Does stroke imaging provide insights into the neural basis of cognition?

Paresh A. Malhotra1 and Charlotte Russell2

1 Centre for Restorative Neuroscience, Division of Brain Sciences, Imperial College London, Charing Cross Hospital, Fulham Palace Road, London, W6 8RF, United Kingdom.

2 Department of Psychology, Institute of Psychiatry, Psychology and Neuroscience, King's College London.

Corresponding author: p.malhotra@imperial.ac.uk

Keywords: Stroke, Neuropsychology, Lesion-mapping, Neglect, AttentionWord Count: 3915

Abstract

Since the advent of in vivo imaging, first with CT, and then MRI, structural

neuroimaging in patients has been widely used as a tool to explore the neural

correlates of a wide variety of cognitive functions. Findings from studies using

this methodology have formed a core component of current accounts of

cognition, but there are a number of problematic issues related to inferring

cognitive functions from structural imaging data in stroke and more generally,

lesion-based neuropsychology as a whole. This review addresses these concerns

in the context of spatial neglect, a common disorder most frequently

encountered following right hemisphere stroke. Recent literature, including

attempts to address some of these questions, is discussed. Novel approaches and

findings from related fields that may help to put stroke-based lesion-mapping

studies into perspective are reviewed, allowing critical but constructive

evaluation of previous work in the field.

Introduction

The birth of Neuropsychology and the lesion-method for exploring the

underpinnings of cognition is widely accepted to have occurred with Paul

Broca’s descriptions of two aphasic patients, Louis Victor Leborgne and Lazare

Lelong, in 1861 [1]. Since that time and particularly following the advent of in

vivo imaging, first with Computed Tomography (CT) and then Magnetic

Resonance Imaging (MRI), there has been a vast number of further studies using

lesion anatomy in order to attempt to determine which brain regions are

responsible for particular aspects of behavior. Crucially, these studies have

formed a major component of current accounts of human cognition. For example,

at the time of writing, one such paper examining the neural underpinnings of

sarcasm recognition has received a great deal of media attention [2]. The

majority of these studies have involved adult patients who have suffered from

hemispheric stroke, most often ischaemic and embolic rather than lacunar or

haemorrhagic, and have ranged from single case descriptions to group studies

involving hundreds of patients [3 4]. However, over the last thirty years or so,

researchers have had access to newer methodologies to investigate human

cognition in healthy individuals. In particular, functional imaging, first with

SPECT and PET, then with fMRI, has produced a huge and rapidly increasing

number of cerebral localization studies. In addition, cognitive neuroscience has

exploited ever-improving ERP and EEG recording and analysis, as well as brain

stimulation techniques which augment specific brain functions or induce ‘virtual

lesions’. However, lesion-based studies, most frequently involving stroke

patients, are still used by a large number of researchers to explore the

underpinnings of cognitive processes and remain vitally important. Increasingly

sophisticated imaging techniques, analysis tools and psychophysical measures

have allowed ever more complex examination of the brain-behaviour

relationship in patient studies. However, modern scanning and analysis tools do

not necessarily overcome problematic issues with the lesion-behaviour mapping

approach and in this review we will discuss these concerns, with a focus on the

spatial neglect syndrome, a disorder that is most commonly caused by strokes

affecting the right cerebral hemisphere. The issue of the anatomy of neglect is

one that has been controversial and widely discussed over the last 15 years [4-

12]. This is not an exhaustive review of the subject - instead, the aim is to use

this field to illustrate some of the more general issues pertaining to the lesion-

behaviour method in stroke.

Stroke Imaging and Lesion Mapping

With the introduction of CT scanning in the 1970s, it became possible to map

stroke patients’ lesions in vivo, and correlate lesion anatomy with behavioural

deficits [13 14]. The development of MRI in the 1980s allowed even more refined

lesion mapping [15]. Since that time investigators have tended to use a

combination of both CT and MR images [5 9 16]. On a single case basis, it is

possible to determine specifically which regions are affected in that patient. , and

where many patients manifest the same disorder, lesions can be overlapped onto

a single template to look for a shared neuroanatomical locus for a disorder. More

recently, it has been possible not only to use analysis software to overlap lesions

of patients with a particular disorder, but also to subtract the lesions of

individuals without that disorder, allowing visualization of both the set of

regions involved when a patient suffers from a particular deficit, and

simultaneously those regions affected in those patients without it. By

incorporating such a behavioural control group into analysis, it is possible to

avoid some of the confounds associated with simply overlapping all the patients

who have a specific deficit or behavioural impairment [17].

In parallel with the development of functional imaging analysis, researchers have

refined statistical methods for examining lesion-behaviour relationships. These

techniques interrogate imaging data at a voxel level and produce statistical data

that relate the presence of damage to particular voxels with specific behavioural

deficits [18]. Such voxel-based lesion symptom mapping has been viewed as

overcoming a large number of problems associated with previous

methodologies, especially in studies with larger patient numbers.

In addition to examining lesion data in this way, investigators have also begun to

directly address the importance of white matter damage in cognitive deficits

[19]. Through probabilistic analysis of group data as well as in vivo dissection on

a smaller scale, it has been possible to determine which white matter tracts are

damaged in stroke patients with a particular impairment or syndrome [20 21].

