Neuropsychologia 204 (2024) 109007

Contents lists available at ScienceDirect

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

Multivariate and network lesion mapping reveals distinct architectures of

domain-specific post-stroke cognitive impairments
Margaret Jane Moore a,* , Jason B. Mattingley a,b, Nele Demeyere c,d
a
Queensland Brain Institute & School of Psychology, The University of Queensland, St Lucia, 4072, Australia
b
Canadian Institute for Advanced Research, Toronto, Canada
c
Department of Experimental Psychology, University of Oxford, United Kingdom
d
Nuffield Department of Clinical Neurosciences, University of Oxford, United Kingdom

A R T I C L E I N F O A B S T R A C T

Keywords: Background: The purpose of this study was to identify patterns of structural disconnection and multivariate
Cognitive screening lesion-behaviour relationships associated with post-stroke deficits across six commonly impacted cognitive do-
Stroke mains: executive function, language, memory, numerical processing, praxis, and visuospatial attention.
Lesion symptom mapping
Methods: Stroke survivors (n = 593) completed a brief domain-specific cognitive assessment (the Oxford
Neuropsychological assessment
Clinical imaging
Cognitive Screen (OCS)) during acute hospitalisation. Network-level and multivariate (sparce canonical corre-
lation) lesion mapping analyses were conducted to identify focal neural correlates and distributed patterns of
structural disconnection associated with impairment on each of the 16 OCS measures.
Results: Network-level and multivariate lesion mapping analyses identified significant correlates for 12/16 and
10/16 OCS measures, respectively which were largely consistent with correlates reported in past work. Language
impairments were reliably localised to network- and voxel-level correlates centred in left fronto-temporal re-
gions. Memory impairments were associated with disconnection in a large network of left hemisphere regions.
Number processing deficits were associated with damage to voxels centred in the left insular/opercular cortex, as
well as disconnection within the surrounding white matter tracts. Within the domain of attention, different
subtypes of visuospatial neglect were linked to distinct but partially overlapping patterns of disconnection and
voxel-level damage. Praxis impairment was not linked to any voxel-level regions but was significantly associated
with disconnection within the left hemisphere dorsal attention network.
Conclusion: These results highlight the utility of routine, domain-specific cognitive assessment and imaging data
for theoretically-driven lesion mapping analyses, while providing novel insight into the complex anatomical
correlates of common and debilitating post-stroke cognitive impairments.

1. Introduction Thiebaut de Schotten et al., 2008). While previous work has reported
in-depth analyses of network-level and multivariate anatomy of single
Post-stroke cognitive impairments are often conceptualised as being neuropsychological deficits, there is a lack of evidence linking perfor-
caused by damage to focal brain regions (Moore and Demeyere, 2022; mance on brief, multi-domain cognitive screens with their associated
Sagnier et al., 2019; Weaver et al., 2021). However, many common complex anatomical correlates. The purpose of the present study was to
post-stroke cognitive deficits can result from disrupted communication identify patterns of structural disconnection and multivariate
between distant areas (Boes et al., 2015; Gleichgerrcht et al., 2017; lesion-behaviour relationships associated with impairments across
Herbet et al., 2015; Mah et al., 2014). These more complex several cognitive domains as captured by the Oxford Cognitive Screen
brain-behaviour relationships are important to characterise as they can (OCS) (Demeyere et al., 2016). The overarching goal was to determine
explain why similar deficits result from a range of different lesion pro- whether brief, standardised post-stroke cognitive screening measures
files, and can help clinicians develop a more comprehensive under- are able to detect complex brain-behaviour relationships in a diverse and
standing of how stroke lesions impact cognition (Bowren et al., 2022; representative sample of acute stroke survivors.

* Corresponding author. Queensland Brain Institute, QBI Building 79, University of Queensland, St. Lucia, QLD, 4067, Australia.
E-mail address: [email protected] (M.J. Moore).

https://doi.org/10.1016/j.neuropsychologia.2024.109007
Received 16 May 2024; Received in revised form 20 August 2024; Accepted 1 October 2024
Available online 2 October 2024
0028-3932/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
M.J. Moore et al. Neuropsychologia 204 (2024) 109007

In studies employing traditional voxel-wise univariate lesion map- et al., 2017; Salvalaggio et al., 2020). This approach is advantageous
ping, domain-specific post-stroke cognitive impairments have been because similar patterns of network-level disconnection can result from
linked to dissociable patterns of lesion damage (Moore and Demeyere, lesions which do not overlap and could therefore not be detected by
2022). This principle has been demonstrated in studies of the neural voxel-wise approaches (Herbet et al., 2015). However, the majority of
correlates of single neuropsychological impairments (e.g., Chechlacz network-level lesion mapping approaches rely on standard disconnec-
et al., 2010; Mirman et al., 2015; Mock et al., 2022; Varjačić et al., tion atlases (rather than in-vivo tractography), which may not capture
2018), in analyses employing lengthy neuropsychological assessment potentially critical patterns of anatomical variability across individuals
batteries (Weaver et al., 2021), and by studies using cognitive screening (Griffis et al., 2021). Both multivariate and network-level lesion map-
data collected as a component of routine clinical care (Moore and ping have been applied to elucidate complex brain-behaviour relation-
Demeyere, 2022). The use of routinely collected, brief screening data ships in a number of individual post-stroke cognitive impairments (e.g.,
helps avoid many of the key practical limitations associated with lesion Ghaleh et al., 2020; Saxena et al., 2022; Wiesen et al., 2019; Yourganov
mapping studies which rely on detailed neuropsychological batteries et al., 2016). Recent work has also leveraged these approaches to
(see Moore et al., 2023a; Moore and Demeyere, 2022), at the cost of a investigate the anatomy of distinct cognitive impairments captured by
degree of behavioural detail. However, past research has demonstrated performance on detailed neuropsychological assessments (Bowren et al.,
that lesion mapping analyses which employ behavioural data from brief 2022). However, these techniques have not yet been applied to char-
cognitive screens generally produce results which agree well with lesion acterise the anatomy of cognitive deficits captured by brief,
mapping studies employing extensive neuropsychological batteries domain-specific cognitive screening tests in acute stroke survivors.
(Moore and Demeyere, 2022). For example, Moore and Demeyere The current study aimed to address this knowledge gap by applying
(2022) used univariate voxel-wise lesion mapping to demonstrate that advanced lesion mapping analyses to identify the neural correlates of
the OCScan reliably differentiate neural correlates supporting 11 common, domain-specific post-stroke cognitive impairments in a large
distinct cognitive functions using behavioural data collected in a short, and representative sample of stroke survivors. Our goal was to expand
standard assessment (<20 min) and routine clinical imaging data. This understanding of brain-behaviour relationships as measured by stan-
past work opens up univariate lesion mapping techniques to a wealth of dard cognitive screening tools, used in routine clinical practice. Brain-
studies that can capitalise on existing clinical data. However, while behaviour relationships are most thoroughly characterised by
univariate lesion mapping approaches can effectively identify regions comparing anatomical results across different levels of analysis, mean-
which may contribute to behaviours of interest, they do not always ing that combining multiple, advanced lesion mapping approaches has
adequately capture the true underlying complexity of brain-behaviour the potential to provide novel insight into the anatomy of common and
relationships (Herbet et al., 2015; Mah et al., 2014). debilitating post-stroke cognitive impairments (Herbet et al., 2015;
Univariate voxel-wise lesion mapping, by design, searches for Moore et al., 2023a). Large-scale studies that characterise the correlates
spatially localised voxel clusters at which the occurrence of lesion of a range of distinct cognitive impairments within single samples are
damage is associated with a specific behavioural impairment (Bates important for expanding theoretical knowledge of brain-behaviour re-
et al., 2003). This approach can generate accurate results for deficits lationships, as well as for establishing the degree of information
which are indeed linked to spatially focal neural correlates, but the complexity that can be extracted from routine clinical data – both in
anatomy of many common post-stroke cognitive deficits is more com- terms of brain imaging and clinically used cognitive screening.
plex. For example, previous studies have suggested that visuospatial
neglect is a disconnection syndrome resulting from non-overlapping 2. Methods
lesions impacting diffuse neural networks responsible for distributing
attention across space (Bartolomeo et al., 2007; Moore et al., 2023b,d; 2.1. Patients
Saxena et al., 2022). Similarly, praxis impairments have been linked to
severed communication between left temporal, parietal, and frontal This retrospective analysis included data from 593 stroke patients
regions (Hoeren et al., 2014; Rosenzopf et al., 2022). Post-stroke exec- recruited in the Oxford Cognitive Screening Programme (Demeyere
utive function, numerical cognition, language, and memory impair- et al., 2016; Milosevich et al., 2024) (Table 1). The Screening Pro-
ments have all been linked to damage over multiple, spatially distinct gramme studies included a consecutive sample of acute stroke survivors,
regions (Lugtmeijer et al., 2021; Mirman et al., 2015; Moore and excluding only patients who were unable to maintain concentration for
Demeyere, 2022; Tsuchida and Fellows, 2013). In cases (such as these) 15 min or were unable to provide informed (witnessed) consent. The
where deficits can result from damage at spatially separated regions, study was conducted in accordance with the Declaration of Helsinki, and
univariate lesion mapping yields results which represent a “spatial the protocol was approved by the National Research Ethics Committee
average” of all involved correlates, but these may not overlap with the
true, underling neural correlates (Mah et al., 2022). Table 1
Multivariate and network-level lesion mapping methods were Demographics and stroke characteristics for the study sample. Where relevant,
developed to address this critical limitation. Multivariate lesion map- sample means are reported followed by standard deviations in parentheses.
ping considers the large-scale spatial distributions of lesion damage Bilateral strokes were classed as single stroke events which crossed the
across all brain regions simultaneously to identify lesion patterns that anatomical midline (e.g., pontine strokes).
distinguish between patients with and without the deficit of interest Demographics:
(Mah et al., 2014; Pustina et al., 2018). Past work has suggested that this N 593
multivariate approach yields more precise results than traditional uni- Age 73.5 (13.1)
Gender 45.5% Female
variate approaches, and is able to more accurately characterise the
Test Timing: ​
anatomy of deficits caused by damage to spatially distinct brain regions Stroke-Test Interval 5.87 (6.28)
(Pustina et al., 2018). Despite these advantages, past work has demon- Stroke-Scan Interval 1.75 (4.17)
strated that multivariate lesion mapping methods can still yield results Stroke Side: ​
that are spatially biased (Ivanova et al., 2021; Sperber et al., 2019) and, Right Hemisphere 286
Left Hemisphere 247
in some cases, may use features other than the deficit of interest to Bilateral 60
distinguish between spared an impaired patients (Moore et al., 2023a). Stroke Type: ​
Conversely, network-level lesion mapping identifies specific patterns Ischemic 80.80%
of structural disconnection for which the extent of damage predicts the Haemorrhagic 19.20%
Lesion volume 3.7 cm3 (54.6)
severity of behavioural impairment (Boes et al., 2015; Gleichgerrcht

