during computation of the transformation matrix (Brett, Leff, Rorden, & Ashburner, 2001). The normalized lesion VOIs were overlaid on the T1-weighted template MRI-scan from the Montreal Neurological Institute (MNI, http://www.bic.mni.mcgill.ca). Regions of interest (ROIs) were taken from the study of He et al. (2007), who determined ROIs based on a meta-analysis of four fMRI studies of young partici-

Table 2 Designation and MNI-coordinates of the four dorsal and four ventral ROIs examined in the study. Dorsal regions

x

y

z

FEF (frontal eye field) pIPS (posterior intraparietal sulcus) vIPS (ventral intraparietal sulcus) MT+ (middle temporal region) Ventral regions MFG (middle frontal gyrus) Ins (anterior insula) TPJ (temporal–parietal junction) STS (superior temporal sulcus)

32 23 30 42

−12 −69 −83 −70

52 49 13 −11

39 38 49 56

11 16 −53 −52

38 2 28 9

pants performing a spatial cueing task. The Talairach atlas (Talairach & Tournoux, 1988) coordinates of the centre of these ROIs were converted to MNI-coordinates (see Table 2 and Fig. 1), and ROIs were then defined as regular spheres with a radius of 10 mm, using SPM with the MarsBar toolbox (http://marsbar.sourceforge.net). Following smoothing the resulting eight ROIs had an average volume of 3.67 ccm. The normalized lesion of each patient was segmented with each of the eight ROIs, and all computations were performed on a voxel-by-voxel basis, using these segmented lesions. When performing VLSM analyses one might be interested in answering two types of questions: first, does a group of patients, whose particular constellation of symptoms differentiates it from brain-injured patients not showing these symptoms, have distinct brain damage? Second, does a discrete behavioural measure correlate with damage to specific brain regions? As regards the present data, the first type of analysis informs us that a brain region is a predictor of neglect, without telling anything about its specific function. In contrast, based on the second type of analysis precise hypotheses may be formulated concerning the role of specific brain regions for spatial attention. Voxel-based lesion-symptom mapping (VLSM) was performed with the NPM software (distributed with MRIcron), which offers two VLSM methods to study the relation between a behavioural measure and anatomy. The nonparametric Liebermeister test is performed on binomial data and thus requires patients to be assigned to two different groups based on a behavioural measure. The more conservative, nonparametric Brunner–Munzel test identifies brain regions that are critical for performance using a continuous behavioural measure. In order to examine whether neglect is associated with damage to specific ROIs, we first performed a comparison between neglect and RH-control patients using the Liebermeister test. For this comparison, it was assumed that neglect patients form a category that is qualitatively distinct from control participants, and VLSM was performed using a binomial measure (neglect present = 0; absent = 1). We then examined to what extent damage to the eight ROIs was correlated with specific measures of spatial attention by computing Brunner–Munzel tests with four performance indices derived from the spatial cueing task. These comparisons were performed on the neglect group only, since the primary aim was to test the prediction that damage to ROIs not previously identified by lesion studies comparing neglect with non-neglect patients is nevertheless an important predictor of specific impairments within the neglect group. Performance indices were computed as follows: Target detection index: this index reflects the speed of target detection in the left relative to the right hemifield, and was computed as lvf − rvf (the mean RT to left targets minus the mean RT to right targets). A positive value indicates slower RTs to left hemifield targets, irrespective of cue validity and cue relevance. Regions identified with the VLSM analysis are predictors of a large target detection index. Validity effect: this index reflects the slowing of RTs to left targets following invalid cues relative to valid cues, and was computed as lvf-invalid − lvf-valid (the mean RT to left targets following invalid cues minus the mean RT to left targets following valid cues). A positive value indicates slower RTs following invalid than valid cues, irrespective of cue relevance. Regions identified with the VLSM analysis are predictors of a large validity effect. Relevance effect: the relevance effect reflects the relative slowing of RTs to invalidly cued targets in the left hemifield when cues were relevant compared to when cues were irrelevant, and was computed as lvf-invalid-relevant − lvf-invalid-irrelevant (the mean RT to left targets following invalid-relevant cues minus the mean RT to left targets following invalid-irrelevant cues). A positive value indicates slower RTs following relevant than irrelevant cues, and regions identified with the VLSM analysis are predictors of a small relevance effect. The critical brain lesion underlying the relevance effect has been identified previously (Ptak & Schnider, 2010); however, since that previous study did not use the ROI approach we included analyses of this effect fore the sake of completeness. Priming effect: the priming effect expresses the degree to which a validdissimilar cue slows down RTs compared to a valid-similar cue, and was computed as lvf-valid-different − lvf-valid-identical (the mean RT to left targets following valid-identical cues minus the mean RT to left targets following valid-different cues). A positive value indicates faster RTs following visually similar compared to dissimilar cues, and regions identified with the VLSM analysis are predictors of a small priming effect. The priming effect reflects perceptual facilitation and is thus the only index that is driven by sensory rather than attentional processes. It was included in our analysis, because there are no specific predictions regarding the role of any of the

