Lesion neuroanatomy of the Sustained Attention to Response task

Neuropsychologia 47 (2009) 2866–2875 Contents lists available at ScienceDirect Neuropsychologia journal homepage: www.elsevier.com/locate/neuropsychologia Lesion neuroanatomy of the Sustained Attention to Response task Pascal Molenberghs a , Céline R. Gillebert a , Hanne Schoofs b , Patrick Dupont a , Ronald Peeters c , Rik Vandenberghe a,b,∗ a Cognitive Neurology Laboratory, Experimental Neurology Section, K.U. Leuven, Belgium Neurology Department, University Hospitals Leuven, Herestraat 49, 3000 Leuven, Belgium c Radiology Department, University Hospitals Leuven, Herestraat 49, 3000 Leuven, Belgium b a r t i c l e i n f o Article history: Received 19 February 2009 Received in revised form 20 May 2009 Accepted 14 June 2009 Available online 21 June 2009 Keywords: Inferior frontal Stroke Cognitive control Go no-go a b s t r a c t The Sustained Attention to Response task is a classical neuropsychological test that has been used by many centres to characterize the attentional deficits in traumatic brain injury, ADHD, autism and other disorders. During the SART a random series of digits 1–9 is presented repeatedly and subjects have to respond to each digit (go trial) except the digit ‘3’ (no-go trial). Using voxel-based lesion symptom mapping (VLSM) in a consecutive series of 44 ischemic unifocal non-lacunar hemispheric stroke patients we determined the neuroanatomy of 4 SART parameters: commission and omission error rate, reaction time variability and post-error slowing. Lesions of the right inferior frontal gyrus significantly increased commission error rate. Lesions of the middle third of the right inferior frontal sulcus (IFS) reduced post-error slowing, a measure of how well subjects can utilize errors to adjust cognitive resource allocation. Omissions and reaction time variability had less localising value in our sample. To conclude, commission errors and post-error slowing in the SART mainly probe right inferior frontal integrity. © 2009 Elsevier Ltd. All rights reserved. 1. Introduction The Sustained Attention to Response task (SART) is a conventional neuropsychological test for sustained attention and cognitive control (Robertson, Manly, Andrade, Baddeley, & Yiend, 1997). In the original version sequences of digits 1–9 are visually presented in random order up to 25 times and subjects have to respond to each digit except for the digit ‘3’. The SART differs from typical vigilance tasks in being relatively short and in requiring very frequent responses and rare withholding of responses. It differs from typical go no-go task by the disproportion of go trials over no-go trials. Over the past 15 years, the SART (Robertson, Manly, Andrade, et al., 1997) has been intensively used in children and adults by a number of centres to analyze the attentional deficits arising in frequent disorders such as traumatic brain injury (Chan, 2001; Robertson, Manly, Andrade, et al., 1997; Whyte, Grieb-Neff, Gantz, & Polanksy, 2006), ADHD (Belgrove, Hawi, Gill, & Robertson, 2006; Johnson et al., 2007; Shallice, Marzocchi, Coser, DelSavio, Meuter, & Rumiati, 2002), and autism (Johnson et al., 2007). The SART has been used as an outcome measure for therapy (O’Connell, Bellgrove, Dockree, Lau, Fitzgerald, & Robertson, 2008) and also to define an endophenotype for studying the genetics of attention (Greene, Bellgrove, ∗ Corresponding author at: Neurology Department, University Hospitals Leuven, Herestraat 49, 3000 Leuven, Belgium. Tel.: +32 16 344280; fax: +32 16 3444285. E-mail address: rik.vandenberghe@uz.kuleuven.ac.be (R. Vandenberghe). 0028-3932/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuropsychologia.2009.06.012 Gill, & Robertson, 2009). The neuroanatomy and electrophysiology has been studied in cognitively intact volunteers by means of functional imaging (Fassbender et al., 2004; Manly et al., 2003) and evoked potentials (Zordan, Sarlo, & Stablum, 2008). One element however that is lacking at the moment is a validation in terms of brain lesion neuroanatomy. We prospectively recruited a consecutive series of non-lacunar hemispheric stroke patients independently of lesion site to determine the cognitive neuroanatomy of the SART. In a voxel-based manner we aimed to define which brain regions are critically implicated in the SART using a similar approach as we have applied in previous studies of spatial attentional deficits (Molenberghs, Gillebert, Peeters, & Vandenberghe, 2008). Different behavioral measures that capture different aspects of sustained attention and cognitive control can be derived from the SART. SART omission errors and reaction time variability likely relate to the ability to sustain attention across trials (Stuss, Shallice, Alexander, & Picton, 1995; Wilkins, Shallice, & McCarthy, 1987). Compared to more typical vigilance tasks which have a low target frequency, the SART can pick up lapses of attention in a sensitive way thanks to the high target frequency allowing to monitor the attentional level nearly continuously during the entire task course. These fluctuations will be manifest as increased reaction time variability or, if they are more severe, response omissions. Commission errors in the SART may also reflect a sustained attention problem: when subjects fail to maintain the task goal on-line (Braver, Reynolds, & Donaldson, 2003; Duncan, Emslie, & Williams, 1996), P. Molenberghs et al. / Neuropsychologia 47 (2009) 2866–2875 an automatic response mode driven by the highly frequent go trials may start to dominate behavior and subjects may neglect the task requirement of perceptual identification of the digit prior to a go response decision. Alternatively, commission errors may arise from impaired inhibitory control (Drewe, 1975): subjects may be aware of the task goal but the time needed to identify the stimulus may exceed the period of time during which the prepotent go response can be stopped (Aron, Robbins, & Poldrack, 2004; Logan & Cowan, 1984). Behaviorally, it is challenging to discriminate between these two accounts. Manly et al. (2003) introduced a variant of the SART in which digits are presented in a fixed series of 1–9 rather than in a random sequence (Manly et al., 2003), increasing predictability. The fixed version is more sensitive to frontal damage than the random version (Manly et al., 2003). During the fixed version a commission error is more typically preceded by electrophysiological markers of a failure to maintain the task goal on-line (Braver et al., 2003), such as increased alpha synchronization or a diminished late positivity potential (O’Connell et al., 2008). In contrast, commission errors during the random version of the SART are associated with a diminished inhibitory event-related potential (ERP) complex (O’Connell et al., 2008), indicative of a problem at the level of inhibitory control. A fourth measure of interest, post-error slowing (Fellows & Farah, 2005; Picton et al., 2007; Rabbitt, 1966), reflects the ability to adjust cognitive resources on-line when a mismatch occurs between actual outcome and task goal. Changes in this feedback loop may arise at the level of error evaluation or error utilization (Konow & Pribram, 1970; Scheffers, Coles, Bernstein, Gehring, & Donchin, 