2
M.J. Moore et al. Neuropsychologia 204 (2024) 109007

(Health Research Authority UK (REC references:14/LO/0648, quantified using the same metric (e.g., right and left neglect), all scores
18/SC/0550, 12/WM/00335). Patients were included in the present not representing significant impairment in the deficit of interest were
analysis if they completed the Oxford Cognitive Screen (OCS) within 31 constrained to 1. This scoring method follows the previously reported
days of stroke and had available acute neuroimaging (within 31 days of approach for lesion mapping studies using OCS data (Moore et al.,
stroke) which was of sufficient quality to facilitate lesion delineation. Of 2023b).
the 1198 patients recruited in the OCS Programme, 398 were excluded
from this analysis for having no useable neuroimaging data. Ninety-six 2.3. Neuroimaging
patients were excluded due to evidence of multiple, temporally
distinct stroke lesions. One hundred and thirteen additional patients Neuroimaging data were collected as a component of routine clinical
were excluded due to OCS or neuroimaging data being collected >31 care (508 CT, 85 MRI (74 T2, 3 T1, 8 FLAIR)). This combination of
days post stroke. different imaging types optimises statistical power by increasing sample
size and maximising the degree of lesion overlap. Previous studies have
2.2. Behavioural data found that these imaging modalities yield comparable lesion mapping
results (de Haan and Karnath, 2018; Moore et al., 2023c). The utilised
Cognitive status was assessed using the Oxford Cognitive Screen lesion dataset has been demonstrated to be of sufficient quality to
(OCS; Demeyere et al., 2015). The OCS is a short, stroke-specific localise established correlates of post-stroke cognitive impairments in
cognitive screen designed for use in acute stroke settings. It yields 16 univariate lesion mapping analyses (Moore and Demeyere, 2022). All
distinct measures summarising cognitive impairments across commonly lesion masks were manually delineated by trained experts in line with
impacted cognitive domains: executive function, language, memory, the standard protocol reported by Moore (2022). Native-space lesion
number processing, praxis, and visuospatial attention (Demeyere et al., masks were smoothed at 5 mm full-width at half-maximum in the z-di-
2015, 2016). This measure is designed to be inclusive for patients with rection, binarized (0.5 threshold), reoriented, warped into 1 × 1 × 1 mm
common stroke-related impairments including aphasia, visuospatial stereotaxic space using Statistical Parametric Mapping (Ashburner et al.,
neglect, primary visual impairments (e.g. hemianopia), and primary 2016) and Clinical Toolbox (Rorden et al., 2012) functions, and were
motor impairments (e.g. hemiplegia) (Demeyere et al., 2015, 2016). visually inspected for quality. All lesions which were unable to be
The OCS executive function domain is assessed using a modified accurately normalised were excluded from this study.
(shape-based) trail making task. The OCS language domain consists of a
picture naming task (images of four objects/animals), a semantics task 2.4. Network-level lesion mapping
(follow verbally presented instructions), and a sentence reading task.
The OCS memory domain includes a delayed recall of the previously Network-level lesion mapping was conducted to identify patterns of
read sentence (verbal recall task), an incidental multiple-choice recall of structural disconnection significantly associated with impairment on
stimuli and tasks completed earlier in the OCS (episodic recognition each OCS measure. To estimate disconnectivity, patient lesion masks
task), and 4 items assessing the participant’s orientation in time and were used to estimate parcel-wise disconnectivity across all cortical
location (orientation task). The OCS numerical processing domain in- areas defined in the Schaefer-Yeo Atlas (100 parcels, 7 networks) (Yeh
cludes a number tasks in which participants are asked to write three et al., 2018) and 35 subcortical/cerebellar areas defined in the AAL
verbally presented numbers (in digits) and a calculation task where (Tzourio-Mazoyer et al., 2002) and Harvard-Oxford subcortical atlases.
participants are asked to solve four simple addition and subtraction These atlases define 7 functional brain networks: Control, Default,
equations. The OCS Praxis domain consists of a gesture and hand- Dorsal Attention, Limbic, Somatic Motor, Ventral Attention, and Visual
position imitation task. Finally, the OCS visuospatial attention domain as well as networks connecting subcortical/cerebellar structures
consists of the hearts cancellation task which yields a total score (Schaefer et al., 2018). These network parcellations were established
(number of targets identified) as well as egocentric neglect and allo- through gradient-weighted Markov Random Field modelling combining
centric neglect scores. Detailed descriptions of OCS subtests and local gradient and global similarity measures applied to a range of
administration procedures are reported elsewhere (Demeyere et al., task-related and resting state fMRI datasets (see Schaefer et al., 2018).
2015, 2016). The Schaefer-Yeo cortical parcellation was compared with the HCP-
Previous research has demonstrated that impairments are correlated 842 atlas to quantify disconnection statistics. The HCP-842 atlas is a
across some OCS domains (Milosevich et al., 2024). For example, found population-averaged disconnectome atlas derived from the diffusion
that domain-level impairments in numeric processing and memory were MRI data of 842 subjects (Yeh et al., 2018). These individual data were
highly correlated (Rtet = 0.69), whereas those between attention and used to map individual white matter tractographies and group these into
praxis were not (Rtet = 0.08). Such patterns are expected because spe- 550,000 trajectories of representative white matter fascicles (stream-
cific profiles of post-stroke cognitive impairment may be due to lines) (Yeh et al., 2018). In cases in which lesions intersect with each of
non-random variations in stroke location (Corbetta et al., 2018) and these defined streamlines, the relevant streamline is considered to be
pre-morbid cognitive status (e.g. MCI, cognitive decline) (Deary et al., disconnected (Griffis et al., 2021). Network-level disconnectivity sta-
2009). Despite domain-level correlations, several large-scale validation tistics were generated by calculating the proportion of streamlines that
studies have provided strong evidence that individual OCS subtests are terminate (end or begin) in each pair of parcels that were disconnected
effectively able to tap their respective functions of interest with minimal (Griffis et al., 2021). This process produced disconnection matrices in
interference from co-morbid impairments (Bormann et al., 2024; which each cell’s value represents the proportion of disconnected
Demeyere et al., 2015; Kong et al., 2016). The OCS has been validated in streamlines (network edges) connecting each pair of defined grey matter
large and representative stroke samples (Bisogno et al., 2023; Bormann parcels (nodes) per patient (135 nodes, 18,225 edges). This disconnec-
et al., 2024; Demeyere et al., 2015; Huygelier et al., 2020; Robotham tion quantification method is standard for use in network-level lesion
et al., 2020) and provides sufficient cognitive detail to link common mapping analyses (Griffis et al., 2021).
post-stroke cognitive impairments to distinct patterns of lesion damage For each network edge impacted in at least five participants,
(Moore and Demeyere, 2022). regression analyses were conducted to determine whether percent
Where relevant, all binary impairment classifications were made disconnection at this edge was significantly associated with poor per-
based on standard impairment thresholds. To ensure consistency across formance on each OCS measure. Each regression included only partici-
scores, patient scores were standardised as proportion correct and pants who completed the relevant subtest, and included lesion volume
ranged between 0 (worst possible score) and 1 (best possible score) for as a covariate of no interest. All disconnection statistics were generated
each measure. In cases where independent cognitive impairments are using Lesion Quantification Toolkit (Griffis et al., 2021).