However, it has been argued by some investigators that these approaches remain

fundamentally flawed, and that lesion mapping in patients who have suffered

from ischaemic stroke is intrinsically biased because of the relationship between

lesion anatomy and the architecture of the vascular tree [22]. These authors have

suggested that a more complex approach using machine learning and much

larger patient datasets is the only solution able to rectify this bias. As discussed

below such anatomical considerations are not the only concern when evaluating

neuroimaging studies of cognition following stroke, and we suggest that an

inclusive approach may prove to be most helpful.

Advantages of the Lesion-Behaviour Approach

In contrast to functional imaging methods in the healthy brain, examining lesion-

behaviour relationships allows researchers to identify which brain regions are

absolutely necessary for a particular cognitive process [17]. Functional imaging

in healthy participants, by demonstrating which regions are activated during a

specific task, is essentially correlational and, as such, cannot provide direct

information about which brain regions are essential for the task being carried

out. Experimental methods which disrupt human brain function also have

significant limitations, including a lack of knowledge about precisely which brain

regions are being affected and to what degree any behavioural effects are caused

by inhibition, excitation, or a combination of the two [23]. Moreover, at present

these techniques are limited with respect to the accessibility of brain regions, in

that the only areas that can be reliably and systematically influenced are

relatively close to the scalp. One further method that provides extremely high

quality data is intraoperative stimulation/ recording but this can only be used in

very specific situations in a highly selective group of patients [24].

Spatial Neglect-A Clinical Syndrome

Neglect is defined as an inability to perceive, report and orient to sensory events

occuring contralateral to the lesion and, although it often coexists with a primary

sensory deficit, it need not do so [25]. Patients with neglect appear to ignore

objects, events and individuals on their neglected side and, when asked to draw

or copy a picture, they often produce remarkable images suggesting a lack of

awareness for one side of a scene or object. There are no directly comparable

animal models, and the syndrome has not only attracted the interest of

neurologists and neuropsychologists, but also philosophers and even novelists

[26 27]. It is widely accepted to be primarily a disorder of spatial attention [28-

30] but, as discussed below, it is a heterogeneous entity rather than a unitary

disorder [31]. Some researchers have described anatomical studies of neglect

patients claiming they demonstrate the neural correlates of spatial awareness,

but these claims must be carefully evaluated and put into context [16].

Neglect is a clinical syndrome which can be diagnosed and evaluated with a wide

variety of tasks (See Figure) [31]. Although investigators have suggested a ‘core

deficit’ that consists of an egocentric attentional bias in space, diagnosis and

classification for the purposes of lesion studies is almost always on the basis of

more than one feature as well as general clinical observation[32]. Furthermore,

it should be noted that neglect can be modulated, or even elicited, through the

use of more demanding or complex paradigms such that it may be present on

one version of a task and absent on another version [33 34] For example, if

patients are tested with one type of cancellation task where they have to find

targets on a sheet of paper, they may show normal performance. However a

similar, more challenging, task with targets embedded amongst distractors is

more likely to elicit neglect behavior [35]. Moreover, neglect can vary on a day-

to-day basis and has been observed to vary within individuals on the same day

[36] [37]. Hence, neglect is a complex construct and more than one cognitive

deficit appears to be involved [28 38]. Neglect is worsened by non-spatial deficits

in attention and related processes and it has been shown that even motivational

influences can affect patients’ performance on diagnostic tests [39 40]. Thus,

equating the anatomy of neglect with the anatomy of a single cognitive process is

problematic and it might be more worthwhile to consider the anatomy of

impaired performance on individual neuropsychological tests, as these are

known to dissociate. For example, individuals may continue to manifest neglect

on a line bisection task while performing normally on cancellation tasks and vice

versa [41 42].

However, even single tasks may require multiple cognitive processes and it is,

therefore, necessary to carefully consider the requirements of any individual

test. For example, it has been shown that a non-lateralised deficit in spatial

working memory appears to contribute to impaired performance on cancellation

tasks [43]. Thus, where possible, the use a targeted experimental paradigm is

preferable, as this is likely to probe fewer cognitive domains than standardized

pen-and-paper tests.

The Anatomy of Neglect

Neglect is most frequently observed in patients with moderate to large strokes

affecting the right middle cerebral artery territory. Until the beginning of this

century it was widely accepted that it tended to be a particular consequence of

parietal damage, particularly the inferior parietal lobule (IPL), although there

were reports of neglect occurring following damage to other regions, including

the frontal lobes and subcortical areas [14 44-46]. In 2001, Karnath and

colleagues published a paper, which controversially stated that neglect localized

to the superior temporal gyrus [16]. The result was based on twenty-five

patients who had suffered from ischaemic and hameorrhagic stroke from 1 to

140 days previously and the main analysis excluded all patients with visual field

deficits, as well as those individuals who suffered from subcortical lesions and

those with subcortical damage in regions known to be associated with neglect.