1996; Stuss et al., 1995). Alternatively, post-error slowing may also reflect a ‘hangover’ from erroneous processes engaged in the previous trial. From paradigms that resemble the SART we derived a couple of a priori hypotheses. On the go no-go paradigm frontal patients make more commission errors than non-frontal patients (Alexander, Stuss, Picton, Shallice, & Gillingham, 2007; Drewe, 1975; Picton et al., 2007) and show increased response time variability (Stuss, Murphy, Binns, & Alexander, 2003). A related paradigm, the stopsignal reaction time (SSRT) (Logan & Cowan, 1984) paradigm, measures the ability to interrupt a planned motor response. SSRT is increased in patients with lesions of the right frontal operculum (Aron, Fletcher, Bullmore, Sahakian, & Robbins, 2003; Aron et al., 2004). Regions outside prefrontal cortex may also contribute: neglect patients are impaired on tasks that require the detection of unfrequent targets in an auditory stream of distracters (Hjaltason, Tegner, Tham, Levander, & Ericson, 1996; Husain & Rorden, 2003; Robertson, Manly, Beschin, et al., 1997; Samuelsson, Hjelmquist, Jensen, Ekholm, & Blomstrand, 1998), similarly to prefrontal lesion patients (Richer & Lepage, 1996; Wilkins et al., 1987). The attentional dwell time, a measure of cognitive resource allocation, is also increased in neglect patients (Husain, Shapiro, Martin, & Kennard, 1997). Since neglect is most often associated with right inferior parietal lesions, this suggested to us that the right inferior parietal lobule contributes to SART performance in addition to its well-established role in spatial attention (Husain & Rorden, 2003; Mennemeier et al., 1994; Robertson, Manly, Beschin, et al., 1997; Robertson, Mattingley, Rorden, & Driver, 1998; Vandenberghe, Gitelman, Parrish, & Mesulam, 2001). Functional magnetic resonance imaging (fMRI) experiments of cognitive control in healthy volunteers provide an anatomically more refined and more complicated picture than the lesion studies. fMRI studies implicate the junction between the inferior frontal sulcus (IFS) and the inferior precentral sulcus (IFJ) (Derrfuss, Brass, Neumann, & von Cramon, 2005; Hon, Epstein, Owen, & Duncan, 2006; Nakahara, Hayashi, Konishi, & Miyashita, 2002), the middle third of the IFS (Derrfuss et al., 2005; Duncan & Owen, 2000), the frontal operculum extending into the anterior insula (Duncan & Owen, 2000), the dorsolateral prefrontal cortex (McDonald, Cohen, 2867 Stenger, & Carter, 2000; Sohn, Ursu, Anderson, Stenger, & Carter, 2000) and the anterior cingulate (Botvinick, Nystrom, Fissell, Carter, & Cohen, 1999; Carter et al., 1998; Duncan & Owen, 2000; Kerns et al., 2004; McDonald et al., 2000) in different cognitive control processes. fMRI studies of response inhibition have often found the right inferior frontal gyrus (IFG) to be part of an extensive network that includes IFS, anterior cingulate (Konishi et al., 1999; Konishi, Nakajima, Uchida, Sekihara, & Miyashita, 1998) and also the inferior parietal lobule (Bunge, Dudukovic, Thomason, Vaidya, & Gabrieli, 2002; Fassbender et al., 2004; Garavan, Ross, & Stein, 1999), particularly in case of go no-go compared to SSRT tasks (Garavan et al., 1999; Rubia et al., 2001). In the majority of previous lesion studies of cognitive control, lesions were classified using a region-based template (Alexander et al., 2007; Alexander, Stuss, Shallice, Picton, & Gillingham, 2005; Aron et al., 2003, 2004; Drewe, 1975; Fellows & Farah, 2005; Picton et al., 2007; Wilkins et al., 1987). The functional demarcations however within prefrontal cortex cannot be directly derived from structural anatomical boundaries (Stuss et al., 1995). Voxel-based lesion-symptom mapping (VLSM) allows one to quantify a deficit on a continuous scale and relate the degree of impairment to lesion site in a voxelwise manner (Bates et al., 2003; Rorden, Karnath, & Bonilha, 2007) without need for an a priori region-based template. VLSM provides coordinates of critical lesion sites that can be directly compared to those obtained using fMRI in healthy volunteers (Molenberghs et al., 2008). We examined which SART parameters are linked to specific cortical lesion sites (Husain & Rorden, 2003; Robertson, Manly, Andrade, et al., 1997; Robertson et al., 1998). We predicted that apart from prefrontal cortex, the inferior parietal lobule would also be involved (Garavan et al., 1999; Hjaltason et al., 1996; Robertson, Manly, Andrade, et al., 1997; Rubia et al., 2001; Samuelsson et al., 1998). 2. Methods 2.1. Subjects All participants gave written informed consent in accordance with the Declaration of Helsinki. The ethical commission, University Hospitals Leuven, approved the experimental protocol. We screened a consecutive series of 616 ischemic stroke patients during their hospitalization at the stroke unit or on occasion of their first follow-up visit to the outpatient clinic. Fifty-six of these patients fulfilled the study criteria, 44 of whom consented to the study (Table 1). All participants had suffered a recent non-lacunar unifocal ischemic hemispheric stroke, confirmed on clinical Fluid Attenuation Inversion Recovery (FLAIR) or Diffusion-Weighted Imaging (DWI) magnetic resonance imaging (MRI). Patients were recruited via the acute stroke unit (n = 29) or via their first follow-up visit to the outpatient stroke clinic (n = 15). Twenty of the patients also participated in a study of spatial attention reported previously (Molenberghs et al., 2008). Exclusion criteria were age above 85 years, pre-existing periventricular or subcortical white matter lesions or a pre-existing stroke on MRI, insufficient balance to sit autonomously and general inability to understand and carry out a computerized task. The anatomical distribution of the ischemic lesions is shown in Fig. 1. Visual fields were intact except in case 10 (left hemianopia), 14 (right lower quadrantanopia), 15 (left upper quadrantanopia) and 32 (left lower quadrantanopia). A control group of 19 healthy subjects (10 women), aged between 51 and 73 years (mean 61.2 years, S.D. 7.1) underwent the same protocol. 2.2. Stimuli and tasks Subjects were seated at 114 cm distance from a 19 in. computer screen. Using Superlab for PC version 2.0 (Cedrus, Phoenix, AR, USA) each of nine digits was foveally presented 25 times in pseudorandom order over a total period of 4.3 min (Fig. 2). In the version used in the current study the digits were randomized across the entire run rather than per sequences of 1–9. Each digit (luminance = 105 cd/m2 ) was presented for 250 ms on a black background, followed by a 900 ms white mask, after which the next trial started immediately. This resulted in a total task duration of 4 min 19 s. The mask consisted of a ring (diameter = 5.51◦ ; width = 105 cd/m2 ) with a diagonal cross in the middle. The digits were presented in one of five randomly allocated font sizes (48, 72, 94, 100 or 120 points symbol font size; height ranging between 2.28 and 5.51 visual degrees). Subjects were instructed to respond to each digit with a key press (response box from Cedrus, Phoenix, AR, USA) except to 2868 P. Molenberghs et al. / Neuropsychologia 47 (2009) 2866–2875 Table 1 Behavioral parameters Case Age Side Size (cm3 ) Time (days) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 