3
M.J. Moore et al. Neuropsychologia 204 (2024) 109007

Bonferroni corrections for multiple comparisons were applied. In 3. Results

cases where no results survived Bonferroni corrections, less conservative
5% FDR corrections were applied. This flexible approach was employed 3.1. Behavioural results
to balance the advantages of both correction methods in lesion mapping.
For example, Bonferroni corrections are extremely conservative but can Table 2 presents OCS test results for the included patient sample.
reliably identify very strong effects in very large patient samples (Moore Across all considered measures, 86.3% of patients exhibited at least one
et al., 2021; Moore and Demeyere, 2022). However, these corrections cognitive impairment, and 72.8% of patients had impairments on two or
have a high false-negative risk when effect sizes (or included samples) more OCS measures. Larger lesions were associated with worse perfor-
are smaller (Mirman et al., 2018). To ensure null results were not due to mance across all OCS measures (all p < 0.001), with the exception of the
the use of highly conservative corrections, less conservative FDR cor- trail-making task (p = 1.87), verbal recall (p = 0.278), and right neglect
rections were applied in cases where no results survived Bonferroni (egocentric p = 0.646, allocentric p = 0.721) (see Supplementary
correction. FDR corrections employ alpha thresholds that are more strict Table 2). Fig. 2 presents lesion overlays for patients impaired on each
than those used in past network-level lesion mapping studies (e.g., p < measure.
0.05, Philippi et al., 2021). FDR corrections were only employed in cases
where Bonferroni corrections did not yield significant results because, in
cases where effects are large, FDR corrections often indicate that nearly 3.2. Network-level lesion mapping
all conducted tests are significant. This flexible correction approach has
been employed in previous network-level lesion mapping studies Network-level lesion mapping analyses identified significant
(Moore et al., 2024,2023b). All cases in which FDR corrections were network edges associated with cognitive impairment on 12 of the 16
used are reported. All reported R2 values are adjusted. considered OCS measures. Fig. 3 presents all network edges that were
significantly associated with each OCS measure. Fig. 4 summarises the
2.5. Multivariate lesion mapping proportion of significant network edges which belong to each of the
seven functional networks defined by the Schaefer-Yeo Atlas.
Multivariate voxel-wise lesion mapping was conducted using the Within the language domain, Picture Naming impairment was asso-
sparse canonical correlation (SCCAN) approach implemented in LESY- ciated with damage to 172 edges spanning left frontal and temporal
MAP (Pustina et al., 2018). This technique is an optimisation method
which identifies multivariate associations between voxel-level patterns Table 2
in lesion damage and behavioural scores (Pustina et al., 2018). The Results from the Oxford Cognitive Screen (OCS). ‘Tested’ refers to the number of
approach builds a best-fit model that leverages multivariate lesion pat- participants assessed on each measure. ‘Number impaired’ is the number and
terns to predict behavioural scores. The proportion of voxels retained in percentage of tested participants meeting impairment criteria. ‘Spared’ and
‘Impaired’ refer to the mean standardised score and standard deviation (in pa-
models (i.e., model sparseness) is determined through a standardised
rentheses) for spared and impaired patients on each measure, respectively.
sparseness optimisation procedure. This procedure involves a four-fold
model cross validation operation (using a 75% training/25% testing Domain/Test

data split) to evaluate the fit of models with varying degrees of sparse- Executive Tested Number Spared Impaired
ness. SCCAN models are only classed as valid if a model’s predicted Impaired
behavioural scores are above chance (p < 0.05) within the Trail Making Task 563 78 (13.8%) 0.07 0.58
cross-validation procedure. This method has been validated for use in (0.11) (0.16)
lesion mapping and has been shown to reduce spatial misallocation of Numeric
Calculation 591 116 (19.6%) 0.08 0.68
voxel-level results relative to traditional univariate lesion mapping ap- (0.12) (0.20)
proaches (Pustina et al., 2018). Number Writing 592 258 (43.6%) 0.00 0.66
Only voxels impacted in at least 5 patients were considered in SCCAN (0.00) (0.29)
models. Lesion volume was corrected for using the direct total lesion Language
Picture Naming 592 194 (32.8%) 0.11 0.71
volume control method to weight input voxel values by lesion size in
(0.14) (0.21)
order to provide a balanced control for the effect of stroke severity on Sentence Reading 583 204 (35.0%) 0.02 0.54
impairment probability (Zhang et al., 2014). SCCAN models were con- (0.03) (0.35)
structed and evaluated for each of the 16 considered OCS measures. Semantics 591 59 (10.0%) 0.00 0.57
Notably, SCCAN models produce voxel values between − 1 and +1, with (0.00) (0.26)
Memory
negative values indicating areas that predict the absence of impairments
Orientation 592 139 (23.5%) 0.00 0.37
and positive values indicating areas that predict the occurrence of def- (0.00) (0.18)
icits of interest (Pustina et al., 2018). Given that the aim of our study was Recall 584 171 (29.3%) 0.07 0.71
to identify areas that, when damaged, lead to significant impairments, (0.11) (0.20)
Recognition 545 122 (22.4%) 0.08 0.62
only significant voxels assigned positive weights are presented and
(0.12) (0.17)
interpreted. Full SCCAN models (including both positive and negative Praxis
values) are provided in supplementary materials. Praxis 586 162 (27.6%) 0.11 0.64
The anatomy of all significant voxel clusters was interpreted relative (0.12) (0.19)
to the Harvard-Oxford Cortical Atlas and Johns Hopkins White Matter Visuospatial
Visual Field Task 591 111 (18.8%) 0.00 0.50
Atlas (Mori et al., 2005). These atlases were selected to generate results
(0.00) (0.22)
that are directly comparable with previous lesion mapping studies (e.g., Cancellation Total 582 332 (57%) 0.07 0.57
Moore and Demeyere, 2022), and to provide anatomical results that are (0.05) (0.28)
balanced in their complexity and ease of interpretation. Egocentric Neglect 547 120 (21.9%) 0.00 0.43
(Left) (0.00) (0.21)
Egocentric Neglect 547 60 (11.0%) 0.00 0.41
2.6. Data availability statement (Right) (0.00) (0.27)
Allocentric Neglect 547 97 (17.7%) 0.00 0.12
All anonymised data, materials, and code associated with this project (Left) (0.00) (0.11)
are openly available at https://osf.io/depjt/, with the exception of Allocentric Neglect 547 55 (10.1%) 0.00 0.07
(Right) (0.00) (0.05)
lesion masks. Lesion masks are available upon request from the authors.