3066

R. Ptak, A. Schnider / Neuropsychologia 49 (2011) 3063–3070

Fig. 1. Axial and lateral view of the MNI-brain with superposed regions of interest (Abbreviations, see text). Employing the distinction made by Corbetta and Shulman (2002) based on functional brain imaging data, regions of interest are coloured according to whether they belong to the dorsal (DAN) or ventral attention network (VAN). The hatched pattern of the MFG indicates that activations of this region are correlated with both attention networks.

eight ROIs in generating sensory priming. This effect therefore provides a control against which anatomical findings with the other three indices can be validated. VLSM-analyses were computed for each of these four indices and each of eight ROIs. In order to control for multiple comparisons, family-wise error rates (FWE) were computed using repeated permutation tests (n = 1000), which provide a less conservative yet equally powerful correction for multiple comparisons than the Bonferroni correction (Rorden et al., 2007).

3. Results 3.1. Behavioural results Target misses were very low in healthy participants (mean omission rate: 0.4%) and RH-control patients (1.3%) and were therefore not analysed further. Neglect patients missed significantly more targets presented in the left (5.5%) compared to the right hemifield (2.1%; Wilcoxon-test: z = 3.1, P < .01), though misses of targets in the left hemifield were comparable between valid (5.7%) and invalid (5.4%) cueing. Fig. 2 shows mean reaction time (RT) data of healthy controls, RH-damaged controls, and neglect patients. The data of each group were submitted to repeated-measures ANOVA with the factors target position (left/right hemifield), validity of the cue (valid/invalid) and cue relevance (relevant-identical/relevantdifferent/irrelevant). Significant effects were followed up with post-hoc Fisher (least-significant difference, LSD) tests. The data of healthy participants (Fig. 2a) and of RH-controls (Fig. 2b) were characterized by a significant main effect of cue relevance (healthy: F2,18 = 34.94, P < .0001; RH-controls: F2,18 = 24.26, P < .0001). In both groups, independently of cue validity relevant-identical cues facilitated target detection relative to relevant-different and irrelevant

cues (all P < .01), while the latter two resulted in indistinguishable performance. No other effect or interaction reached significance. In contrast, the ANOVA of neglect patient data (Fig. 2c) revealed main effects of target position (F1,19 = 28.35, P < .0001), cue validity (F1,19 = 8.2, P < .01) and cue relevance (F2,38 = 7.84, P < .01). These effects were not further analysed because they were embedded in higher-order interaction effects of validity × relevance (F2,38 = 23.77, P < .0001) and validity × target position (F1,19 = 15.16, P < .001). The interaction between validity and relevance reflected the finding, that neglect patients reacted faster to validly than to invalidly cued targets when cues were relevant-identical (P < .0001) or relevant-different (P < .0001), but not when they were irrelevant. The interaction between validity and target position was due to the fact that the slowing of RTs following invalid compared to valid cues was more pronounced for left than right hemifield targets. This effect is the critical finding suggesting a deficit of attentional disengagement from ipsilesional cues, which several authors consider as distinctive feature of spatial neglect (Losier & Klein, 2001; Morrow & Ratcliff, 1988; Posner et al., 1987). In order to extract from these relatively complex experimental data condensed measures of spatial attention that could be entered in a VLSM analysis, we computed four performance indices respectively reflecting the global speed of target detection, the influence of invalid cues (validity effect), of relevant cues (relevance effect), and of visually similar cues on performance (priming effect). Fig. 3 shows the average performance of the three groups on these four performance indices. A Pearson correlation analysis revealed no significant correlations between these indices, suggesting that they reflected independent measures. Between-group differences were analysed with a one-way ANOVA. For the target detection

Fig. 2. Reaction-time data of healthy controls, right-hemisphere damaged controls, and neglect patients. Note that different scales were used to represent data of healthy controls and the two patient groups. LVF/RVF: left/right visual field. Error bars show SEM.

R. Ptak, A. Schnider / Neuropsychologia 49 (2011) 3063–3070

Fig. 3. Average (±SEM) target-detection index, validity effect, relevance effect and priming effect.