43 82 44 69 53 88 72 65 80 74 73 79 79 47 52 68 64 79 75 74 84 61 37 76 79 65 62 37 42 54 42 64 77 34 66 55 64 61 62 35 60 44 71 80 R R R L L R L R R R L L L L R R R R R R L L L L R R R R R R L R L L L R R L L L R R L R 26.9 20.2 302.7 19.0 108.0 84.1 46.8 17.0 29.8 173.0 16.4 4.8 2.1 13.9 14.3 11.0 216.0 191.0 15.4 117.0 12.5 1.0 11.2 4.0 40.8 49.5 89.7 84.8 43.4 30.2 13.8 197.0 17.2 64.9 95.1 2.6 107.0 18.5 17.0 64.4 29.6 161.0 25.8 64.6 4 5 4 6 4 7 3 5 6 6 4 3 6 5 147 154 5 4 3 7 6 217 21 5 14 10 4 14 6 5 133 196 126 168 126 140 196 7 133 63 168 91 14 126 Bells Bisection (%) L/M/R 0/0/1 1/0/0 1/0/1 0/0/0 3/0/3 2/2/4 2/4/0 2/0/0 0/0/0 14/0/1 0/0/0 2/1/1 0/0/1 0/1/0 2/1/0 0/0/0 15/4/2 15/4/0 2/1/0 0/0/1 0/0/0 0/0/0 0/0/0 1/1/2 2/1/1 1/1/1 2/0/0 2/0/1 4/3/1 2/0/0 2/1/1 2/0/0 0/0/1 0/1/0 1/1/2 1/0/0 3/0/1 0/0/1 0/0/0 0/0/0 1/1/0 0/0/0 1/0/1 3/1/0 +5.8 +3.7 −5.3 +0.7 +4.1 +8.1 −1.7 +5.8 +1.2 +20.5 −0.3 +1.9 +1.7 +3.8 −5.9 −3.0 +18.4 +33.4 +1.2 −1.7 +9.6 +2.0 +0.4 +4.6 −3.8 +4.0 +5.3 +0.9 +18.7 +6.6 +2.1 −5.3 −5.9 3.9 +0.5 −3.6 +4.1 +5.2 +0.4 +0.1 −1.8 +5.0 +2.1 +6.3 Sustained Attention to Response task comm./25 omiss./200 post (ms) mean RT (ms) S.D./mean (ms) 2 5 8 7 12 2 1 2 3 3 6 9 7 3 7 2 8 6 8 1 5 5 7 5 2 0 1 15 8 7 8 17 6 9 6 7 18 12 11 11 2 18 2 1 6 12 2 10 4 3 8 1 17 2 15 5 5 3 20 23 2 8 12 3 10 2 5 2 4 1 5 9 8 7 6 4 1 2 5 5 3 3 1 2 1 8 4 13 105 288 −20 276 −45 31 114 7 94 117 8 62 −37 124 51 79 38 −19 −16 81 −47 21 47 101 94 no errors 10 −54 104 102 −8 38 6 11 318 260 64 103 74 −46 234 1 62 −31 408 526 485 473 580 655 704 414 508 363 625 471 638 647 415 486 539 465 308 581 400 423 346 558 625 501 483 392 661 421 404 491 502 360 389 460 477 422 414 332 533 490 613 521 0.23 0.34 0.17 0.38 0.15 0.15 0.27 0.35 0.34 0.25 0.32 0.21 0.21 0.19 0.25 0.31 0.17 0.20 0.15 0.15 0.20 0.20 0.16 0.16 0.13 0.30 0.14 0.43 0.42 0.31 0.23 0.18 0.23 0.21 0.25 0.36 0.17 0.26 0.35 0.21 0.19 0.46 0.17 0.28 Bold: Significantly different from normals at P < 0.05 (Crawford & Howell, 1998). Legend: side: hemispheric side of the lesion. L: Left. R: Right. M: Middle. Size: Lesion size. time = time to stroke onset. post = post error slowing. comm.: total number of commission errors; omiss.: total number of omission errors. the digit 3. In case of a “3” they had to withhold from responding. Throughout the experiment subjects used their preferred hand. Subjects were asked to give equal importance to accuracy and speed in doing the task. Each subject started with a practice run consisting of 16 go and 2 no-go trials. Our 2 accuracy measures (Table 1) were: 1. Total number of commission errors: Key press responses to a digit “3”. 2. Total number of omission errors: Missed responses to a digit other than 3. Our 2 reaction time (RT) measures (Stuss et al., 2003; Picton et al., 2007) were: 1. Post-error slowing: RT during the trial following a commission error minus that preceding the commission error, divided by the mean RT in that subject. 2. RT variability: Standard deviation of the RTs on correct response trials, divided by the mean RT in that subject Subjects also received a more extensive conventional neuropsychological test battery consisting, among other tests, of target cancellation (Bells test) (Gauthier, Dehaut, & Joanette, 1989), line bisection (Schenkenberg, Bradford, & Ajax, 1980), the number location test from the Visual Object Space Perception Battery (VOSP) (Warrington & James, 1991), Coloured Progressive Matrices (Raven, Court, & Raven, 1995), visual extinction (Bender, 1952), BORB (Riddoch & Humphreys, 1993) and digit span (Wechsler, 1998). 2.3. Image acquisition and preprocessing A 3 T Philips Intera system (Best, Netherlands) equipped with a head volume coil provided T1 images (TR = 1975 ms, TE = 30 ms, in-plane resolution 1 mm) as well as Fluid Attenuation Inversion Recovery (FLAIR) 3D images (TR = 10741 ms, TE = 150 ms) in each patient. Using SPM2 (http://www.fil.ion.ucl.ac.uk, Welcome Trust Centre for Neuroimaging, London, UK) the T1 and FLAIR images were coregistered. The T1 scan was normalized to the Montreal Neurological Institute (MNI) T1 template in Talairach space (Friston et al., 1995; Talairach & Tournoux, 1988). The spatial normalization involved both linear (12 affine transformations) and nonlinear (7 × 9 × 7 basis functions, 16 reiterations) transformations (Ashburner & Friston, 1999). High regularization was used to constrain the nonlinear part of the algorithm and penalize unlikely deformations associated with the presence of lesions (Ashburner & Friston, 1999; Tyler, Marslen-Wilson, & Stamatakis, 2005). The same normalization matrix was applied to the FLAIR images. The match between each patient’s normalized brain and the brain template was carefully evaluated through visual inspection and use of a cross-hair yoked between the template image and the normalized image. After verification of the normalization, lesions were semi-automatically delineated using MRIcro version 1.37 (http://www.sph.sc.edu/comd/rorden/mricro.html) and manually set intensity thresholds (Rorden et al., 2007). Subsequently the lesion volumes were imported into the MRIcron lesion-symptom mapping software (http://www.sph.sc.edu/comd/rorden/mricron). A voxel was included in the analysis only if it was lesioned in at least 4 of the subjects. P. Molenberghs et al. / Neuropsychologia 47 (2009) 2866–2875 2869 Fig. 1. Total group. Lesion distribution volume. The color code indicates in how many individuals of our patient sample (n = 44) a given voxel was lesioned (ranging from 1 to 13). (A) Transverse sections. (B) Projection on the Population-Average, Landmark and Surface based (PALS) (van Essen, 2005) atlas using Computerized Anatomical Reconstruction and Editing Toolkit (sideview and ventral view) (CARET, Washington University School of Medicine, Department of Anatomy and Neurobiology, http://brainmap.wustl.edu(van Essen et al., 2001)). 2.4. Statistical analysis 3. Results 2.4.1. Factor analysis of neuropsychological test scores The neuropsychological data were subjected to a principal components factor analysis with orthogonal rotation (Statistical Package for the Social Sciences, http://www.spss.com/spss/). We included the 4 SART parameters (commission error rate, omission error rate, post-error slowing and within-subject RT variability) as well as contra- versus ipsilesional omissions on the Bells test, mean percentage deviation on the line bisection (ipsilesional deviation being coded positively), and the score on the number location test of the VOSP (Table 1). 3.1. Factor analysis 2.4.2. Voxel-based lesion symptom mapping of SART Each of the 4 SART parameters was entered separately into a VLSM analysis (Rorden et al., 2007). We examined which voxels when lesioned were associated with significantly worse scores compared to patients in whom these voxels were intact (Brunner and Munzel t test; Brunner & Munzel, 2000). Our significance threshold was set at P < 0.05, with a Bonferroni correction for the brain search volume (FamilyWise Error correction; Rorden et al., 2007). When VLSM yielded a significant effect in a particular region, we determined the origin of the finding in further detail: we evaluated the average scores in the patient group who had a lesion in that region and the group who did not have a lesion in that region. Furthermore, we evaluated how many of the individuals with a lesion in that region had a significant deficit at the individual level compared to cognitively intact controls using a modified t test (Crawford & Howell, 1998). 