4
M.J. Moore et al. Neuropsychologia 204 (2024) 109007

Fig. 1. Lesion overlay for the Study sample (n = 593). Voxel colour denotes the number of patients with lesions impacting each region. Only regions impacted in at
least 5 participants are visualised. MNI z-slices -32 – 48 are presented.

regions (R2 values = 0.002–0.005, all p < 3.64 × 10− 5). Damage to the Left Egocentric Neglect was associated with disconnection across a
edge connecting the left hemisphere dorsal attentional network (poste- wide network, spanning the majority of the right hemisphere (n = 415,
rior division 1) and the left default network (parietal division 2) was the R2 = 0.001–0.002, all p < 3.65 × 10− 5). Left Allocentric Neglect was
strongest predictor of Picture Naming impairment. Impairment on the associated with damage to 107 network edges, spanning right hemi-
Semantics task was associated with disconnection across 79 edges within sphere temporo-parietal regions (R2 = 0.0003–0.0006, all p < 3.47 ×
left hemisphere temporo-parietal regions (R2 values = 0.001–0.002, all 10− 5). Disconnection within streamlines connecting the right somatic
p < 3.63 × 10− 5). Impairment in the Semantics task was most strongly motor network (division 6) and the dorsal attention network (posterior
predicted by disconnection between the left hemisphere visual network division 2) were the strongest predictors of both Left Egocentric Neglect
(node 7) and the left hemisphere default network (parietal division 2). and Allocentric Neglect impairments. There was substantial, but not
Sentence Reading impairment was associated with damage to 172 left complete, overlap between the networks associated with left egocentric
hemisphere edges (R2 values = 0.002–0.005, all p < 2.98 × 10− 5), with and allocentric neglect. Specifically, 90.7% of edges associated with left
damage to the edge connecting the left hemisphere frontal eye fields allocentric neglect were also significantly associated with left egocentric
(dorsal attention network) and the left putamen being the strongest neglect. The edges associated with allocentric, but not egocentric neglect
predictor of impairment. Across all three Language domain tasks, 18.2% primarily involved connections between the control and dorsal atten-
of significant network edges were common across all three tasks, with an tional network (15%), within the control network (10%), and between
additional 44.2% of significant edges being common across at least two the ventral attention and somatic motor networks (10%).
tasks. Right Egocentric Neglect was associated with damage to 43 edges (R2
Within the Memory domain, Episodic Recognition deficits were asso- = 0.0004–0.0004, all p < 0.016 (FDR corrected)), with disconnection
ciated with damage to 103 edges mainly located in left temporo-parietal between the left hemisphere temporal pole (limbic network) and left
regions (R2 values = 0.001–0.003, all p < 3.61 × 10− 5). Impairment on prefrontal cortex (default network) being the strongest predictor or
the Episodic Recognition task was most strongly predicted by damage to impairment. Right Allocentric Neglect was linked to disconnection at 16
connections between posterior division 1 of the dorsal attention network edges (R2 = 0.0001–0.0002, all p < 3.31 × 10− 5). Disconnection be-
and parietal division 2 of the default network. Orientation impairment tween the left somatic motor networks (divisions 1 and 2) was the best
was significantly associated with disconnection of 140 edges (R2 values predictor of Right Allocentric Neglect impairment. Only 3.4% of network
= 0.0005–0.001, all p < 0.018 (FDR corrected)). Disconnection between edges associated with Right Egocentric Neglect and Right Allocentric
the left hemisphere prefrontal cortex (division 7, default network) and Neglect were significantly associated with both deficits.
the left pallidum was the single strongest predictor of Orientation Visual Field Task impairment was associated with damage to 111
impairment. Verbal Recall deficits were associated with damage to 173 network edges (R2 = 0.001–0.003, all p < 3.67 × 10− 5). Damage to
edges (R2 values = 0.002–0.005, all p < 3.26 × 10− 5). The strongest streamlines connecting the right visual network (division 6) and the
predictor of impairment on this task was damage to streamlines con- right precuneus (control network) was the strongest predictor of
necting the left frontal eye fields (dorsal attentional network) and the impairment on this task. Praxis impairment was linked to disconnection
left putamen. Across all Memory domain subtests, 19.2% of all signifi- in 60 edges (R2 = 0.0007–0.002, all p < 0.017 (FDR corrected)), with
cant edges were common across all three tasks. An additional 50.7% of damage to connections between the left default network (parietal divi-
significant nodes were shared across at least two tasks. sion 2) and the left pallidum being the strongest predictor of Praxis
Number Writing deficits were associated with disconnection in 133 impairment. Within the Executive Function domain, no network edges
edges (R2 = 0.002–0.006, all p < 3.28 × 10− 5). Damage to the edge were significantly associated with poor Trail Making Task performance
connecting the left thalamus and the left frontal opercular cortex (di- (FDR corrected).
vision 2, ventral attention network) was the strongest predictor of
impairment. Calculation impairment was linked to disconnection of 9
edges (R2 = 0.001–0.002, all p < 2.74 × 10− 5). Damage to streamlines 3.3. Multivariate lesion mapping results
connecting the left pallidum and the left somatic motor network (divi-
sion 4) was the strongest predictor of poor Calculation performance. All SCCAN lesion mapping yielded significant models for 10 of the 16
network edges significantly associated with Calculation impairment considered measures. Fig. 5 visualises all voxel clusters that, when
were also associated with Number Writing impairment. damaged, were associated with impairment. Table 3 reports the anat-
Low Cancellation Total scores (regardless of the presence of response omy of each of these significant clusters. Detailed SCCAN results,
asymmetry) were associated with disconnection in 14 edges (R2 = including optimum sparseness values and full voxel weight maps, are
0.002–0.003, all p < 3.54 × 10− 5). Cancellation Total impairment was available in supplementary materials.
most strongly predicted by disconnection between the right hemisphere Within the Language domain, Picture Naming impairment was linked
precuneus (control network) and the precuneus/posterior parietal cor- to damage to large voxel clusters (n = 79,960), spanning left hemisphere
tex (default network). anterior temporal regions. Sentence Reading impairment was associated
with a more restricted voxel cluster (n = 5130), primarily impacting the

5
M.J. Moore et al. Neuropsychologia 204 (2024) 109007

Fig. 2. Lesion overlays for patients exhibiting impairment on each OCS measure. Voxel colour represents the number of impaired patients with lesions overlapping at
each region. MNI z-slices -32 – 48 are presented. Total number impaired on each measure is reported in Table 2.