The remaining patients with neglect were compared with control subjects who

had suffered from right hemisphere stroke, but did not manifest neglect (or have

visual deficits or involvement of the subcortical regions previously linked to

neglect). Although the patients were described as having ‘pure’ neglect because

of the exclusion of patients with hemianopia, it should be noted that diagnosis

was made on the basis of clinical observation and the presence of neglect on at

least two out of four diagnostic tests. The finding of a superior temporal locus

associated with neglect was controversial, and since then there has been a great

deal of further work in the field, including by the authors of that paper,

addressing some of the concerns raised with this study as well as more general

issues relating to lesion-mapping in stroke. These issues include the nature of the

behavioural assessments being used, recruitment of appropriate control groups,

chronicity of patient assessment, the utilisation of MRI versus CT (and the use of

different MRI modalities), the importance of white matter damage, and the need

to take a network-based rather than region-based approach to the problem [4 5

7 47-49].

Behavioural Tasks in Neglect

Although neglect can often be diagnosed by observation, this is clearly

dependent to an extent on the clinician and does not allow easy quantification of

neglect severity. It has long been known that impaired performance on the two

main neuropsychological tests used to diagnose and quantify neglect, line

bisection and cancellation tasks, appears to localise to different brain regions

[50]. A number of investigators have carried out further studies in order to

characterize different types of neglect behavior and localize them to specific

brain regions [6 51 52]. In a recent paper, Verdon et al used principal

components analysis to demonstrate that tests examining allocentric, or object-

based neglect (where patients show reduced awareness for the contralesional

side of individual objects, regardless of their location) localized more to deep

temporal regions, whereas more visuo-perceptive impairment related to the IPL

and visuo-motor components to the dorsolateral prefrontal cortex [9]. Meta-

analyses of multiple lesion–mapping studies have also suggested such a

dissociation between egocentric and allocentric neglect [53 54]. Thus, the use of

test batteries to characterize patients with a clinical syndrome may be

particularly unhelpful when considering lesion anatomy. Investigating the neural

correlates of a single task or carrying out a principal components analysis to

determine the underlying factors linking impaired task performance with lesion

location is likely to constitute a more fruitful approach [54].

Chronicity

Up to 82% of right hemisphere stroke patients manifest some degree of neglect

in the acute stage and one third of patients may manifest neglect when tested a

year later [47 55], although it should be noted that this proportion increases if

more sophisticated psychophysical tests are used [56]. One issue that has

confronted researchers is how to address this and whether the neuroanatomy of

chronic neglect is somehow more indicative of the neural correlates of

visuospatial attention [5]. In the acute stage, cognitive deficits may not only be

directly due to the lesion, but there may be related impaired function of nearby,

or distant but functionally connected, brain regions (See below for further

discussion). In addition, there is often acute oedema, such that the effective

lesion size is greater than that visualized by standard MRI. Thus, some authors

have looked only at chronic neglect and others have compared lesion anatomy

between patients who had neglect acutely and then recovered, and patients who

have longstanding chronic neglect [5 47]. However, it should be noted that

plastic changes in those individuals who have recovered can make it difficult to

interpret their lesion anatomy [32].

Scanning Methodology

Although a number of major studies have used CT images to explore the anatomy

of neglect, MRI is the modality of choice [7 14]. Lesion mapping methods using

CT have improved, but it is still the case that CT scans obtained for clinical

purposes tend to be of far lower resolution than MRI, and may be less likely to

demonstrate ischaemic strokes, particularly in the hyperacute stage but also 2 to

4 weeks after the ischaemic event [57]. Therefore, carrying out a voxel level

interrogation of lesion maps derived from such low quality imaging data is not

appropriate. With respect to structural MRI, DWI scans are particularly helpful in

the hyperacute stage and are predictive of eventual lesion size [58]. Again, scans

obtained for clinical purposes may well be of lower resolution, and specific high-

resolution protocols can provide more detailed anatomical information [7].

One further MR modality that has been employed to investigate neglect is

perfusion weighted imaging (PWI). Subcortical regions have long been known to

be associated with neglect, and Hillis and colleagues explored the mechanism

using PWI/DWI mismatch [59]. By examining which areas were hypoperfused,

they were able to show that subcortical strokes caused neglect via the

hypoperfusion of neighbouring cortical areas that were structurally intact, rather

than via direct subcortical damage per se. Thus, without the added use of

perfusion imaging (or other dynamic/functional modalities), interpretation of

structural imaging alone in the acute stage is problematic.

The Importance of White Matter

Although earlier studies had demonstrated that focal white matter damage was

relevant in patients with neglect [46 60], its importance has been increasingly

noted over the last decade following a key study in patients undergoing

intraoperative brain stimulation during surgery and the evolution of methods for

white matter tract delineation [19 24]. It has been now been shown that damage

to the superior longitudinal fasciculus, a structure connecting the frontal and

parietal lobes, is extremely common in neglect patients [20]. In addition, the

importance of the SLF has been highlighted by the occurrence of neglect in two

individuals with iatrogenic SLF damage during neurosurgery [61]. Furthermore,

when other studies that did not include dedicated white matter analysis have

been closely re-examined, similar findings have been demonstrated [10]. A

recent study confirmed the importance of the SLF in acute and chronic neglect

patients, but also highlighted the potential role of damage to the forceps major of

the corpus callosum, a tract that connects the two occipital lobes [11]. Umarova

and colleagues have examined white matter integrity in the unaffected left

hemisphere of neglect cases and found that changes in the superior parietal lobe,

optic radiation and corpus callosum were related to delayed recovery [62]. The

authors of the study suggested that these left hemisphere changes induced white

matter re-modelling, but further work will be needed to determine exactly how

such changes relate to the primary stroke, or whether they in fact pre-date it.