index the analysis revealed a significant difference between groups (F2,37 = 10.76, P < .001). While neglect patients had a significantly higher target detection index than healthy controls (P < .001) and RH-controls (P < .01) the control groups did not differ from each other. Also, neglect patients were the only group whose target detection index significantly differed from zero (t19 = 5.32, P < .001). Similarly, the analysis of the validity effect showed a significant effect of group (F2,37 = 6.55, P < .01), indicating that neglect patients had a larger validity effect than healthy and RH-controls (both P < .01). Only neglect patients had a significantly positive validity effect (t19 = 3.40, P < .01). A comparable analysis of the relevance effect was also significant (F2,37 = 3.66, P < .05) due to a larger relevance effect of neglect patients compared to both control groups (both P < .05). Again, only neglect patients had a relevance effect that was significantly different from zero (t19 = 2.28, P < .05). Finally, the analysis of the priming effect showed no significant difference between groups (F2,37 = 0.75). The priming effect was the only index on which all three groups had significantly positive values (healthy controls: t9 = 5.16, P < .001; RH-controls: t9 = 6.10, P < .001; neglect: t19 = 2.48, P < .05). In sum, performance of neglect patients differed from the two control groups only on the three indices that evaluate the shifting and disengagement of attention, while it was comparable on the sole index that evaluates sensory facilitation of target detection in the left hemifield. 3.2. Lesion-symptom mapping results One potentially confounding variable that might explain differential involvement of specific brain regions when comparing neglect and control patients is lesion volume. Previous studies consistently reported higher lesion volume in neglect compared to RH-damaged control patients (e.g., Golay et al., 2008; Mort et al., 2003a). In order to assess whether lesion volume is a predictor of damage to specific ROIs independently of the presence of neglect, we first performed VLSM analyses across the neglect

3067

and RH-control group with lesion volume as independent predictor variable. These analyses revealed significant effects for all eight ROIs. Therefore, if lesion volume were the critical factor predicting involvement of specific brain regions in neglect, we would expect neglect patients to have more frequent damage to all ROIs compared to control patients. Instead, VLSM analyses using a binary classification (Liebermeister test) of patients as belonging to the neglect or the non-neglect group only revealed significant differences in three ROIs (Fig. 4; FWE-corrected significance level of P < .05): the TPJ (1460 voxels), the pIPS (870 voxels) and the MFG (140 voxels). These results indicate that, while the TPJ, pIPS and MFG specifically predict the presence of neglect, lesion volume only predicts whether an ROI will be damaged, independently of whether neglect is actually present. Our next analyses aimed at identifying regions within the neglect group that were correlated with specific patterns of performance in the spatial cueing task (Fig. 5). A VLSM analysis with the target detection index as continuous measure (Brunner–Munzel test) revealed that the extent of general slowing of contralesional target detection in neglect patients was only predicted by damage to the pIPS (270 voxels). A similar result was obtained when the analysis was performed using the validity effect as control measure: again, damage to the pIPS (190 voxels), followed by the MT (70 voxels) turned out to be a significant predictor of large validity effects. In contrast, VLSM analysis using the relevance effect revealed a negative correlation between this measure and damage to the FEF (50 voxels) and the pIPS (170 voxels). This finding indicates that damage to these two ROIs was associated with a small or absent relevance effect. Finally, we performed a VLSM analysis with the priming effect as predictor; in contrast to the other measures this analysis did not reveal any significant effects.

4. Discussion Previous studies of the anatomy of spatial neglect did not report an association between neglect and regions identified by functional imaging studies as critical nodes of a dorsal attention network. The present study differs from these studies in two important aspects: first, in order to increase the statistical power of the lesion analyses we defined ROIs based on previous functional imaging findings, and performed VLSM analyses only on these selected brain regions. Second, similarly to functional imaging studies we searched for brain areas whose damage is a predictor of specific patterns of performance in a cueing task measuring dynamic aspects of spatial attention. Due to the exclusion of patients with hemianopia our patient group is not necessarily representative of a clinical sample of neglect patients, many of whom present visual field defects. However, the advantage is that performance of neglect patients reflects the expression of pure neglect and is not contaminated by effects of visual impairment. Our results reveal that regions of the dorsal attention network, which were not previously considered as critical for the occurrence of spatial neglect, are significant pre-

Fig. 4. Sagittal sections through the MNI-brain showing results of the binary-measure voxel-based lesion-symptom analysis. The figure shows brain regions that are significant predictors of spatial neglect.

3068

R. Ptak, A. Schnider / Neuropsychologia 49 (2011) 3063–3070

Fig. 5. Sagittal sections through the MNI-brain showing results of continuous-measures voxel-based lesion-symptom analyses. The figure shows brain regions that are significant predictors of a large target-detection index, a large validity effect, and a small relevance effect in patients with spatial neglect.