2.4.3. Linear regression analyses We conducted a multiple linear regression with age, lesion extent and time to stroke onset as independent variables and the different SART parameters as dependent variables. We also conducted two simple linear regression analyses with reaction times as independent variable and omission error rate and commission error rate, respectively, as dependent variables. In the total study population, a factor analysis of the neuropsychological data revealed 3 factors with an Eigenvalue higher than 1: the first factor explained 31.0% of the total variance (Eigenvalue = 2.17) and clustered the target cancellation (r = 0.94), line bisection (r = 0.86), and the number location score (r = 0.72). The second factor (Eigenvalue = 1.36) explained 18.7% of the total variance and clustered RT variability (r = 0.84) and omission error rate (r = 0.63). A third factor (Eigenvalue = 1.16) explained 13.9% of the variance and clustered the commission error rate (r = 0.85) and post-error slowing (r = −0.61). 3.2. VLSM of SART When we used commission error rate as input for the VLSM analysis, lesions of the right IFG (centre of mass x = 43, y = 25, z = 14, ext. 16,586 mm3 , P < 0.05) were associated with significantly more commission errors than was seen in patients in whom this region was intact (Fig. 3A). Six patients (cases 3, 17, 28, 32, 37, 42) had a lesion that overlapped with this right IFG region. Their average commission error rate was 14.0 (S.D. 4.77) whereas the average commission error rate in the remainder of the patients was 5.29 (S.D. 3.36), similarly to what we saw in controls (mean 7.6, S.D. 2.53). At the individual level, 4 out of 6 patients with a right IFG lesion (cases 28, 32, 37, 42) had a significantly higher commis- 2870 P. Molenberghs et al. / Neuropsychologia 47 (2009) 2866–2875 Fig. 2. Experimental paradigm. In the main experiment subjects were instructed to respond to each digit with a key press except to the digit 3 (no-go trial). sion error rate than controls (P < 0.05; Crawford & Howell, 1998) (Fig. 4). In order to examine the apparent right-sided laterality we flipped the right IFG VLSM result and examined whether the flipped region belonged to the VLSM lesion distribution volume and how subjects behaved who had a left-sided lesion that overlapped with the flipped region. Four left-hemispheric patients (cases 5, 7, 35, 40) had a lesion that overlapped with the flipped region. The average commission error rate in these 4 patients was 7.5 (S.D. 5.06) closely similar to the normal values. At the individual level, none of the patients with a left IFG lesion showed a pathological commission error rate (Crawford & Howell, 1998). Overall, left-hemispheric lesions (29.3 voxels, S.D. 32.5) were significantly smaller than right-hemispheric lesions (84.7 voxels, S.D. 79.6) (P < 0.01) (Figs. 1 and 5). This could cause a spurious laterality effect in favor of the right hemisphere. For that reason we conducted a VLSM analysis with the left-hemispheric patients only. This did not yield any significant effect in the left IFG, even if we lowered the significance threshold to an uncorrected P < 0.05. A similar analysis with the right-sided lesion patients confirmed the results from the main analysis (corrected P < 0.05). Within the right-hemispheric group, commission error rate tended to correlate with lesion size (Pearson r = 0.376, P = 0.06) but this could not account for the right IFG result since patients with similar-sized lesions sparing the right IFG did not show an increased commission error rate (Fig. 5). When we used post-error slowing as input for the VLSM analysis, lesions of the middle third of the right IFS (centre of mass x = 35 mm, y = 33 mm, z = 27 mm, ext. 718 mm3 , P < 0.05) were associated with a significant decrease in post-error slowing compared to what was seen in patients in whom this region was intact (Fig. 3B). Six patients (cases 3, 17, 18, 28, 32 and 42) had a lesion overlapping with this right IFS lesion. They responded as fast after a commission error (479 ms, S.D. 84) as before (481 ms, S.D. 58) while patients who did not have an IFS lesion showed post-error slowing (post-commission error: 539 ms (S.D. 125), prior to commission error: 463 ms (S.D. 116)) similar in degree to what we saw in the healthy controls (post-error: 477 ms, S.D. 206; prior to error 360 ms, S.D. 47) (Fig. 6). At the individual level, the reduction in post-error slowing did not reach significance in any of the 6 patients compared to controls (Crawford & Howell, 1998). The functional dissociation between the right IFG and the right middle IFS was confirmed when we lowered the threshold of the VLSM analysis to an uncorrected P < 0.001: even at this lower threshold, commission error rate and post-error slowing localised to two nearby but anatomically distinct areas. If we used the difference in reaction times before and after a commission error as VLSM input without dividing by the individual’s mean RT, results were entirely comparable (corrected P < 0.05). At a slightly lower significance threshold, this was also true if the reaction times following a correctly withheld response rather than those preceding a commission error were subtracted from the reaction times following a commission error. VLSM analysis of omission error rate and RT variability did not yield any significant results at the pre-set significance threshold. In order to evaluate whether the absence of any parietal involvement in the SART was due to lack of sensitivity for parietal involvement in the specific patient sample examined, we tested whether a VLSM with the spatial bias measure of the target cancellation task as input yielded the predicted parietal effects (Molenberghs et al., 2008; Mort et al., 2003). Lesions of right inferior parietal cortex extending into the lower bank of IPS were associated with a significantly higher rightward bias than what was seen in the absence of inferior parietal lesions. Our a priori measures of sustained attention did not capture performance changes over time. In order to evaluate whether this explained the absence of parietal involvement, we calculated for each subject the difference between commission error rate in the first 12 no-go trials and in the last 12 no-go trials. Likewise, for the omission error rate and the reaction time variability we calculated the difference between the first 100 go trials and the last 100 go trials. Each of these measures were used as input for VLSM but none of these analyses yielded any inferior parietal effect, even if we lowered the threshold to an uncorrected P < 0.05. Furthermore we performed a linear regression analysis with reaction time as dependent variable and trial position as the independent variable and used the regression coefficient as input for VLSM. No parietal effect was seen, even when the threshold was lowered at an uncorrected P < 0.05. In a final effort to relate any changes in task performance over time to inferior parietal cortex, we divided the trial series into 5 epochs (one epoch per 45 trials) and conducted a linear regression analysis with epoch order as independent variable (1–5) and commission or omission error rates per 45-trials epoch as dependent variable. These coefficients were used as input for VLSM. No significant effects were obtained in IPL. 3.3. Linear regression analyses A multiple linear regression analysis with commission error rate as outcome and with age, lesion size and time-to-stroke onset as independent variables showed a significant effect (F(3, 40) = P. Molenberghs et al. / Neuropsychologia 47 (2009) 2866–2875 2871 Fig. 3. Total group. Lesion-symptom maps thresholded at Bonferroni corr. P < 0.05 (Brunner-Munzel test). (A) VLSM with post-error slowing as input. (B) VLSM with commission error rate as input. 