A Network-based approach

The importance of white matter damage as described above highlights the need

to understand cognitive deficits at the level of brain networks rather than single

cortical regions. With respect to spatial neglect, this improved understanding has

come from close interrogation of structural imaging as discussed above and also

from task-related and ‘resting-state’ functional imaging studies in patient groups

[42 [63] [64] [49]. Although authors have consistently attempted to find a critical

lesion site for neglect (see above), it has long been appreciated that it is caused

by damage to brain networks that are responsible for spatial attention and

related processes [45 65]. Evidence from neglect patients, when examined in

tandem with data from functional imaging studies in healthy volunteers,

suggests that neglect arises from the combination of a critical deficit in a spatial

attention network, resulting in an ipsilesional bias, along with damage to other

right-dominant frontoparietal networks responsible for processes including

arousal, salience and spatial working memory [28 38]. By using functional

imaging in neglect patients over the last ten years, investigators have been able

to directly examine whether these networks are affected in neglect and what

changes take place during recovery [49 63]. For example, by using functional

connectivity analysis, researchers have shown that there appears to be an acute

disruption of interparietal connectivity in neglect patients [63] although other

work has suggested that interparietal imbalance is not critical [64]. The most

prevalent current account of neglect stems from this work, utilising results from

both patient studies and functional imaging findings in healthy volunteers. This

suggests that neglect involves structural damage to a ventral right-lateralized

network involved in arousal and directing attention to salient and novel events,

as well as functional disruption of a dorsal attention network responsible for the

deployment of spatial attention [28]. Evidence from stimulation work may be

critical here, as it has shown that the spatial bias in neglect can be improved by

systematically influencing interparietal balance with transcranial direct current

stimulation (tDCS), even to the unlesioned hemisphere [66].

Future Methodologies

As mentioned in the introduction, it has recently been suggested that previous

methods of lesion-mapping are inherently biased due to the propensity of

ischaemic strokes to damage certain sets of voxels because of their location with

respect to the arterial tree [22]. The authors of this study have stated that

correcting this systematic error requires more sophisticated analysis using novel

machine learning techniques in much larger patient numbers than have been

previously been recruited. Two recent studies have employed multivariate

analysis using machine-learning in order to address this issue [67] [48]. The first

of these, by Smith and colleagues, demonstrated that multi-region patterns of

damage were more predictive of neglect than any single voxel, but that damage

to the superior temporal gyrus seemed particularly predictive. However, it

should be noted that this study utilized data from a previous paper, which again

used a combination of neuropsychological tests and clinical observation to

classify patients into neglect and non-neglect groups [4 48]. Moreover, it has

been suggested that the large region-of-interest method that was used does not

overcome the previously described confounds associated with vascular anatomy

[68].

In a second study, Corbetta and colleagues did not just address neglect but

looked at multiple cognitive domains across a relatively large group of patients

presenting with a first stroke. They found that the majority of lesions were

subcortical, affecting white matter [67]. As per the multivariate study by Karnath

et al, each deficit was associated with damage to multiple regions but - consistent

with the concept of neglect as a multi-component syndrome - three component

factors were recognised (deficits in lateralised spatial attention, sustained

attention and attentional re-orienting). Regions involved included the white

matter relating to the superior longitudinal fasciculus, the inferior frontal gyrus,

insula, thalamus, and parahippocampal gyrus but not the temporo-parietal

junction, although it should be noted that this was not well sampled.

Conclusions

A clear conclusion from the number and diversity of the studies reviewed here is

that stroke imaging is indeed a viable and valuable method for exploring the

anatomy of cognitive processes. We have focused on the neglect syndrome but

the issues that have been raised are pertinent to other cognitive domains. The

majority of these studies have employed a lesion-mapping approach with stroke

patients, but it should be noted that this appears to be most effective when

supported by data from other methodologies including functional imaging and

non-invasive brain stimulation. Moreover, thorough neuropsychological testing

and targeted behavioural tasks are critical to a deeper understanding of how

anatomical damage might actually relate to cognitive deficits. Advanced imaging

and analysis methodology can only be helpful if combined with considered

behavioural paradigms and appropriate control groups.