dictors of the degree of impairment shown by neglect patients in experimental measures of attention. Based on neurophysiological and neuroimaging findings several authors conceptualized the impairments associated with spatial neglect within network models of spatial attention (Corbetta & Shulman, 2002; Mesulam, 1990). However, these models faced the problem that important components of these networks appeared to be preserved in neglect. Only recently did clinical studies using intraoperative electrical stimulation (Thiebaut de Schotten et al., 2005) or diffusion tensor imaging (Urbanski et al., 2011) of frontoparietal fibre tracts provide support for a network model of neglect (see Bartolomeo, Thiebaut de Schotten, & Doricchi, 2007; Ptak, 2011). Here, we performed VLSM analyses with continuous measures as indices of performance in the spatial cueing task in an attempt to follow the logic of functional imaging studies and to seek for structural predictors of specific disturbances associated with spatial neglect. The regions determined with this approach represent nodes of a distributed cortical network comprising the TPJ, the IPS, and the dorsal premotor cortex including the FEF and MFG. The categorical VLSM analysis identified the TPJ as being critical for the occurrence of spatial neglect, a finding that is consistent with previous lesion studies reporting that damage to the angular or supramarginal gyrus is a strong predictor of neglect (Golay et al., 2008; Mort et al., 2003a; Vallar & Perani, 1986). This consistence with previous findings suggests that we examined a representative neglect group that shares anatomical features with neglect patients examined in other studies. However, though the TPJ was associated with neglect, none of the continuous-measures VLSM analyses revealed significant effects in this region. The fact that the TPJ was not a predictor of a particular pattern of performance in the cueing task suggests that it is crucial for the occurrence of neglect, yet does not have specific attention functions. This finding appears to contradict the conclusion suggested by some anatomical and functional imaging studies that the TPJ is specifically involved in the disengagement of attention (Friedrich, Egly, Rafal, & Beck, 1998) and re-orienting toward behaviourally relevant stimuli (Corbetta,

Patel, & Shulman, 2008; Indovina & Macaluso, 2007; Kincade et al., 2005; Natale et al., 2010). However, re-orienting cannot be the only function of the TPJ, as this region not only exhibits activity associated with attention switching, but is also important for non-spatial aspects of attention such as maintaining alertness, sustaining attention, and detecting novelty (Singh-Curry & Husain, 2009). Our findings indicate that the TPJ may serve functions whose conjoint impairment characterizes spatial neglect, and thus possibly represents a general-purpose re-orienting and alerting system. In addition to the TPJ, neglect was also associated with damage to the IPS. At the first sight this finding seems surprising, as previous lesion studies reported a more inferior localization of damage in spatial neglect, which was centred on the IPL, superior temporal cortex, or subcortical white matter (Doricchi & Tomaiuolo, 2003; Golay et al., 2008; Mort et al., 2003a; Vallar & Perani, 1986). However, this is the first study that combined an ROI approach with a VLSM analysis instead of performing a simple subtraction analysis. Such an approach is more sensitive in revealing the specific impact of IPS damage on attentional impairments. Several lines of evidence suggest that the IPS is one of the central nodes of the attention network of the brain. In the present study damage to this region was not only associated with spatial neglect, but was also a predictor of the degree of contralesional slowing, of the disengagement deficit, and of the effect of behavioural relevance on dynamic aspects of attention. In contrast, IPS damage did not predict the degree of priming from visually similar stimuli in the contralesional hemifield, suggesting that it is important for attention, but not for low-level sensory processes. These findings, as well as functional characteristics of the IPS indicate that this region is central for dynamic aspects of spatial attention and for the selection of environmental stimuli. Thus, the IPS contains several interacting spatial maps supporting coordinate transformations that are important for the planning of goal-directed eye and hand movements (Colby & Goldberg, 1999). Neurons in the lateral intraparietal area (LIP) and in neighbouring regions of the monkey brain exhibit characteristic enhancement of activity in tasks requiring a shift of attention to peripheral visual stimuli (Bushnell et al., 1981; Cohen et al., 2004).