left central opercular cortex and underlying white matter. No significant capsule). Within the Visuospatial Domain, Cancellation Total score was
voxels were associated with worse performance on the OCS Semantics associated with damage to 139,745 voxels distributed through right
task. In the Memory domain, Orientation impairment was not associated posterior parietal areas and right occipital/superior temporal regions.
with any voxel clusters. However, Episodic Recognition was linked to Left Egocentric Neglect was linked to damage to several voxel clusters (n
damage to 4256 voxels and Verbal Recall deficits were predicted by = 45,697) spanning right temporo-parietal and occipital regions. Right
damage to 4885 voxels. The voxel clusters associated with both Verbal Egocentric Neglect, Left Allocentric Neglect, and Right Allocentric Neglect
Recall and Episodic Recognition involved overlapping, but not identical, were not found to be associated with any significant voxel clusters. Vi-
regions centred in the left basal ganglia. sual Field Task impairment was associated with damage to 86,493 voxels
Number Writing impairment was associated with 27,235 voxels, primarily located in the bilateral primary visual cortices and right pos-
whereas Calculation impairment was predicted by damage to 3677 terior parietal regions. Gesture Imitation Task impairment was not asso-
voxels. Both these impairments’ significant correlates were centred on ciated with any voxels. Trail Making Task impairment was significantly
the opercular/insular cortices and underlying white matter (external associated with 3248 voxels, located in the bilateral white matter

6
M.J. Moore et al. Neuropsychologia 204 (2024) 109007

Fig. 3. Visualisation of significant network edges associated with impairment on each OCS measure. Network nodes are represented by white dots, and edge colour
represents adjusted R2 value for each comparison. On each slice, node locations are collapsed into two dimensions and plotted in ascending order of edge R2 value so
that stronger predictors of impairment are plotted on top. Axial slices are MNI z = 14, coronal are MNI y = − 16, and sagittal are MNI x = 8. Names, MNI coordinates,
and lesion mapping statistics for each node are openly available (https://osf.io/depjt/).

(external capsule) and right middle/superior temporal gyrus. patterns of structural disconnection that were associated with language
impairments in the present study.
4. Discussion Within the memory domain, verbal recall, episodic recognition, and
orientation task impairment were associated with large and diverse
Taken together, our findings suggest that brief, domain-specific network-level correlates spanning the majority of the left hemisphere. In
cognitive assessment data are able to capture complex network-level the SCCAN analysis, verbal recall and episodic recognition were linked
and multivariate brain-behaviour relationships in post-stroke cognitive to left fronto-temporal correlates, whereas orientation impairment was
impairment. The identified correlates generally agree well with those not associated with any significant correlates. Past multivariate lesion
documented in past studies that used lesion mapping methods to iden- mapping analyses have linked memory deficits to similar voxel-level
tify regions associated with impairments identified by detailed neuro- correlates (e.g., left fusiform, parahippocampal and insular cortices,
psychological assessments. This finding is important as it demonstrates and deep frontal white matter; Bowren et al., 2022). Notably, orienta-
that even simple behavioural data can be used to elucidate complex tion impairment was most strongly predicted by damage to the default
brain-behaviour relationships. The results highlight the utility of mode network edge connecting the left pallidum and prefrontal cortex.
routinely collected cognitive assessment and imaging data for Previous functional imaging work has found that signals associated with
theoretically-driven lesion mapping analyses, and provide novel insight orientation in time and space strongly overlap with the default network
into the complex anatomical correlates of common and debilitating post- (Peer et al., 2015). This suggestion is in line with our study’s finding that
stroke cognitive impairments. a higher proportion of default network edges (relative to any other
OCS Language domain impairments were linked to distinct but network) were significantly associated with orientation impairment.
overlapping network- and voxel-level neural correlates largely centred Previous studies have indicated that verbal components of memory
in left fronto-temporal regions. These findings are in line with past tasks rely on left-hemisphere language areas while non-verbal memory
network-level and multivariate analyses of language-related post-stroke functions are dependent on right hemisphere cortical and subcortical
cognitive impairments. Past work has suggested that post-stroke aphasia areas (Mock et al., 2022). Considered in the context of the present
is a multidimensional deficit which can involve deficits in non-verbal study’s result, this suggests that the identified correlates of OCS Memory
cognition as well as classical language deficits (Schumacher et al., tasks may largely represent the verbal components of these cognitive
2019). This conceptualisation helps to account for the large and diverse functions. These results are unlikely to have arisen because patients with

7
M.J. Moore et al. Neuropsychologia 204 (2024) 109007

Fig. 4. Identity of network edges significantly associated with each OCS measure. Each cell represents connections between a given network (listed on x-axis) and
another network (y-axis). In cases where both networks are the same, cells represent connections within a single functional network. Cell colour (and label) denotes
the proportion of significant network edges which fall into each category. S/C denotes subcortical and cerebellar network edges.

language impairments were unable to complete the OCS memory tasks. et al., 2023a). This method therefore yields a summary of the most
Several large-scale validation studies have demonstrated that even pa- strongly associated regions, rather than capturing the wide variability of
tients with severe language impairments are able to complete OCS underlying lesions which can be associated with any given deficit. This
verbal recall, episodic recognition, and orientation tasks (Bormann methodological feature is important to consider when evaluating the
et al., 2024; Demeyere et al., 2015; Huygelier et al., 2020). However, results of the current study, particularly with regard to cases in which
such patients may nevertheless have weaker verbal encoding, leading to only significant unilateral correlates arose from lesion distributions
poorer memory performance (even though the tasks require only which were clearly bilateral.
recognition from multiple choice options, rather than free recall). It is also important to note that not all memory impairments iden-
Notably, many patients with exclusively right hemisphere lesions tified in stroke survivors are causally related to lesion damage. This is
exhibited impairments on these tasks (see Fig. 1), indicating that the because there is a high comorbidity of degenerative brain conditions
memory domain tasks likely capture both verbal and non-verbal mem- within the stroke population (e.g. small vessel disease) (Rost et al.,
ory functions. This was not, however, reflected in the lesion-mapping 2022). These neurodegenerative conditions commonly result in memory
results. This is likely due to the comparatively greater lesion overlap and executive impairments, meaning that a portion of the memory im-
of impaired patients with left hemisphere versus right hemisphere le- pairments identified by the OCS may have been present prior to stroke
sions. Bias toward clusters impacted by lesions in more patients onset and related to overall brain health rather than the stroke-specific
(ignoring critical correlates impacted in fewer patients) is a well- lesion. Our analyses cannot effectively distinguish between impairments
established limitation of univariate lesion mapping analyses (Mah that were pre-morbid and impairments that followed the stroke. This
et al., 2014). These results illustrate that such prevalence imbalances limitation may account for the documented variability of lesion loca-
may impact the results of multivariate, voxel-wise lesion mapping an- tions in patients with memory impairments.
alyses as well. Lesion mapping analyses simplify complex lesion distri- Past research has indicated that the anatomy of visuospatial neglect
butions to identify spatially contiguous regions which are responsible deficits is very complex, with previous studies reporting a wide range of
for the largest portion of behavioural variance (Mah et al., 2014; Moore voxel- and network-level correlates as being associated with this

8
M.J. Moore et al. Neuropsychologia 204 (2024) 109007

Fig. 5. Significant correlates of poor performance as reported by multivariate lesion mapping. All voxels which, when damaged, were associated with worse per-
formance are presented. Voxel colour represents the model weight assigned by SCCAN analysis, with larger values representing stronger relationships between
damaged and impairment. MNI slices -40 – 58 are presented. Full SCCAN voxel weight maps are visualised in Supplementary Materials and are openly available at htt
ps://osf.io/depjt/.