Although the lesion method of stroke neuroimaging can be improved to some

extent by more sophisticated analysis, there remain a number of concerns that

cannot easily be overcome. These include the presence of concurrent small

vessel disease and atrophy, both of which have been shown to influence

performance [69 70]. Genotype is a factor that has been shown to influence

attentional asymmetry in healthy populations, but has not been systematically

explored in the context of cognitive disorders following stroke [71]. In addition,

it should be noted that strokes are not the only cause of focal pathology in the

brain. For a comprehensive understanding of the neural basis of cognition, a

more inclusive position may be the most fruitful approach [72]. As well as using

multiple imaging modalities and methodologies, and incorporating data from

techniques such as brain stimulation, it is essential to consider relevant findings

from other patient groups, including individuals with focal dementias as well as

patients with brain tumours [73-75]. Although neuroimaging in these

populations is also associated with a number of confounds, these are largely

different to those associated with focal stroke lesion-mapping. In fact, a recent

study looking at frontal damage across a number of pathologies suggested that

there was no major difference in the cognitive performance of patients with

damage caused by different aetiologies, after accounting for the effects of age and

premorbid IQ [76]. Thus, an inclusive approach to the problems associated with

stroke neuroimaging and cognition might allow us to consolidate previous

findings in the light of current concerns, without ‘throwing out the baby with the

bathwater’.

Figure Legend

Tests used to examine for the presence of neglect.

Panel A shows a standard cancellation task (Behavioural Inattention Test). These

tasks require subjects to search for targets (here the small stars), often amongst

distractors) on a piece of paper. Patients with neglect tend to find the most

ipsilesional targets and omit the more contralesional targets, and often begin by

marking target items on the ipsilesional side of the array. Panel B shows a

neglect patient’s line bisection. Although the patient has been asked to find and

mark the true centre of the line, they have clearly erred to the right. As well as

these pen-and-paper tests for neglect, patients are often diagnosed by clinical

observation of their general behaviour. Panel C shows how a patient responded

when asked to name objects around him in the room. All the objects are well to

the right of the midline.

Patients with neglect can show deficits on all three of these tasks or can be

diagnosed on the basis of impaired performance on one or two of them. This

behavioural heterogeneity is likely to underlie many of the conflicting findings in

the related neuroimaging research.

References

1. Broca P. Perte de la parole, ramollissement chronique et destruction partielle du lobe antérieur gauche du cerveau. Bulletin de la Société Anthropologique 1861;2:235-38

2. Davis CL, Oishi K, Faria AV, et al. White matter tracts critical for recognition of sarcasm. Neurocase 2015:1-8 doi: 10.1080/13554794.2015.1024137[published Online First: Epub Date]|.

3. Leff AP, Schofield TM, Crinion JT, et al. The left superior temporal gyrus is a shared substrate for auditory short-term memory and speech comprehension: evidence from 210 patients with stroke. Brain 2009;132(Pt 12):3401-10 doi: 10.1093/brain/awp273[published Online First: Epub Date]|.

4. Karnath HO, Fruhmann Berger M, Kuker W, et al. The anatomy of spatial neglect based on voxelwise statistical analysis: a study of 140 patients. Cerebral cortex 2004;14(10):1164-72 doi: 10.1093/cercor/bhh076[published Online First: Epub Date]|.

5. Doricchi F, Tomaiuolo F. The anatomy of neglect without hemianopia: a key role for parietal-frontal disconnection? Neuroreport 2003;14(17):2239-43 doi: 10.1097/01.wnr.0000091132.75061.64[published Online First: Epub Date]|.

6. Hillis AE, Newhart M, Heidler J, et al. Anatomy of spatial attention: insights from perfusion imaging and hemispatial neglect in acute stroke. J Neurosci 2005;25(12):3161-7

7. Mort DJ, Malhotra P, Mannan SK, et al. The anatomy of visual neglect. Brain 2003;126(Pt 9):1986-97

8. Ringman JM, Saver JL, Woolson RF, et al. Frequency, risk factors, anatomy, and course of unilateral neglect in an acute stroke cohort. Neurology 2004;63(3):468-74

9. Verdon V, Schwartz S, Lovblad KO, et al. Neuroanatomy of hemispatial neglect and its functional components: a study using voxel-based lesion-symptom

mapping. Brain 2010;133(Pt 3):880-94 doi: 10.1093/brain/awp305[published Online First: Epub Date]|.

10. Doricchi F, Thiebaut de Schotten M, Tomaiuolo F, et al. White matter (dis)connections and gray matter (dys)functions in visual neglect: gaining insights into the brain networks of spatial awareness. Cortex 2008;44(8):983-95 doi: S0010-9452(08)00114-7 [pii]

10.1016/j.cortex.2008.03.006[published Online First: Epub Date]|.11. Lunven M, Thiebaut De Schotten M, Bourlon C, et al. White matter lesional

predictors of chronic visual neglect: a longitudinal study. Brain : a journal of neurology 2015 doi: 10.1093/brain/awu389[published Online First: Epub Date]|.

12. Urbanski M, Thiebaut de Schotten M, Rodrigo S, et al. Brain networks of spatial awareness: evidence from diffusion tensor imaging tractography. J Neurol Neurosurg Psychiatry 2008;79(5):598-601

13. Naeser MA, Hayward RW. Lesion localization in aphasia with cranial computed tomography and the Boston Diagnostic Aphasia Exam. Neurology 1978;28(6):545-51

14. Vallar G, Perani D. The anatomy of unilateral neglect after right-hemisphere stroke lesions. A clinical/CT-scan correlation study in man. Neuropsychologia 1986;24(5):609-22

15. DeWitt LD, Grek AJ, Buonanno FS, et al. MRI and the study of aphasia. Neurology 1985;35(6):861-5

16. Karnath HO, Ferber S, Himmelbach M. Spatial awareness is a function of the temporal not the posterior parietal lobe. Nature 2001;411(6840):950-3 doi: 10.1038/35082075[published Online First: Epub Date]|.