R. Ptak, A. Schnider / Neuropsychologia 49 (2011) 3063–3070

Importantly, population activity of LIP neurons predicts the locus of attention, be it captured by a task-irrelevant, abrupt-onset distracter, or voluntarily directed to a visual target (Bisley & Goldberg, 2003). LIP is therefore critically involved in voluntary and stimulusdriven shifts of attention. A further important feature of the IPS is its capacity to bias stimulus representations in downstream areas, such as area MT (Saalmann, Pigarev, & Vidyasagar, 2007). Finally, one of the most important properties of neurons lying in the IPS and adjacent regions is their capacity to represent the sensory quality as well as the behavioural relevance of environmental stimuli (Constantinidis & Steinmetz, 2001; Gottlieb et al., 1998). This property indicates that the IPS is a zone of convergence for low-level sensory and motivational inputs, which computes a saliency map of the environment by integrating these inputs (Gottlieb, 2007; Mesulam, 2002; Ptak, 2011). The general importance of the IPS for the representation and selection of environmental stimuli is also supported by the frequent finding of blood-flow increases in this area when participants shift attention voluntarily or reflexively (Hopfinger et al., 2000; Kastner et al., 1999; Mort et al., 2003b; Yantis et al., 2002). Our observation that damage to the IPS is a predictor of attentional deficits associated with neglect is therefore in line with the central role of this region for the selection of environmental stimuli based on their sensory saliency and behavioural priority. The third region predicting impairments of attention in spatial neglect was the dorsal premotor cortex. We found that the relevance effect was negatively correlated to damage to the FEF; damage to this region was thus associated with a particularly small relevance effect. This finding was expected based on the results of a previous whole-brain analysis on the same data set (Ptak & Schnider, 2010) and virtual lesion studies showing that inhibition of the FEF affects the discriminability of targets and nontargets as well as the speed of target detection (Grosbras & Paus, 2002; Muggleton, Juan, Cowey, & Walsh, 2003). In addition, we now also found that the MFG is a global predictor of the occurrence of spatial neglect. The MFG ROI lies in dorsal premotor area 8, whose impairment has occasionally – though not systematically – been reported in spatial neglect (Halligan et al., 2003; Husain & Kennard, 1997), suggesting that it might possibly be associated with particular forms of neglect. Based on findings of functional connectivity analyses (Corbetta et al., 2008; He et al., 2007) argued that the MFG constitutes a link between a dorsal (including the IPS and the FEF) and a ventral (including the TPJ and ventral prefrontal cortex) attention network. The fact that MFG damage, together with pIPS and TPJ damage was a predictor of spatial neglect is compatible with this finding, though our previous results (Ptak & Schnider, 2010) more specifically suggest that the premotor cortex modulates spatial attention in compliance with behavioural goals and expectancies. Several findings show, that neglect is strongly affected by top-down expectancies and behavioural goals. Thus, focusing attention on a specific feature affects the degree of contralesional extinction (Baylis, Driver, & Rafal, 1993; Ptak, Valenza, & Schnider, 2002) and modulates the attention-capturing effect of ipsilesional distracters in the spatial cueing task (Ptak & Schnider, 2006). It is important to note that in the present study patients were instructed to ignore all cues – irrespective of their resemblance to the target. The fact that ipsilesional cues nevertheless captured attention and slowed reactions to contralesional targets indicates that the attentional set of the patient affects performance in a purely reflexive manner (Folk, Remington, & Johnston, 1992). Our finding that the dorsal premotor cortex is crucial for this function suggests that this region constitutes a system whose role it is to bias the selection of environmental stimuli in compliance with current behavioural goals. In support of a network model of spatial neglect the cortical regions considered in this study are richly interconnected. Thus, LIP has heavy projections to the FEF and adjacent premotor cortex

3069