5.39, P < 0.005). Surprisingly, commission error rate correlated significantly and inversely with age (r = −0.33, P = 0.021). The correlation with lesion size (r = 0.27, P = 0.056) and time-to-stroke onset (r = 0.25, P = 0.079) did not reach significance. Post-error slowing, reaction time variability and omission error rate did not correlate with age, lesion size or time-to-stroke onset (F(3, 39) = 0.90, P = 0.45, F(3, 40) = 1.11, P = 0.36 and F(3, 40) = 1.08, P = 0.37, respectively). According to a linear regression analysis with reaction times as independent variable and commission error rate as dependent variable, commission errors tended to be more frequent with faster reaction times (Pearson correlation coefficient −0.30, P = 0.051). Fig. 4. (A–D) Lesions of the four subjects with a pathological commission error rate on the SART. 2872 P. Molenberghs et al. / Neuropsychologia 47 (2009) 2866–2875 of go trials. When however no-go trials had a low probability, as in the original SART version, they failed to withhold their responses. Fig. 5. Regression plot. X-axis: Lesion size. Y-axis: Commission error rate. Legend: Empty squares: left-hemispheric patients. Full circles: Right-hemispheric patients. The datapoints from subjects with a lesion that overlapped with the right IFG result are labelled with their case number. Scores that fall above the dashed line are pathological compared to healthy controls. A similar analysis with omission error rate as dependent variable remained far below significance. 3.4. Ancillary experiments 3.4.1. Commission errors We investigated whether the pathological commission error rate in case of a right-sided IFG lesion was due to the low probability of the ‘no-go’ events, the ‘no-go’ character of those events, or a combination of both. We contacted each of the 7 cases who had a right IFG lesion. Cases 3 and 42 were available for re-testing, 4 and 2 years after the first evaluation, respectively. The tasks were the SART using a randomization scheme per sequence of 1–9 and a variant of the SART in which subjects had to respond to a digit ‘3’ but not to any of the other digits. Subjects first performed the original SART version, subsequently an unrelated visuoperceptual task (45 min) and the SART variant thereafter. On the original SART version cases 3 and 42 made 18 and 15 commission errors, respectively, out of 25 no-go trials, confirming the results obtained in the main study. On the SART variant with proportionally more no-go trials, case 3 made only one commission error out of 200 no-go trials and one omission error out of 25 go trials. Case 42 made one commission error and no omission errors. These subjects therefore were able to correctly withhold responses to nogo trials when the no-go trials were highly expected and were also able to correctly respond to low-probability trials if these consisted Fig. 6. Mean reaction times before and after a commission error in controls, patients with and patients without an IFS lesion. Light gray: Mean reaction time in trials preceding a commission error (and S.D.). Dark gray: Mean reaction time in trials following a commission error (and S.D.). 3.4.2. Post-error slowing We conducted two further variants of the SART to further elucidate the source of the reduced post-error slowing. Cases 3 and 42 each participated in one of these variants, which was performed after another unrelated visuoperceptual task (duration 5 min). Stimuli and task were identical to those of the original SART (i.e. high-probability go trials and low-probability no-go trials) with one exception: we instructed case 3 to pronounce the word ‘hit’ when he had made a commission error to a no-go trial. Case 3 made 22 commission errors and pronounced the word ‘hit’ in 20 of those 22 trials. He also pronounced ‘hit’ to the 3 no-go trials where he correctly withheld responses and never did in go trials. The subject therefore gave proof of knowledge of a distinction between the digit ‘3’ trials and the other ones and was able to use this knowledge to positively respond to the ‘3’ trials in a discriminate way even when he was not able to withhold his response to this type of trial. Finally, in case 42 we provided auditory feedback with a high tone for correct responses and a low tone for errors to examine whether increased error awareness was able to restore post-error slowing. When case 42 was tested on this version, he made 18 commission errors but did not show post-error slowing even when negative feedback was provided explicitly (mean RT pre-error: 545 ms, S.D. 99; mean RT post-error: 493 ms, S.D. 99). 4. Discussion Accurate performance of the SART in the current patient sample relied principally on the integrity of right inferior frontal cortex, in agreement with our first a priori hypothesis. Our data did not provide evidence for our second hypothesis regarding the contribution of the right inferior parietal lobule. The two outcome measures with most localising value were the commission error rate and post-error slowing. Our two other outcome measures, omission error rate and reaction time variability, were not linked to any particular lesion site. The reduction in post-error slowing with IFS lesions cannot be accounted for by a ‘hang-over’ from the erroneous processes engaged in the previous trial. If that were the case, one would expect an increase in post-error slowing in patients compared to controls rather than a decrease. The right-sided laterality must be interpreted with caution. The average lesion size was smaller in left than in right hemisphere patients due to exclusion of patients with significant language comprehension deficits. This may bias the sensitivity of the VLSM analysis in favor of the right hemisphere. When we conducted a VLSM analysis separately in the left and in the right hemisphere patients, this confirmed the right-sided effect in the absence of a significant left IFG effect. Previous studies have also suggested that right IFG lesions may cause more response inhibition problems than left IFG lesions (Aron et al., 2004). Our consecutive series did not include cases of anterior cerebral artery ischemia. Our data therefore are neutral to the possible contribution of anterior cingulate or other medial frontal regions to cognitive control (Alexander et al., 2005, 2007; Fellows & Farah, 2005; Picton et al., 2007). Medial frontal lesions may cause an increase in both omission (Picton et al., 2007) and commission errors during a go-no go task (Drewe, 1975) as