17. Rorden C, Karnath HO. Using human brain lesions to infer function: a relic from a past era in the fMRI age? Nat Rev Neurosci 2004;5(10):813-9

18. Bates E, Wilson SM, Saygin AP, et al. Voxel-based lesion-symptom mapping. Nat Neurosci 2003;6(5):448-50

19. Catani M. Diffusion tensor magnetic resonance imaging tractography in cognitive disorders. Curr Opin Neurol 2006;19(6):599-606 doi: 10.1097/01.wco.0000247610.44106.3f[published Online First: Epub Date]|.

20. Thiebaut de Schotten M, Tomaiuolo F, Aiello M, et al. Damage to white matter pathways in subacute and chronic spatial neglect: a group study and 2 single-case studies with complete virtual "in vivo" tractography dissection. Cerebral cortex 2014;24(3):691-706 doi: 10.1093/cercor/bhs351[published Online First: Epub Date]|.

21. Forkel SJ, Thiebaut de Schotten M, Dell'Acqua F, et al. Anatomical predictors of aphasia recovery: a tractography study of bilateral perisylvian language networks. Brain 2014;137(Pt 7):2027-39 doi: 10.1093/brain/awu113[published Online First: Epub Date]|.

22. Mah YH, Husain M, Rees G, et al. Human brain lesion-deficit inference remapped. Brain : a journal of neurology 2014;137(Pt 9):2522-31 doi: 10.1093/brain/awu164[published Online First: Epub Date]|. Mah et al used high dimensional analysis to examine lesion data from over 581 patients, and suggesting that the majority of stroke lesion mapping studies to-date are subject to an inherent anatomical bias. All the lesions were

caused by ischaemic stroke, and it is unlikely that the results apply to haemorrhagic lesions or damage caused by other pathology.

23. Bolognini N, Ro T. Transcranial magnetic stimulation: disrupting neural activity to alter and assess brain function. J Neurosci 2010;30(29):9647-50 doi: 10.1523/JNEUROSCI.1990-10.2010[published Online First: Epub Date]|.

24. Thiebaut de Schotten M, Urbanski M, Duffau H, et al. Direct evidence for a parietal-frontal pathway subserving spatial awareness in humans. Science 2005;309(5744):2226-8

25. Heilman KM, Watson RT, Valenstein E. Neglect and related disorders. In: Heilman KM, Valenstein E, eds. Clinical Neuropsychology. 4th ed. New York: OUP, 2003:296-346.

26. Churchland PS. Brain-wise : studies in neurophilosophy. Cambridge, Mass.: MIT Press, 2002.

27. Genova L. Left neglected : a novel. 1st Gallery Books hardcover ed. New York: Gallery Books, 2011.

28. Corbetta M, Shulman GL. Spatial neglect and attention networks. Annu Rev Neurosci 2011;34:569-99 doi: 10.1146/annurev-neuro-061010-113731[published Online First: Epub Date]|.

29. Heilman KM, Van Den Abell T. Right hemisphere dominance for attention: the mechanism underlying hemispheric asymmetries of inattention (neglect). Neurology 1980;30(3):327-30

30. Mesulam MM. Spatial attention and neglect: parietal, frontal and cingulate contributions to the mental representation and attentional targeting of salient extrapersonal events. Philosophical transactions of the Royal Society of London. Series B, Biological sciences 1999;354(1387):1325-46 doi: 10.1098/rstb.1999.0482[published Online First: Epub Date]|.

31. Vuilleumier P. Mapping the functional neuroanatomy of spatial neglect and human parietal lobe functions: progress and challenges. Ann N Y Acad Sci 2013;1296:50-74 doi: 10.1111/nyas.12161[published Online First: Epub Date]|. A clear-sighted review of the field, with an acute appreciation of the problems that beset attempts to map a clinical syndrome to a single brain region.

32. Karnath HO, Rorden C. The anatomy of spatial neglect. Neuropsychologia 2012;50(6):1010-7 doi: 10.1016/j.neuropsychologia.2011.06.027[published Online First: Epub Date]|.

33. Stone SP, Wilson B, Wroot A, et al. The assessment of visuo-spatial neglect after acute stroke. Journal of neurology, neurosurgery, and psychiatry 1991;54(4):345-50

34. Azouvi P, Samuel C, Louis-Dreyfus A, et al. Sensitivity of clinical and behavioural tests of spatial neglect after right hemisphere stroke. Journal of neurology, neurosurgery, and psychiatry 2002;73(2):160-6

35. Gauthier L., F. D, Y J. The bells test: a quantitative and qualitative test for visual neglect. International Journal of Clinical Neuropsychology 1989;11:49-54

36. Jehkonen M, Laihosalo M, Koivisto AM, et al. Fluctuation in spontaneous recovery of left visual neglect: a 1-year follow-up. Eur Neurol 2007;58(4):210-4

37. Bailey MJ, Riddoch MJ, Crome P. Treatment of visual neglect in elderly patients with stroke: a single-subject series using either a scanning and cueing strategy or a left-limb activation strategy. Physical therapy 2002;82(8):782-97

38. Husain M, Rorden C. Non-spatially lateralized mechanisms in hemispatial neglect. Nature reviews. Neuroscience 2003;4(1):26-36 doi: 10.1038/nrn1005[published Online First: Epub Date]|.