syndrome (Moore et al., 2023b,d; Saxena et al., 2022). The results of the associated neural correlates. This is the first study to identify
current study are broadly in line with this past work, as left egocentric network-level correlates of right allocentric neglect. Past lesion studies
and allocentric neglect were both linked to damage to large and diverse have indicated that this deficit can arise from a diverse range of
structural networks, spanning a large portion of the right hemisphere non-overlapping lesions (Moore et al., 2021). The results of the current
(Moore et al., 2023b; Saxena et al., 2022). In line with past work, left study provide evidence that allocentric neglect, like egocentric neglect,
egocentric and allocentric neglect were linked to distinct, but over- may represent a disconnection syndrome. This possibility, in turn, can
lapping patterns of network-level disconnection (Moore et al., 2023b). help account for why diverse, often non-overlapping lesions have pre-
Notably, left egocentric neglect impairment was associated with viously been associated with allocentric neglect (Chechlacz et al., 2012).
disconnection of some interhemispheric network edges. This result is in Cancellation task impairment (regardless of neglect impairment) was
line with previous research reporting that damage to interhemispheric linked to a diverse range of neural correlates. In SCCAN analyses,
network edges (Moore et al., 2023b,d; Saxena et al., 2022) and key hubs cancellation total was primarily linked to damage within right posterior
for interhemispheric communication (e.g. the corpus callosum parietal and occipital areas. This finding likely reflects the fact that one
(Bartolomeo et al., 2007; Corbetta and Shulman, 2011b; Heilman and of the most common reasons for patients to fail to identify targets on this
Adams, 2003);) are associated with egocentric neglect. task is the occurrence of left neglect or left visual field impairments
Right egocentric and allocentric neglect were associated with (Demeyere et al., 2016). These impairments are also linked to compar-
disconnection within largely non-overlapping left-hemisphere net- atively homologous lesion patterns (e.g. lesions to the primary visual
works. The identified network associated with right egocentric neglect cortex), which past work has shown are more likely to be identified as
impairment differs from the networks identified in previous studies (e. significant in voxel-wise lesion mapping analyses relative to lesion
g., Moore et al., 2023b). This is likely due to differences in how right patterns that are more spatially variable (Mah et al., 2014; Moore et al.,
egocentric neglect severity was quantified. Past studies have used 2023). These findings were mirrored in network-level analyses in which
detailed “centre of cancellation” metrics to capture the “centre of mass” low cancellation total score was best predicted by damage to visual
of participant responses, whereas the present study used basic OCS network edges. In line with past work, visual field impairments were
asymmetry scores (Rorden and Karnath, 2010). These standard asym- primarily associated with damage to, and disconnection between, pri-
metry scores are not able to capture the level of impairment severity mary visual areas (Bridge, 2020; Swienton and Thomas, 2014; Zhang
detail represented by centre of cancellation scores but were selected to et al., 2006). Notably, multivariate analyses primarily yielded right
accomplish this study’s goal of linking standard OCS scores with their hemisphere correlates of visual field impairment, despite the fact that

9
M.J. Moore et al. Neuropsychologia 204 (2024) 109007

Table 3
Anatomical descriptive statistics for significant voxels yielded by SCCAN analysis. Each atlas-defined ROI con-
taining significant voxels is listed. Side denotes ROI hemisphere and Fraction indicates the percentage of ROI
voxels classed as significant. In cases where measures were associated with damage to >10 ROIs, only ROIs with
fractions greater than 5% are reported.
Executive Side Fraction

Trail Making Task

Middle Temporal Gyrus (posterior division) R 6.94%

Superior Temporal Gyrus (posterior division) R 6.30%
Planum Polare R 3.90%
Posterior Limb of internal capsule L 2.99%
Heschl’s Gyrus L 2.58%
Supramarginal Gyrus (posterior division) R 2.56%
Superior corona Radiata L 2.38%
External capsule L 1.97%
Superior Temporal Gyrus (anterior division) R 1.74%
External capsule R 1.65%

Language

Picture Naming
Heschl’s Gyrus L 81.59%
Central Opercular Cortex L 72.39%
External capsule L 68.54%
Insular Cortex L 67.16%
Frontal Operculum Cortex L 58.00%
Parietal Operculum Cortex L 51.87%
Superior Longitudinal fasciculus L 45.02%
Planum Temporale L 41.85%
Inferior Frontal Gyrus (pars opercularis) L 41.25%
Planum Polare L 38.86%
Temporal Occipital Fusiform Cortex L 32.06%
Precentral Gyrus L 26.75%
Superior corona Radiata L 25.97%
Posterior Limb of internal capsule L 21.00%
Tapetum L 20.73%
Uncinate fasciculus L 19.90%
Middle Temporal Gyrus (temporooccipital part) L 17.20%
Angular Gyrus L 15.28%
Sagittal stratum L 14.84%
Posterior thalamic Radiation L 12.47%
Anterior Limb of internal capsule L 10.38%
Superior Temporal Gyrus (posterior division) L 10.36%
Retrolenticular part of internal capsule L 9.57%
Occipital Fusiform Gyrus L 9.26%
Superior fronto-occipital fasciculus L 8.93%
Intracalcarine Cortex L 8.88%
Middle Frontal Gyrus L 8.42%
Supramarginal Gyrus (posterior division) L 6.56%
Lingual Gyrus L 6.39%
Postcentral Gyrus L 5.88%
Sentence Reading
Superior Longitudinal fasciculus L 9.34%
External capsule L 9.03%
Central Opercular Cortex L 7.98%
Insular Cortex L 5.37%
Superior corona Radiata L 4.16%
Precentral Gyrus L 2.66%
Anterior Limb of internal capsule L 0.43%
Inferior Frontal Gyrus pars (opercularis) L 0.04%
Postcentral Gyrus L 0.02%
Posterior Limb of internal capsule L 0.02%

Memory Side Fraction

Episodic Recognition
External capsule L 27.68%
Posterior Limb of internal capsule L 10.80%
Insular Cortex L 4.63%
Anterior Limb of internal capsule L 3.83%
Superior fronto-occipital fasciculus L 1.93%
Superior corona Radiata L 1.20%
Central Opercular Cortex L 0.16%
Retrolenticular part of internal capsule L 0.09%
Superior Longitudinal fasciculus L 0.02%
Verbal Recall
External capsule L 34.50%
Posterior Limb of internal capsule L 8.03%
(continued on next page)

10
M.J. Moore et al. Neuropsychologia 204 (2024) 109007

Table 3 (continued )
Executive Side Fraction

Trail Making Task

Anterior Limb of internal capsule L 5.13%

Superior corona Radiata L 3.16%
Insular Cortex L 2.25%
Superior fronto-occipital fasciculus L 1.93%
Central Opercular Cortex L 0.30%
Superior Longitudinal fasciculus L 0.17%

Number

Calculation
External capsule L 10.83%
Central Opercular Cortex L 7.83%
Superior Longitudinal fasciculus L 3.52%
Insular Cortex L 3.21%
Precentral Gyrus L 1.51%
Superior corona Radiata L 1.39%
Supramarginal Gyrus (posterior division) L 0.77%
Anterior Limb of internal capsule L 0.43%
Angular Gyrus L 0.36%
Inferior Frontal Gyrus (pars opercularis) L 0.20%
Number Writing
Insular Cortex L 64.67%
External capsule L 56.86%
Central Opercular Cortex L 40.98%
Parietal Operculum Cortex L 24.73%
Uncinate fasciculus L 22.65%
Heschl’s Gyrus L 20.01%
Planum Polare L 19.04%
Superior Longitudinal fasciculus L 16.28%
Planum Temporale L 14.73%
Frontal Operculum Cortex L 13.44%
Superior corona Radiata L 9.10%
Anterior Limb of internal capsule L 5.60%