well as increased response variability (Picton et al., 2007). According to single neuron recording studies (Schall, Stuphorn, & Brown, 2002), patient lesion studies (Swick & Turken, 2002) and fMRI of the intact brain (Garavan, Ross, Kaufman, & Stein, 2003; Lütcke & Frahm, 2008) the anterior cingulate plays a role in monitoring of conflict (Botvinick P. Molenberghs et al. / Neuropsychologia 47 (2009) 2866–2875 et al., 1999; Carter et al., 1998; McDonald et al., 2000) and error. In a case study of an anterior cingulate lesion, error-related negativity, an event-related potential thought to reflect the activity of a neural system responsible for error detection (Scheffers et al., 1996), was abolished (Swick & Turken, 2002). Inferior parietal lesions were not associated with changes in any of the SART parameters compared to other lesion sites, contrary to our a priori hypothesis (Hjaltason et al., 1996; Robertson, Manly, Beschin, et al., 1997; Samuelsson et al., 1998). In previous studies (Hjaltason et al., 1996; Robertson, Manly, Beschin, et al., 1997; Samuelsson et al., 1998), patients were selected on the basis of the presence of a clinical neglect syndrome, regardless of lesion extent or site. The lesion in individual neglect patients however often extends beyond the parietal lobe. Possibly, the sustained attention deficits seen in neglect patients are due to collateral rightsided inferior frontal damage. Alternatively, if response variability or other measures of sustained attention rely on a brain circuit that includes other areas besides IPL, patients with IPL lesions may not be significantly different on this parameter from patients with lesions of this circuit outside IPL. Conceivably, variations of the current paradigm could modify the exact sustained attention demands increasing the sensitivity of this task for parietal damage but this remains to be tested empirically. For instance, had we reversed the go no-go rule, the need to inhibit automatic responses would have been much reduced and the relative contribution of inferior frontal versus inferior parietal cortex might have been reversed. The SART requires subjects to attend to the identity of the digit rather than its location (Malhotra, Coulthard, & Husain, 2009). Had we placed the digits at peripheral locations and requested the subjects to monitor stimulus location rather digit identity, we would predict that right inferior parietal involvement would have been much stronger: According to a recent lesion study in neglect patients (Malhotra et al., 2009), right IPL lesions impair sustained attention mainly when attention is directed to locations compared to object identity. This is in agreement with a previous fMRI study implicating the right angular gyrus in spatial short-term memory (Vandenberghe et al., 2001). Had we presented the digits at varying peripheral locations and instructed subjects to maintain attention to a selection of these locations, right inferior parietal involvement could have been much more prominent (Malhotra et al., 2009). We did not obtain any significant lesion correlate for intrasubject response variability. This differs from an earlier region-based lesion study implicating the lateral convexity and medial wall of prefrontal cortex in increased individual performance variability (Stuss et al., 2003) during forced-choice simple target detection tasks and even more so during more complex tasks, such as detection of a target defined by feature conjunction (Stuss et al., 2003). In that study only orbitofrontal lesions did not affect RT variability. The effect of prefrontal lesions on RT variability in previous studies was principally based on comparisons with normal control populations (Picton et al., 2007; Stuss et al., 2003). If RT variability can be affected by a relatively wide range of prefrontal regions, it may not be detected by VLSM since VLSM compares the measure of interest between patients who have a lesion in a given voxel to those in whom the voxel is spared: the comparison group, i.e. the patients in whom a given voxel is intact, may also show a deficit of this parameter. Our findings confirm the critical role of the right IFG in inhibitory control (Bunge et al., 2002; Fassbender et al., 2004; Garavan et al., 1999; Konishi et al., 1998, 1999). In the SSRT paradigm, subjects with right inferior opercular lesions are able to abort their responses only when the delay interval between go- and stop-signal is relatively brief compared to controls or patients with lesions elsewhere in prefrontal cortex. The extent of the right inferior frontal lesion correlates positively with SSRT (Aron et al., 2003, 2004). In healthy controls, transcranial magnetic stimulation of the right frontal operculum shortens the delay between go- and stop-signal 2873 that is required to interrupt the go-response, in the absence of any effect upon go trials (Chambers et al., 2006). Because no-go trials are relatively rare in the SART and the intertrial interval is fixed, subjects will have a propensity to respond to each trial. Identification of a “3” may be analogous to a stop signal. In patients with right inferior frontal lesions the time required to identify a stimulus may exceed that needed to interrupt the motor plan, resulting in a commission error. Post-error slowing relies on a reaction time contrast between two temporally juxtaposed trials within a same subject. Post-error slowing can be calculated in a variety of ways, e.g. by subtracting either the reaction time prior to the false-positive trial or by subtracting the reaction time following a correctly withheld response. One can normalize the post-error slowing by mean response time or not. Whichever formula is used, the VLSM results were closely similar. Post-error slowing was significantly reduced in patients with lesions confined to the middle third of the right IFS (Fig. 3), superiorly and anteriorly to the relatively large IFG region implicated in commission errors. In a previous region-based study, post-error slowing was reversed in patients with left dorsolateral prefrontal lesions (more rapid responses following errors) and reduced in patients with right dorsolateral prefrontal lesions (Stuss et al., 2003). Several of the lesions in that study included the right IFS (Stuss et al., 2003). Reduced post-error slowing results from alterations of a feedback loop that encompasses the evaluation of mismatches between actual outcome and task goal and the adaptation of behavioral schemata (Shallice, 1988) on the basis of the mismatch. In a classical case study, a frontal lesion patient was able to verbally report errors he or others made but failed to use these errors to modify his behavior (Konow & Pribram, 1970). A set of follow-up experiments we carried out provides evidence that subjects were aware of the distinction between go and no-go trials. Even when they received feedback, they did not slow their responses, compatible with a deficiency in error utilization (Konow & Pribram, 1970) rather than error evaluation. According to a recent and influential model of the neuroanatomy of spatial and nonspatial attention (Corbetta & Shulman, 2002), a ventral network consisting of the right temporoparietal junction, inferior and middle frontal gyrus, is implicated