39. Malhotra PA, Soto D, Li K, et al. Reward modulates spatial neglect. J Neurol Neurosurg Psychiatry 2013;84(4):366-9 doi: 10.1136/jnnp-2012-303169[published Online First: Epub Date]|. The first empirical demonstration of the effect of anticipated reward on impaired attention in stroke patients. This result suggest that motivational function may have an important role to play in attentional performance following stroke.

40. Mesulam MM. Principles of behavioral neurology. Philadelphia: F.A. Davis, 1985.

41. Na DL, Adair JC, Williamson DJ, et al. Dissociation of sensory-attentional from motor-intentional neglect. Journal of neurology, neurosurgery, and psychiatry 1998;64(3):331-8

42. Bisiach E, Geminiani G, Berti A, et al. Perceptual and premotor factors of unilateral neglect. Neurology 1990;40(8):1278-81

43. Malhotra P, Jager HR, Parton A, et al. Spatial working memory capacity in unilateral neglect. Brain 2005;128(Pt 2):424-35 doi: 10.1093/brain/awh372[published Online First: Epub Date]|.

44. Husain M, Kennard C. Visual neglect associated with frontal lobe infarction. Journal of neurology 1996;243(9):652-7

45. Watson RT, Valenstein E, Heilman KM. Thalamic neglect. Possible role of the medial thalamus and nucleus reticularis in behavior. Archives of neurology 1981;38(8):501-6

46. Leibovitch FS, Black SE, Caldwell CB, et al. Brain-behavior correlations in hemispatial neglect using CT and SPECT: the Sunnybrook Stroke Study. Neurology 1998;50(4):901-8.

47. Karnath HO, Rennig J, Johannsen L, et al. The anatomy underlying acute versus chronic spatial neglect: a longitudinal study. Brain 2011;134(Pt 3):903-12 doi: 10.1093/brain/awq355[published Online First: Epub Date]|.

48. Smith DV, Clithero JA, Rorden C, et al. Decoding the anatomical network of spatial attention. Proceedings of the National Academy of Sciences of the United States of America 2013;110(4):1518-23 doi: 10.1073/pnas.1210126110[published Online First: Epub Date]|.

49. Corbetta M, Kincade MJ, Lewis C, et al. Neural basis and recovery of spatial attention deficits in spatial neglect. Nature neuroscience 2005;8(11):1603-10 doi: 10.1038/nn1574[published Online First: Epub Date]|.

50. Binder J, Marshall R, Lazar R, et al. Distinct syndromes of hemineglect. Arch Neurol 1992;49(11):1187-94

51. Rorden C, Fruhmann Berger M, Karnath HO. Disturbed line bisection is associated with posterior brain lesions. Brain Res 2006;1080(1):17-25 doi: 10.1016/j.brainres.2004.10.071[published Online First: Epub Date]|.

52. Committeri G, Pitzalis S, Galati G, et al. Neural bases of personal and extrapersonal neglect in humans. Brain : a journal of neurology 2007;130(Pt 2):431-41 doi: 10.1093/brain/awl265[published Online First: Epub Date]|.

53. Chechlacz M, Rotshtein P, Humphreys GW. Neuroanatomical Dissections of Unilateral Visual Neglect Symptoms: ALE Meta-Analysis of Lesion-Symptom Mapping. Frontiers in human neuroscience 2012;6:230 doi: 10.3389/fnhum.2012.00230[published Online First: Epub Date]|.

54. Molenberghs P, Sale MV, Mattingley JB. Is there a critical lesion site for unilateral spatial neglect? A meta-analysis using activation likelihood estimation. Frontiers in human neuroscience 2012;6:78 doi: 10.3389/fnhum.2012.00078[published Online First: Epub Date]|.

55. Stone SP, Halligan PW, Greenwood RJ. The incidence of neglect phenomena and related disorders in patients with an acute right or left hemisphere stroke. Age Ageing 1993;22(1):46-52

56. Bonato M. Neglect and extinction depend greatly on task demands: a review. Frontiers in human neuroscience 2012;6:195 doi: 10.3389/fnhum.2012.00195[published Online First: Epub Date]|. An important paper stressing the point that neglect and related deficits may go undiagnosed unless sophisticated targeted behavioural testing is carried out.