Visuospatial Side Fraction

Cancellation Total

Intracalcarine Cortex R 91.97%

Occipital Fusiform Gyrus R 64.63%
Lingual Gyrus R 61.17%
Supracalcarine Cortex R 58.53%
Temporal Occipital Fusiform Cortex R 53.53%
Lateral Occipital Cortex (superior division) R 42.82%
Temporal Fusiform Cortex (posterior division) R 39.63%
External capsule R 36.42%
Lateral Occipital Cortex (inferior division) R 36.27%
Posterior Limb of internal capsule R 35.39%
Posterior thalamic Radiation R 34.82%
Angular Gyrus R 32.23%
Middle Temporal Gyrus (posterior division) R 30.19%
Anterior Limb of internal capsule R 28.62%
Cingulum (hippocampus) R 28.51%
Supramarginal Gyrus (posterior division) R 27.61%
Parahippocampal Gyrus (posterior division) R 27.40%
Supramarginal Gyrus (anterior division) R 26.31%
Occipital Pole R 19.34%
Uncinate fasciculus R 19.32%
Sagittal stratum R 16.89%
Cuneal Cortex R 16.82%
Superior Temporal Gyrus (posterior division) R 16.25%
Cerebral peduncle R 15.90%
Inferior Temporal Gyrus (posterior division) R 15.65%
Parietal Operculum Cortex R 13.68%
Middle Frontal Gyrus L 12.51%
Anterior corona Radiata L 12.11%
Retrolenticular part of internal capsule R 11.90%
Superior Temporal Gyrus (anterior division) L 11.82%
Frontal Medial Cortex R 11.65%
Lateral Occipital Cortex (superior division) L 10.57%
Middle Temporal Gyrus (temporooccipital part) R 10.11%
Fornix (cres) Stria terminalis R 9.68%
Cingulate Gyrus (anterior division) L 9.63%
Planum Temporale R 8.61%
Tapetum R 8.30%
Subcallosal Cortex R 8.15%
(continued on next page)

11
M.J. Moore et al. Neuropsychologia 204 (2024) 109007

Table 3 (continued )
Visuospatial Side Fraction

Cancellation Total

External capsule L 6.77%

Precuneus Cortex R 6.56%
Insular Cortex R 6.19%
Heschl’s Gyrus R 5.96%
Inferior Temporal Gyrus (temporooccipital part) R 5.02%
Left Egocentric Neglect
Posterior Limb of internal capsule R 58.65%
Parietal Operculum Cortex R 44.48%
Intracalcarine Cortex R 37.47%
Central Opercular Cortex R 25.75%
Superior Longitudinal fasciculus R 25.32%
Supramarginal Gyrus (anterior division) R 21.29%
Postcentral Gyrus R 17.55%
Planum Temporale R 15.83%
Lateral Occipital Cortex (superior division) R 13.20%
Heschl’s Gyrus R 13.18%
Precentral Gyrus R 11.23%
Lingual Gyrus R 10.67%
Retrolenticular part of internal capsule R 10.59%
Supracalcarine Cortex R 9.80%
Inferior Frontal Gyrus (pars opercularis) R 9.65%
External capsule R 7.06%
Superior corona Radiata R 6.52%
Planum Polare R 5.27%
Visual Field Task
Intracalcarine Cortex R 95.55%
Supracalcarine Cortex R 62.17%
Lingual Gyrus R 58.95%
Parietal Operculum Cortex R 53.10%
Cuneal Cortex R 50.48%
Temporal Occipital Fusiform Cortex R 46.40%
Occipital Fusiform Gyrus R 45.61%
Posterior thalamic Radiation R 39.79%
Superior Temporal Gyrus (posterior division) R 33.84%
Heschl’s Gyrus R 27.29%
Planum Temporale R 26.82%
Temporal Occipital Fusiform Cortex L 26.26%
Supramarginal Gyrus (anterior division) R 26.22%
Intracalcarine Cortex L 23.90%
Lateral Occipital Cortex (superior division) R 20.39%
Middle Temporal Gyrus (temporooccipital part) R 19.59%
Supramarginal Gyrus (posterior division) R 19.13%
Temporal Fusiform Cortex (posterior division) R 17.36%
Angular Gyrus R 15.01%
Lingual Gyrus L 14.25%
Occipital Pole R 13.32%
Lateral Occipital Cortex (inferior division) R 11.30%
Precuneus Cortex R 10.83%
Occipital Fusiform Gyrus L 10.28%
Parahippocampal Gyrus (posterior division) R 7.35%
Middle Temporal Gyrus (posterior division) R 7.29%
Posterior thalamic Radiation L 5.90%

these deficits can occur from lesions impacting either hemisphere. suggested that this ability relies upon a fronto-parietal network
Rather than representing an asymmetry in the correlates of visual field including the fronto-parietal association fibres, superior longitudinal
deficits, these more lateralised results are likely caused by bias in lesion fasciculus, and the external capsule white matter (Moeller et al., 2015).
mapping results due to the comparative imbalance of patients exhibiting While some previous studies have reported that connectivity between
visual field impairments following left- and right-hemisphere lesions in left and right hemispheres plays a key role in calculation tasks (Cho
the study sample (see Fig. 1). Notably, network-level analysis success- et al., 2012; Park et al., 2013), previously reported correlates of nu-
fully identified bilateral correlates of visual field deficits, suggesting that merical cognition have primarily been localised to the left hemisphere
this approach may help address some key limitations present in (Benn et al., 2012; Moeller et al., 2015). The results of this study are
voxel-wise analyses. broadly in line with this previous work. However, given the diversity of
Very little network-level or multivariate lesion mapping work specific error patterns which could lead to impairment on numerical
exploring the neural correlates of numerical processing has been con- cognition tasks (e.g., Haupt et al., 2017; McCloskey et al., 1985), more
ducted. Past univariate lesion mapping analyses have linked OCS detailed lesion mapping work is needed to identify neural correlates of
calculation impairments to left parietal operculum damage, and number more specific impairment patterns.
writing impairments to voxels located in the left central/parietal oper- No significant correlates of impairment on the Trail Making Task
culum, insula, Heschl’s gyrus, and underlying white matter (Moore and were identified within either network-level and multivariate lesion
Demeyere, 2022). Previous work examining patterns of brain connec- mapping analyses. There are several potential explanations for this
tivity supporting numerical cognition in healthy individuals has finding. First, executive function is a diverse cognitive function in which