in stimulus-driven reorienting. This ventral network has been proposed as an alerting system that detects novel stimuli (Corbetta & Shulman, 2002). Our data call for some modification of this model. The contribution of IFS in our study was, strictly speaking, not exogenously driven: it was driven by the error made in response to the no-go stimulus. The reduction of post-error slowing following lesions of the middle third of the right IFS provides strong evidence that this region is involved in on-line readjustment of attentional resources and that the role of this region is not limited to exogenous, stimulus-driven reorienting conditions. To conclude, commission error rate and post-error slowing in the random version of the SART rely in particular on the integrity of the right inferior frontal cortex. We did not find any positive arguments in favor of a contribution of the inferior parietal lobule. Acknowledgements Supported by FWO grants G.0076.02 and G0668.07 (EuroCores) (R.V.), KU Leuven Research grants OT/04/41 and EF/05/014 (R.V.), and Inter-University Attraction Pole P6/29. R.V. is a Clinical Investigator of the Fund for Scientific Research (FW0), Flanders (Belgium), and CRG an FWO research fellow. References Alexander, M., Stuss, D., Picton, T., Shallice, T., & Gillingham, S. (2007). Regional frontal injuries cause distinct impairments in cognitive control. Neurology, 68, 1515–1523. 2874 P. Molenberghs et al. / Neuropsychologia 47 (2009) 2866–2875 Alexander, M., Stuss, D., Shallice, T., Picton, T., & Gillingham, S. (2005). Impaired concentration due to frontal lobe damage from two distinct lesion sites. Neurology, 65, 572–579. Aron, A., Fletcher, P., Bullmore, E., Sahakian, B., & Robbins, T. (2003). Stop-signal inhibition disrupted by damage to right inferior frontal gyrus in humans. Nature Neuroscience, 6, 115–116. Aron, A., Robbins, T., & Poldrack, R. (2004). Inhibition and the right inferior frontal cortex. Trends of Cognitive Neuroscience, 8, 170–177. Ashburner, J., & Friston, K. (1999). Nonlinear spatial normalization using basis functions. Human Brain Mapping, 7, 254–266. 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, 448–450. Belgrove, M., Hawi, Z., Gill, M., & Robertson, I. (2006). The cognitive genetics of attention deficit hyperactivity disorder: Sustained attention as a candidate phenotype. Cortex, 42, 838–845. Bender, M. (1952). Disorders in perception (with particular reference to the phenomena of extinction and displacement). Springfield, IL: C. C. Thomas. Botvinick, M., Nystrom, L., Fissell, K., Carter, C., & Cohen, J. (1999). Conflict monitoring versus selection-for-action in anterior cingulate cortex. Nature, 402, 179–181. Braver, T., Reynolds, J., & Donaldson, D. (2003). Neural mechanisms of transient and sustained cognitive control during task switching. Neuron, 39, 713–726. Brunner, E., & Munzel, U. (2000). The nonparametric Behrens-Fisher problem: Asymptotic theory and a small-sample approximation. Biometric Journal, 1, 17–25. Bunge, S., Dudukovic, N., Thomason, M., Vaidya, C., & Gabrieli, J. (2002). Immature frontal lobe contributions to cognitive control in children: Evidence from fMRI. Neuron, 33, 301–311. Carter, C., Braver, T., Barch, D., Botvinick, M., Noll, D., & Cohen, J. (1998). Anterior cingulate cortex, error detection, and the online monitoring of performance. Science, 280, 747–749. Chambers, C., Bellgrove, M., Stokes, M., Henderson, T., Garavan, H., Robertson, I. ,., et al. (2006). Executive “brake failure” following deactivation of human frontal lobe. Journal of Cognitive Neuroscience, 18, 444–455. Chan, R. (2001). A further study on the sustained attention to response task (SART): The effect of age, gender and education. Brain Injury, 15, 819–829. Corbetta, M., & Shulman, G. (2002). Control of goal-directed and stimulus-driven attention in the brain. Nature Review Neuroscience, 3, 201–215. Crawford, J., & Howell, D. (1998). Comparing an individual’s test score against norms derived from small samples. Clinical Neuropsychology, 12, 482–486. Derrfuss, J., Brass, M., Neumann, J., & von Cramon, D. (2005). Involvement of the inferior frontal junction in cognitive control: Meta-analyses of switching and Stroop studies. Human Brain Mapping, 25, 22–34. Drewe, E. (1975). An experimental investigation of Luria’s theory on the effects of frontal lobe lesions in man. Neuropsychologia, 13, 421–429. Duncan, J., Emslie, H., & Williams, P. (1996). Intelligence and the frontal lobe: The organization of goal-directed behavior. Cognitive Psychology, 30, 257–303. Duncan, J., & Owen, A. M. (2000). Common regions of the human frontal lobe recruited by diverse cognitive demands. Trends in Neurosciences, 23, 475–483. Fassbender, C., Murphy, K., Foxe, J. J., Wylie, G., Javitt, D. C., & Robertson, I. (2004). A topography of executive functions and their interactions revealed by functional magnetic resonance imaging. Cognitive Brain Research, 20(2), 132–143. Fellows, L., & Farah, M. (2005). Is anterior cingulate cortex necessary for cognitive control? Brain, 128, 788–796. Friston, K., Holmes, A., Worsley, K., Poline, J., Frith, C., Heather, J. ,., et al. (1995). Statistical parametric maps in functional imaging: A general approach. Human Brain Mapping, 2, 189–210. Garavan, H., Ross, T., Kaufman, J., & Stein, E. (2003). A midline dissociation between error-processing and response-conflict monitoring. Neuroimage, 20, 1132–1139. Garavan, H., Ross, T., & Stein, E. (1999). Right hemispheric dominance of inhibitory control: An event-related functional MRI study. Proceedings of National Academy of Sciences of United States of America, 96, 8301–8306. Gauthier, L., Dehaut, F., & Joanette, Y. (1989). The bells test: A quantitative and qualitative test for visual neglect. International Journal of Clinical Neuropsychology, 11, 49–54. Greene, C., Bellgrove, M., Gill, M., & Robertson, I. (2009). Noradrenergic genotype predicts lapses in sustained attention. Neuropsychologia, 47, 591–594. Hjaltason, H., Tegner, R., Tham, K., Levander, M., & Ericson, K. (1996). Sustained attention and awareness of disability in chronic neglect. Neuropsychologia, 34(12), 1229–1233. Hon, N., Epstein, R., Owen, A., & Duncan, J. (2006). Frontoparietal activity with minimal decision and control. Journal of Neuroscience, 26, 9805–9809. Husain, M., & Rorden, C. (2003). Non-spatially lateralized mechanisms in hemispatial neglect. Nature Review Neuroscience, 4, 26–36. Husain, M., Shapiro, K., Martin, J., & Kennard, C. (1997). Abnormal temporal dynamics of visual attention in spatial neglect patients. Nature, 385, 154–156. Johnson, K. A., Robertson, I. H., Kelly, S. P., Silk, T. J., Barry, E., Dáibhis, A. ,., et al. (2007). Dissociation in performance of children with adhd and high-functioning autism on a task of sustained attention. Neuropsychologia, 45, 2234–2245. Kerns, J., Cohen, J., MacDonald, A., Cho, R., Stenger, V., & Carter, C. (2004). Anterior cingulate conflict monitoring and adjustments in control. Science, 303, 1023–1026. Konishi, S., Nakajima, K., Uchida, I., Kikyo, H., Kameyama, M., & Miyashita, Y. (1999). Common inhibitory mechanism in human inferior prefrontal cortex revealed by event-related functional mri. Brain, 122, 981–991. Konishi, S., Nakajima, K., Uchida, I., Sekihara, K., & Miyashita, Y. (1998). No-go