(Brodmann areas 6 and 8, Averbeck, Battaglia-Mayer, Guglielmo, & Caminiti, 2009; Schall, Morel, King, & Bullier, 1995), which may explain the frequent coactivations of these regions in functional imaging studies of spatial attention. Studies on monkeys have shown that projections between the FEF and the angular gyrus travel through the superior longitudinal fasciculus (SLF), which is the main association fibre tract linking parietal with frontal cortex (Makris et al., 2005; Schmahmann & Pandya, 2006). Interestingly, several studies suggested that damage to the SLF might be more predictive of spatial neglect than cortical damage (Bartolomeo et al., 2007; Doricchi & Tomaiuolo, 2003; Thiebaut de Schotten et al., 2005). However, the SLF is not systematically involved in all neglect patients, but particularly in those patients in whom relevant or irrelevant stimuli capture attention to a similar degree (Ptak & Schnider, 2010). This pathway might therefore convey information about the relevance of environmental stimuli elaborated in the dorsal premotor cortex to regions around the TPJ. Task-relevant signals are also exchanged between the dorsal premotor cortex and the IPS, as suggested by the association of both regions with the relevance effect. Finally, the inferior and superior parietal lobe have rich local interconnections providing the anatomical basis for interactions between a specialized system for the shifting of attention toward task-relevant stimuli (located in the TPJ) and a saliency map of the environment (located in the IPS). In conclusion, our results complement functional imaging and neurophysiological studies of spatial attention by showing that the IPS, dorsal premotor cortex and TPJ are associated with neglect, and that interactions within a network formed by these regions predict the performance of neglect patients in a task measuring dynamic aspects of spatial attention. Acknowledgements This work was supported by the Swiss National Science Foundation (grant 320030-134591) and the de Reuter Foundation (grant 458). We thank L. Golay for testing some of the participants of this study. References Averbeck, B. B., Battaglia-Mayer, A., Guglielmo, C., & Caminiti, R. (2009). Statistical analysis of parieto-frontal cognitive-motor networks. Journal of Neurophysiology, 102(3), 1911–1920. Bartolomeo, P., Siéroff, E., Decaix, C., & Chokron, S. (2001). Modulating the attentional bias in unilateral neglect: the effects of the strategic set. Experimental Brain Research, 137, 432–444. Bartolomeo, P., Thiebaut de Schotten, M., & Doricchi, F. (2007). Left unilateral neglect as a disconnection syndrome. Cerebral Cortex, 17, 2479–2490. Bates, E., Wilson, S. M., Saygin, A. P., Dick, F., Sereno, M. I., Knight, R. T., et al. (2003). Voxel-based lesion-symptom mapping. Nature Neuroscience, 6(5), 448–450. Baylis, G. C., Driver, J., & Rafal, R. D. (1993). Visual extinction and stimulus repetition. Journal of Cognitive Neuroscience, 5(4), 453–466. Bichot, N. P., & Schall, J. D. (1999). Effects of similarity and history on neural mechanisms of visual selection. Nature Neuroscience, 2(6), 549–554. Binder, J., Marshall, R., Lazar, R., Benhamin, J., & Mohr, J. P. (1992). Distinct syndromes of hemineglect. Archives of Neurology, 49(11), 1187–1194. Bisley, J. W., & Goldberg, M. E. (2003). Neuronal activity in the Lateral Intraparietal Area and spatial attention. Science, 299(January), 81–86. Brett, M., Leff, A. P., Rorden, C., & Ashburner, J. (2001). Spatial normalization of brain images with focal lesions using cost function masking. NeuroImage, 14(2), 486–500. Bruce, C. J., & Goldberg, M. E. (1985). Primate frontal eye fields. I. Single neurons discharging before saccades. Journal of Neurophysiology, 53(3), 603–635. Bushnell, M. C., Goldberg, M. E., & Robinson, D. L. (1981). Behavioral enhancement of visual responses in monkey cerebral cortex. I. Modulation in posterior parietal cortex related to selective visual attention. Journal of Neurophysiology, 46(4), 755–772. Cohen, Y. E., Cohen, I. S., & Gifford, G. W. (2004). Modulation of LIP activity by predictive auditory and visual cues. Cerebral Cortex, 14, 1287–1301. Colby, C. L., & Goldberg, M. E. (1999). Space and attention in parietal cortex. Annual Review of Neuroscience, 22, 319–349. Constantinidis, C., & Steinmetz, M. A. (2001). Neuronal responses in area 7a to multiple-stimulus displays. I. Neurons encode the location of the salient stimulus. Cerebral Cortex, 11, 581–591.