57. Becker H, Desch H, Hacker H, et al. CT fogging effect with ischemic cerebral infarcts. Neuroradiology 1979;18(4):185-92

58. Ricci PE, Burdette JH, Elster AD, et al. A comparison of fast spin-echo, fluid-attenuated inversion-recovery, and diffusion-weighted MR imaging in the first 10 days after cerebral infarction. AJNR. American journal of neuroradiology 1999;20(8):1535-42

59. Hillis AE, Wityk RJ, Barker PB, et al. Subcortical aphasia and neglect in acute stroke: the role of cortical hypoperfusion. Brain 2002;125(Pt 5):1094-104

60. Samuelsson H, Jensen C, Ekholm S, et al. Anatomical and neurological correlates of acute and chronic visuospatial neglect following right hemisphere stroke. Cortex 1997;33(2):271-85.

61. Shinoura N, Suzuki Y, Yamada R, et al. Damage to the right superior longitudinal fasciculus in the inferior parietal lobe plays a role in spatial neglect. Neuropsychologia 2009;47(12):2600-3 doi: 10.1016/j.neuropsychologia.2009.05.010[published Online First: Epub Date]|.

62. Umarova RM, Reisert M, Beier TU, et al. Attention-network specific alterations of structural connectivity in the undamaged white matter in acute neglect. Human brain mapping 2014;35(9):4678-92 doi: 10.1002/hbm.22503[published Online First: Epub Date]|. A very interesting study showing contralesional white matter changes in patients with neglect. Further work is necessary to pin down how exactly this relates to dysfunction and recovery.

63. He BJ, Snyder AZ, Vincent JL, et al. Breakdown of functional connectivity in frontoparietal networks underlies behavioral deficits in spatial neglect. Neuron 2007;53(6):905-18 doi: 10.1016/j.neuron.2007.02.013[published Online First: Epub Date]|.

64. Umarova RM, Saur D, Kaller CP, et al. Acute visual neglect and extinction: distinct functional state of the visuospatial attention system. Brain : a journal of neurology 2011;134(Pt 11):3310-25 doi: 10.1093/brain/awr220[published Online First: Epub Date]|.

65. Mesulam MM. A cortical network for directed attention and unilateral neglect. Annals of neurology 1981;10(4):309-25 doi: 10.1002/ana.410100402[published Online First: Epub Date]|.

66. Sparing R, Thimm M, Hesse MD, et al. Bidirectional alterations of interhemispheric parietal balance by non-invasive cortical stimulation. Brain : a journal of neurology 2009;132(Pt 11):3011-20 doi: 10.1093/brain/awp154[published Online First: Epub Date]|.

67. Corbetta M, Ramsey L, Callejas A, et al. Common behavioral clusters and subcortical anatomy in stroke. Neuron 2015;85(5):927-41 doi: 10.1016/j.neuron.2015.02.027[published Online First: Epub Date]|. An anatomical study looking at a large sample of relatively unselected stroke patients, allowing the investigators to carefully examine anatomy and multiple behavioural deficits in tandem. The results highlight the importance of white matter damage and network dysfunction, and also begin to dissect the behavioural factors that contribute to impaired performance in separate cognitive domains.

68. Nachev P. The first step in modern lesion-deficit analysis. Brain : a journal of neurology 2014 doi: 10.1093/brain/awu275[published Online First: Epub Date]|.

69. Levine DN, Warach JD, Benowitz L, et al. Left spatial neglect: effects of lesion size and premorbid brain atrophy on severity and recovery following right cerebral infarction. Neurology 1986;36(3):362-6

70. Bahrainwala ZS, Hillis AE, Dearborn J, et al. Neglect performance in acute stroke is related to severity of white matter hyperintensities. Cerebrovascular diseases 2014;37(3):223-30 doi: 10.1159/000357661[published Online First: Epub Date]|.

71. Bellgrove MA, Chambers CD, Johnson KA, et al. Dopaminergic genotype biases spatial attention in healthy children. Molecular psychiatry 2007;12(8):786-92 doi: 10.1038/sj.mp.4002022[published Online First: Epub Date]|.

72. Bartolomeo P. The quest for the 'critical lesion site' in cognitive deficits: problems and perspectives. Cortex; a journal devoted to the study of the nervous system and behavior 2011;47(8):1010-2 doi: 10.1016/j.cortex.2010.11.007[published Online First: Epub Date]|.

73. Shallice T, Skrap M. Localisation through operation for brain tumour: a reply to Karnath and Steinbach. Cortex; a journal devoted to the study of the nervous system and behavior 2011;47(8):1007-9 doi: 10.1016/j.cortex.2010.12.006[published Online First: Epub Date]|.

74. Zilli EM, Heilman KM. Allocentric spatial neglect with posterior cortical atrophy. Neurocase 2015;21(2):190-7 doi: 10.1080/13554794.2013.878731[published Online First: Epub Date]|.

75. Andrade K, Samri D, Sarazin M, et al. Visual neglect in posterior cortical atrophy. BMC neurology 2010;10:68 doi: 10.1186/1471-2377-10-68[published Online First: Epub Date]|.

76. Cipolotti L, Healy C, Chan E, et al. The impact of different aetiologies on the cognitive performance of frontal patients. Neuropsychologia 2015;68:21-30 doi: 10.1016/j.neuropsychologia.2014.12.025[published Online First: Epub Date]|. An important study examining whether different pathological aetiologies led to variation in cognitive profiles in frontal patients. The lack of difference between patient subgroups suggests that grouping patients together may not be as problematic as was previously thought.