dominant brain activity in human inferior prefrontal cortex revealed by functional magnetic resonance imaging. European Journal of Neuroscience, 10, 1209–1213. Konow, A., & Pribram, K. (1970). Error recognition and utilization produced by injury to the frontal cortex in man. Neuropsychologia, 8, 489–491. Logan, G., & Cowan, W. (1984). On the ability to inhibit thought and action: A theory of an act of control. Psychological Review, 91, 295–327. Lütcke, H., & Frahm, J. (2008). Lateralized anterior cingulate function during error processing and conflict monitoring as revealed by high-resolution fMRI. Cerebral Cortex, 18, 508–515. Malhotra, P., Coulthard, E., & Husain, M. (2009). Role of right posterior parietal cortex in maintaining attention to spatial locations over time. Brain, 132, 645–660. Manly, T., Owen, A. M., McAvinue, L., Datta, A., Lewis, G. H., Scott, S. K. ,., et al. (2003). Enhancing the sensitivity of a sustained attention task to frontal damage: Convergent clinical and functional imaging evidence. Neurocase, 9(4), 340–349. McDonald, A., Cohen, J., Stenger, V., & Carter, C. (2000). Dissociating the role of the dorsolateral prefrontal and anterior cingulate cortex in cognitive control. Science, 288, 1835–1838. Mennemeier, M., Chatterjee, A., Watson, R., Wertman, E., Carter, L., & Heilman, K. (1994). Contributions of the parietal and frontal lobes to sustained attention and habituation. Neuropsychologia, 32, 703–716. Molenberghs, P., Gillebert, C., Peeters, R., & Vandenberghe, R. (2008). Convergence between lesion-symptom mapping and fmri of spatially selective attention in the intact brain. Journal of Neuroscience, 28, 3359–3373. Mort, D., Malhotra, P., Mannan, S., Rorden, C., Pambakian, A., Kennard, C. ,., et al. (2003). The anatomy of visual neglect. Brain, 126, 1986–1997. Nakahara, K., Hayashi, T., Konishi, S., & Miyashita, Y. (2002). Functional MRI of Macaque monkeys performing a cognitive set-shifting task. Science, 295, 1532–1536. O’Connell, R., Bellgrove, M., Dockree, P., Lau, A., Fitzgerald, M., & Robertson, I. (2008). Self-alert training: Volitional modulation of autonomic arousal improves sustained attention. Neuropsychologia, 46, 1379–1390. Picton, T., Stuss, D., Alexander, M., Shallice, T., Binns, M., & Gillingham, S. (2007). Effects of focal frontal lesions on response inhibition. Cerebral Cortex, 17, 826–838. Rabbitt, P. (1966). Errors and error-correction in choice-response tasks. Journal of Experimental Psychology, 71, 264–272. Raven, J., Court, J., & Raven, J. (1995). Coloured progressive matrices. Oxford: Oxford Psychologists Press. Richer, F., & Lepage, M. (1996). Frontal lesions increase post-target interference in rapid stimulus streams. Neuropsychologia, 34, 509–514. Riddoch, M., & Humphreys, G. (1993). Birmingham object recognition battery. Hove, UK: Lawrence Erlbaum Associates Ltd. Robertson, I., Manly, T., Andrade, J., Baddeley, B., & Yiend, J. (1997). ‘oops!’: Performance correlates of everyday attentional failures in traumatic brain injured and normal subjects. Neuropsychologia, 35(6), 747–758. Robertson, I., Manly, T., Beschin, N., Daini, R., aeske Dewick, H., Homberg, V. ,., et al. (1997). Auditory sustained attention is a marker of unilateral spatial neglect. Neuropsychologia, 35(12), 1527–1532. Robertson, I., Mattingley, J., Rorden, C., & Driver, J. (1998). Phasic alerting of neglect patients overcomes their spatial deficit in visual awareness. Nature, 395, 169–172. Rorden, C., Karnath, H., & Bonilha, L. (2007). Improving lesion-symptom mapping. Journal of Cognitive Neuroscience, 19, 1081–1088. Rubia, K., Russell, T., Overmeyer, S., Brammer, M., Bullmore, E., Sharma, T. ,., et al. (2001). Mapping motor inhibition: Conjunctive brain activations across different versions of go/no-go and stop tasks. NeuroImage, 13, 250–261. Samuelsson, H., Hjelmquist, E., Jensen, C., Ekholm, S., & Blomstrand, C. (1998). Nonlateralized attentional deficits: An important component behind persisting visuospatial neglect? Journal of Clinical and Experimental Neuropsychology, 20, 73–88. Schall, J., Stuphorn, V., & Brown, J. (2002). Monitoring and control of action by the frontal lobes. Neuron, 36, 309–322. Scheffers, M., Coles, M., Bernstein, P., Gehring, W., & Donchin, E. (1996). Event-related brain potentials and error-related processing: An analysis of incorrect responses to go and no-go stimuli. Psychophysiology, 33, 42–53. Schenkenberg, T., Bradford, D., & Ajax, E. (1980). Line bisection and unilateral visual neglect in patients with neurologic impairment. Neurology, 30, 509–517. Shallice, T. (1988). From neuropsychology to mental structure. Cambridge: Cambridge University Press. Shallice, T., Marzocchi, G., Coser, S., DelSavio, M., Meuter, R., & Rumiati, R. (2002). Executive function profile of children with attention deficit hyperactivity disorder. Developmental Neuropsychology, 21, 43–71. Sohn, M., Ursu, S., Anderson, J., Stenger, V., & Carter, C. (2000). The role of prefrontal cortex and posterior parietal cortex in task switching. Proceedings of National Academy of Sciences of United States of America, 97, 13448–13453. Stuss, D., Murphy, K., Binns, M., & Alexander, M. (2003). Staying on the job: The frontal lobes control individual performance variability. Brain, 126, 2363–2380. Stuss, D., Shallice, T., Alexander, M., & Picton, T. (1995). A multidisciplinary approach to anterior attentional functions. Annals of New York Academy of Sciences, 769, 191–212. Swick, D., & Turken, A. (2002). Dissociation between conflict detection and error monitoring in the human anterior cingulate cortex. Proceedings of National Academy of Sciences of United States of America, 99, 16354–16359. Talairach, J., & Tournoux, P. (1988). Co-planar stereotaxic atlas of the human brain. New York: Thieme Medical Publishers, Inc. Tyler, L., Marslen-Wilson, W., & Stamatakis, E. (2005). Dissociating neuro-cognitive component processes: Voxel-based correlational methodology. Neuropsychologia, 43, 771–778. P. Molenberghs et al. / Neuropsychologia 47 (2009) 2866–2875 van Essen, D. (2005). A population-average, landmark- and surface-based (PALS) atlas of human cerebral cortex. Neuroimage, 28, 635–662. van Essen, D., Drury, H., Dickson, J., Harwell, J., Hanlon, D., & Anderson, C. (2001). An integrated software suite for surface-based analyses of cerebral cortex. Journal of American Medical Informatics Association, 8, 443–459. Vandenberghe, R., Gitelman, D., Parrish, T., & Mesulam, M. (2001). Location- or feature-based targeting of peripheral attention. NeuroImage, 14, 34–47. Warrington, E., & James, M. (1991). Visual object and space perception battery. Thamys Valley Test Company Ltd. 2875 Wechsler, D. (1998). Wechsler Memory Scale-III. San Antonio, TX: The Psychological Corporation. Whyte, J., Grieb-Neff, P., Gantz, C., & Polanksy, M. (2006). Measuring sustained attention after traumatic brain injury: Differences in key findings from the sustained attention to response task. Neuropsychologia, 44, 2007–2014. Wilkins, A., Shallice, T., & McCarthy, R. (1987). Frontal lesions and sustained attention. Neuropsychologia, 25, 359–365. Zordan, L., Sarlo, M., & Stablum, F. (2008). ERP components activate by the ‘GO’ and ‘withhold’ conflict in the random SART. Brain Cognition, 66, 57–64.