3070

R. Ptak, A. Schnider / Neuropsychologia 49 (2011) 3063–3070

Corbetta, M., Akbudak, E., Conturo, T. E., Snyder, A. Z., Ollinger, J. M., Drury, H. A., et al. (1998). A common network of functional areas for attention and eye movements. Neuron, 21, 761–773. 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 Neuroscience, 3(March), 292–297. Corbetta, M., Kincade, M. J., Lewis, C., Snyder, A. Z., & Sapir, A. (2005). Neural basis and recovery of spatial attention deficits in spatial neglect. Nature Neuroscience, 8(11), 1603–1610. Corbetta, M., Patel, G., & Shulman, G. L. (2008). The reorienting system of the human brain: from environment to theory of mind. Neuron, 58, 306–324. Corbetta, M., & Shulman, G. L. (2002). Control of goal-directed and stimulus-driven attention in the brain. Nature Reviews: Neuroscience, 3(March), 201–215. Doricchi, F., & Tomaiuolo, F. (2003). The anatomy of neglect without hemianopia: a key role for parietal–frontal disconnection? NeuroReport, 14(17), 2239–2243. Eriksen, C. W., & Hoffman, J. E. (1972). Temporal and spatial characteristics of selective encoding from visual displays. Perception & Psychophysics, 12, 201–204. Fecteau, J. H., & Munoz, D. P. (2006). Salience, relevance, and firing: a priority map for target selection. Trends in Cognitive Sciences, 10(8), 382–390. Folk, C. L., Remington, R. W., & Johnston, J. C. (1992). Involuntary covert orienting is contingent on attentional control settings. Journal of Experimental Psychology: Human Perception and Performance, 18, 1030–1044. Friedrich, F. J., Egly, R., Rafal, R. D., & Beck, D. (1998). Spatial attention deficits in humans: A comparison of superior parietal and temporal–parietal junction lesions. Neuropsychology, 12(2), 193–207. Gauthier, L., Dehaut, F., & Joanette, Y. (1989). The Bells Test: A quantative and qualitative test for visual neglect. International Journal of Clinical Neuropsychology, 11, 49–54. Gitelman, D. R., Nobre, A. C., Parrish, T. B., LaBar, K. S., Kim, Y.-H., Meyer, J. R., et al. (1999). A large-scale distributed network for covert spatial attention. Brain, 122, 1093–1106. Golay, L., Hauert, C. A., Greber, C., Schnider, A., & Ptak, R. (2005). Dynamic modulation of visual detection by auditory cues in spatial neglect. Neuropsychologia, 43(9), 1258–1265. Golay, L., Schnider, A., & Ptak, R. (2008). Cortical and subcortical anatomy of chronic spatial neglect following vascular damage. Behavioral and Brain Functions, 4, 43. Gottlieb, J. (2007). From thought to action: the parietal cortex as a bridge between perception, action, and cognition. Neuron, 53, 9–16. Gottlieb, J., Kusunoki, M., & Goldberg, M. E. (1998). The representation of visual salience in monkey parietal cortex. Nature, 391(January), 481–484. Grosbras, M.-H., & Paus, T. (2002). Transcranial magnetic stimulation of the human frontal eye field: effects on visual perception and attention. Journal of Cognitive Neuroscience, 14(7), 1109–1120. Halligan, P. W., Fink, G. R., Marshall, J. C., & Vallar, G. (2003). Spatial cognition: Evidence from visual neglect. Trends in Cognitive Sciences, 7(3), 125–133. He, B. J., Snyder, A. Z., Vincent, J. L., Epstein, A., Shulman, G. L., & Corbetta, M. (2007). Breakdown of functional connectivity in frontoparietal networks underlies behavioral deficits in spatial neglect. Neuron, 53, 905–918. Hillis, A. E., Newhart, M., Heidler, J., Barker, P. B., Herskovits, E. H., & Degaonkar, M. (2005). Anatomy of spatial attention: Insights from perfusion imaging and hemispatial neglect in acute stroke. Journal of Neuroscience, 25(12), 3161–3167. Hopfinger, J. B., Buonocure, M. H., & Mangun, G. R. (2000). The neural mechanisms of top-down attentional control. Nature Neuroscience, 3(3), 284–291. Husain, M., & Kennard, C. (1997). Distractor-dependent frontal neglect. Neuropsychologia, 35(6), 829–841. Indovina, I., & Macaluso, E. (2007). Dissociation of stimulus relevance and saliency factors during shifts of visuospatial attention. Cerebral Cortex, 17, 1701–1711. Karnath, H. O., Fruhmann Berger, M., Küker, W., & Rorden, C. (2004). The anatomy of spatial neglect based on voxelwise statistical analysis: A study of 140 patients. Cerebral Cortex, 14, 1164–1172. Kastner, S., Pinsk, M. A., De Weerd, P., Desimone, R., & Ungerleider, L. G. (1999). Increased activity in human visual cortex during directed attention in the absence of visual stimulation. Neuron, 22(4), 751–761. Kerkhoff, G. (2001). Spatial hemineglect in humans. Progress in Neurobiology, 63, 1–27. Kimberg, D. Y., Coslett, H. B., & Schwartz, M. F. (2007). Power in voxel-based lesionsymptom mapping. Journal of Cognitive Neuroscience, 19(7), 1067–1080. Kincade, J. M., Abrams, R. A., Astafiev, S. V., Shulman, G. L., & Corbetta, M. (2005). An event-related functional magnetic resonance imaging study of voluntary and stimulus-driven orienting of attention. Journal of Neuroscience, 25(18), 4593–4604. Losier, B. J. W., & Klein, R. M. (2001). A review of the evidence for a disengage deficit following parietal lobe damage. Neuroscience and Biobehavioral Reviews, 25, 1–13. Makris, N., Kennedy, D. N., McInerney, S., Sorensen, A. G., Wang, R., Caviness, V. S., et al. (2005). Segmentation of subcomponents within the superior longitudinal fascicle in humans: A quantitative, in vivo, DT-MRI study. Cerebral Cortex, 15, 854–869. Mesulam, M. M. (1990). Large-scale neurocognitive networks and distributed processing for attention. Language and Memory. Annals of Neurology, 28, 597–613. Mesulam, M. M. (2002). Functional anatomy of attention and neglect: From neurons to networks. In H.-O. Karnath, D. Milner, & G. Vallar (Eds.), The cognitive and neural bases of spatial neglect (pp. 33–45). Oxford: Oxford University Press.

Milner, A. D., & McIntosh, R. D. (2005). The neurological basis of visual neglect. Current Opinion in Neurology, 18, 748–753. Morrow, L. A., & Ratcliff, G. (1988). The disengagement of covert attention and the neglect syndrome. Psychobiology, 16(3), 261–269. Mort, D. J., Malhotra, P., Mannan, S. K., Rorden, C., Pambakian, A., Kennard, C., et al. (2003). The anatomy of visual neglect. Brain, 126, 1986–1997. Mort, D. J., Perry, R. J., Mannan, S. K., Hodgson, T. L., Anderson, E., Quest, R., et al. (2003). Differential cortical activation during voluntary and reflexive saccades in man. Neuroimage, 18(2), 231–246. Muggleton, N. G., Juan, C.-H., Cowey, A., & Walsh, V. (2003). Human frontal eye fields and visual search. Journal of Neurophysiology, 89, 3340–3343. Natale, E., Marzi, C. A., & Macaluso, E. (2010). Right temporal–parietal junction engagement during spatial reorienting does not depend on strategic attention control. Neuropsychologia, 48(4), 1160–1164. Paus, T. (1996). Location and function of the human frontal eye-field: A selective review. Neuropsychologia, 34(6), 475–483. Perry, R. J., & Zeki, S. (2000). The neurology of saccades and covert shifts in spatial attention. Brain, 123, 2273–2288. Pierrot-Deseilligny, C., Ploner, C. J., Müri, R. M., Gaymard, B., & Rivaud-Péchaux, S. (2002). Effects of cortical lesions on saccadic eye movements in humans. Annals of the New York Academy of Sciences, 956, 216–229. Posner, M. I. (1980). Orienting of attention. Quarterly Journal of Experimental Psychology, 32, 3–25. Posner, M. I., & Petersen, S. E. (1990). The attention system of the human brain. Annual Review of Neuroscience, 13, 25–42. Posner, M. I., Walker, J. A., Friedrich, F. A., & Rafal, R. D. (1987). How do the parietal lobes direct covert attention? Neuropsychologia, 25(1A), 135–145. Ptak, R. (2011). The frontoparietal attention network of the human brain: action, saliency, and a priority map of the environment. Neuroscientist, doi:10.1177/1073858411409051 Ptak, R., & Schnider, A. (2006). Reflexive orienting in spatial neglect is biased towards behaviourally salient stimuli. Cerebral Cortex, 16, 337–345. Ptak, R., & Schnider, A. (2010). The dorsal attention network mediates orienting toward behaviorally relevant stimuli in spatial neglect. Journal of Neuroscience, 30(38), 12557–12565. Ptak, R., & Valenza, N. (2005). The inferior temporal lobe mediates distracterresistant visual search of patients with spatial neglect. Journal of Cognitive Neuroscience, 17(5), 788–799. Ptak, R., Valenza, N., & Schnider, A. (2002). Expectation-based attentional modulation of visual extinction in spatial neglect. Neuropsychologia, 40, 2199–2205. Rorden, C., Karnath, H.-O., & Bonilha, L. (2007). Improving lesion-symptom mapping. Journal of Cognitive Neuroscience, 19(7), 1081–1088. Saalmann, Y. B., Pigarev, I. N., & Vidyasagar, T. R. (2007). Neural mechanisms of visual attention: How top-down feedback highlights relevant locations. Science, 316(June), 1612–1615. Schall, J. D., Morel, A., King, D. J., & Bullier, J. (1995). Topography of visual cortex connections with frontal eye field in macaque: Convergence and segregation of processing streams. Journal of Neuroscience, 15(6), 4464–4487. Schenkenberg, T., Bradford, D. C., & Ajax, E. T. (1980). Line bisection and unilateral visual neglect in patients with neurologic impairment. Neurology, 30, 509–517. Schmahmann, J. D., & Pandya, D. N. (2006). Fiber pathways of the brain. Oxford: Oxford University Press. Shinoura, N., Suzuki, Y., Yamada, R., Tabei, Y., Saito, K., & Yagi, K. (2009). Damage to the right superior longitudinal fasciculus in the inferior parietal lobe plays a role in spatial neglect. Neuropsychologia, 47(12), 2600–2603. Singh-Curry, V., & Husain, M. (2009). The functional role of the inferior parietal lobe in the dorsal and ventral stream dichotomy. Neuropsychologia, 47(6), 1434–1448. Talairach, T., & Tournoux, P. (1988). Co-planar stereotaxic atlas of the human brain: 3-Dimensional proportional system – An approach to cerebral imaging. New York: Thieme. Thiebaut de Schotten, M., Urbanski, M., Duffau, H., Volle, E., Lévy, R., Dubois, B., et al. (2005). Direct evidence for a parietal–frontal pathway subserving spatial awareness in humans. Science, 309, 2226–2228. Thompson, K. G., Hanes, D. P., Bichot, N. P., & Schall, J. D. (1996). Perceptual and motor processing stages identified in the activity of macaque frontal eye field neurons during visual search. Journal of Neurophysiology, 76(6), 4040–4055. Urbanski, M., Thiebaut de Schotten, M., Rodrigo, S., Oppenheim, C., Touze, E., Meder, J. F., et al. (2011). DTI-MR tractography of white matter damage in stroke patients with neglect. Experimental Brain Research, 208(4), 491–505. Vallar, G., & Perani, D. (1986). The anatomy of unilateral neglect after righthemisphere stroke lesions. A clinical/CT-scan correlation study in man. Neuropsychologia, 24(5), 609–622. Verdon, V., Schwartz, S., Lovblad, K. O., Hauert, C. A., & Vuilleumier, P. (2010). Neuroanatomy of hemispatial neglect and its functional components: a study using voxel-based lesion-symptom mapping. Brain, 133(Pt 3), 880–894. Vul, E., Harris, C., Winkielman, P., & Pashler, H. (2009). Puzzlingly high correlations in fMRI studies of emotion, personality, and social cognition. Perspectives on Psychological Science, 4(3), 274–290. Wilson, B., Cockburn, J., & Halligan, P. (1987). Behavioural inattention test. Bury St Edmunds: Thames Valley Test Company. Yantis, S., Schwarzbach, J., Serences, J. T., Carlson, R. L., Steinmetz, M. A., Pekar, J. J., et al. (2002). Transient neural activity in human parietal cortex during spatial attention shifts. Nature Neuroscience, 5(10), 995–1002.