12
M.J. Moore et al. Neuropsychologia 204 (2024) 109007

a number of dissociable impairments (e.g., inhibition, set-switching, that this inherent noise may have biased reported anatomical correlates.
working memory) could result in impairment on the OCS trail making Past research has suggested that distinct patterns of performance on OCS
task (Muir et al., 2015; Varjacic et al., 2018; Varjačić et al., 2018). Given subtests may be explained by reduced number of underlying dimensions
that the anatomy of each of these constituent abilities is diverse, it is and that these dimensions may map on to distinct patterns of lesion
plausible that performance on this single task may not map onto a damage (Corbetta et al., 2018; Sperber et al., 2023). This possibility is
distinct pattern of lesion damage (Cipolotti et al., 2016; Moore et al., not explored in the present study but represents an important topic for
2023a,2024; Varjacic et al., 2018). In line with this possibility, past future research to explore. Despite our study’s large sample size, not all
research has shown that executive impairment is strongly associated potentially relevant brain regions are impacted by a sufficient number of
with non-lesion related factors (e.g., general cortical atrophy severity, lesions to enable statistical analyses (see Fig. 1). The lesion distribution
white matter integrity, pre-morbid cognitive decline) (Hobden et al., is representative of the stroke population (e.g. primarily MCA strokes)
2022, 2023; Kirova et al., 2015; Rost et al., 2022). These results high- but has comparatively lower power to detect contributions of neural
light that not all cognitive impairments should be expected to map onto regions supplied by the anterior cerebral artery, posterior cerebral ar-
distinct damage profiles in lesion mapping analyses. This is because tery, and anterior cerebral communicating artery.
some deficits may be linked to pre-morbid cognitive decline or
non-stroke related brain integrity issues including general cortical at- Funding
rophy or white matter lesions. Similarly, not all cognitive impairments
can be expected to map onto distinct patterns of lesion damage, and not JBM was supported by a National Health and Medical Research
all plausible underlying correlates can be effectively detected by even (NHMRC) Investigator Grant (Leadership 3; GNT2010141). ND is fun-
the most advanced lesion mapping methods (Moore et al., 2023a). These ded by the National Institute for Health Research (NIHR) (Advanced
possibilities are important to consider in the context of the null lesion Fellowship NIHR302224). The study was further supported by the Na-
mapping results reported in this, as well as in previous studies. tional Institute for Health Research (NIHR) Clinical Research Network
Considered cumulatively, the results of this study suggest that the and the NIHR Oxford Biomedical Research Centre (BRC) based at Oxford
OCS is able to tap complex brain-behaviour relationships. This finding University Hospitals NHS Trust andUniversity of Oxford. The views
further emphasises the value of standard, domain-specific post-stroke expressed are those of the author(s) and not necessarily those of the
cognitive assessment as this practice is able to capture complex NHS, the NIHR or the Department of Health.
functional-anatomical relationships which may be overlooked in popu-
lar domain-general cognitive screens (Demeyere et al., 2015, 2016,
Declaration of competing interest
2019). This practice is also valuable for clinically relevant lesion map-
ping investigations as these studies can employ comparatively simple
None.
and widely available cognitive data to complex brain-behaviour re-
lationships. Future studies should aim to investigate the prognostic
CRediT authorship contribution statement
utility of such complex brain-behaviour relationships in order to deter-
mine whether detailed lesion location metrics may help predict patient
Margaret Jane Moore: Writing – original draft, Visualization,
outcomes, over and above other established clinical and cognitive pre-
Software, Methodology, Investigation, Formal analysis, Data curation,
dictors (Bowren et al., 2022; Milosevich et al., 2024).
Conceptualization. Jason B. Mattingley: Writing – review & editing,
Supervision. Nele Demeyere: Writing – review & editing, Supervision,
4.1. Limitations
Conceptualization.
The results presented in this study highlight some of the limitations
Data availability
inherent in all lesion mapping methodologies, regardless of whether
univariate, network-level, or multivariate analyses are used. Univariate
All data (with the exception of lesion masks) is available here:
voxel-wise analyses are biased toward regions in which the lesions of
https://osf.io/depjt/. Lesion masks are available upon request.
impaired patients are spatially homologous, regardless of underlying
functional structures (Mah et al., 2014). Multivariate lesion mapping
approaches may help reduce this bias, but this more advanced approach Acknowledgements
does not appear to entirely negate this risk (Ivanova et al., 2021; Sperber
et al., 2019). Multivariate approaches can, in theory, identify significant We would like to thank the participants who took part in the Oxford
regions which are unrelated to the function of interest, but incidentally Cognitive Screening study and all research staff who contributed to-
help distinguish between spared and impaired patients (Moore et al., wards data collection. In particular, we acknowledge the contributions
2023a). Multivariate analyses are also intended to identify the set of to data collection and curation for the OCS data made by Ms Ellie
regions which most effectively distinguish between spared and impaired Slavkova, Ms Grace Chiu, and Ms Romina Basting.
patients, rather than capturing the full lesion location variability present
in the lesion distributions of impaired patients (Pustina et al., 2018). Appendix A. Supplementary data
Similarly, mass univariate network-level analyses are unable to distin-
guish between edges which are frequently damaged in deficit-causing Supplementary data to this article can be found online at https://doi.
lesions and edges which are critically associated with the impairment org/10.1016/j.neuropsychologia.2024.109007.
of interest (Mah et al., 2014). Overall, each lesion mapping method has
associated strengths and weaknesses. This investigation aims to combine References
the relative strengths and weaknesses of multiple methods in order to
provide a comprehensive investigation of brain-behaviour relationships. Ashburner, J., Barnes, G., Chen, C., Daunizeau, J., Flandin, G., Friston, K., Kiebel, S.,
Kilner, J., Litvak, V., Moran, R., 2016. SPM12 manual. URL: Http://Www. Fil. Ion.
Ideally, cognitive impairment should be diagnosed based on agree-
Ucl. Ac. Uk/Spm/Doc/Spm12 Manual. Pdf.
ment between multiple independent assessments (Moore et al., 2022; Bartolomeo, P., Thiebaut de Schotten, M., Doricchi, F., 2007. Left unilateral neglect as a
Quinn et al., 2021). As this study involved a retrospective analysis, this disconnection syndrome. Cerebr. Cortex 17 (11), 2479–2490. https://doi.org/
was not possible. Importantly, different underlying problems can result 10.1093/cercor/bhl181.
Bates, E., Wilson, S.M., Saygin, A.P., Dick, F., Sereno, M.I., Knight, R.T., Dronkers, N.F.,
in apparent cognitive impairment on any given task. This investigation’s 2003. Voxel-based lesion–symptom mapping. Nat. Neurosci. 6 (5). https://doi.org/
large sample size reduces, but does not entirely eliminate, the possibility 10.1038/nn1050. Article 5.

13
M.J. Moore et al. Neuropsychologia 204 (2024) 109007

Benn, Y., Zheng, Y., Wilkinson, I.D., Siegal, M., Varley, R., 2012. Language in calculation: Huygelier, H., Schraepen, B., Demeyere, N., Gillebert, C.R., 2020. The Dutch version of
a core mechanism? Neuropsychologia 50 (1), 1–10. https://doi.org/10.1016/j. the Oxford Cognitive Screen (OCS-NL): normative data and their association with
neuropsychologia.2011.09.045. age and socio-economic status. Aging Neuropsychol. Cognit. 27 (5), 765–786.
Bisogno, A.L., Franco Novelletto, L., Zangrossi, A., De Pellegrin, S., Facchini, S., Basile, A. https://doi.org/10.1080/13825585.2019.1680598.
M., Baracchini, C., Corbetta, M., 2023. The Oxford cognitive screen (OCS) as an Ivanova, M.V., Herron, T.J., Dronkers, N.F., Baldo, J.V., 2021. An empirical comparison
acute predictor of long-term functional outcome in a prospective sample of stroke of univariate versus multivariate methods for the analysis of brain–behavior
patients. Cortex 166, 33–42. https://doi.org/10.1016/j.cortex.2023.04.015. mapping. Hum. Brain Mapp. 42 (4), 1070–1101.
Boes, A.D., Prasad, S., Liu, H., Liu, Q., Pascual-Leone, A., Caviness Jr., V.S., Fox, M.D., Kirova, A.-M., Bays, R.B., Lagalwar, S., 2015. Working memory and executive function
2015. Network localization of neurological symptoms from focal brain lesions. Brain decline across normal aging, mild cognitive impairment, and Alzheimer’s disease.
138 (10), 3061–3075. https://doi.org/10.1093/brain/awv228. BioMed Res. Int. 2015.
Bormann, T., Kaller, C.P., Kulyk, C., Demeyere, N., Weiller, C., 2024. The German version Kong, A.P.-H., Lam, P.H.-P., Ho, D.W.-L., Lau, J.K., Humphreys, G.W., Riddoch, J.,
of the Oxford Cognitive Screen (D-OCS): normative data and validation in acute Weekes, B., 2016. The Hong Kong version of the Oxford Cognitive Screen (HK-OCS):
stroke and a mixed neurological sample. J. Neuropsychol. 12359. https://doi.org/ validation study for Cantonese-speaking chronic stroke survivors. In:
10.1111/jnp.12359 jnp. Neuropsychology, Development, and Cognition. Section B, Aging, Neuropsychology
Bowren, M., Bruss, J., Manzel, K., Edwards, D., Liu, C., Corbetta, M., Tranel, D., Boes, A. and Cognition, vol. 23, pp. 530–548. https://doi.org/10.1080/
D., 2022. Post-stroke outcomes predicted from multivariate lesion-behaviour and 13825585.2015.1127321, 5.
lesion network mapping. Brain 145 (4), 1338–1353. Lugtmeijer, S., Geerligs, L., de Leeuw, F.E., de Haan, E.H.F., Kessels, R.P.C., Smits, A.R.,
Bridge, H., 2020. Loss of visual cortex and its consequences for residual vision. Current Schmand, B.A., de Haan, E.H.F., de Leeuw, F.E., Luijckx, G. jan, Scholte, H.S.,
Opinion in Physiology 16, 21–26. Spikman, J.M., Kappelle, L.J., Geerligs, L., van Zandvoort, M.J.E., Caan, M.W.A.,
Chechlacz, M., Rotshtein, P., Bickerton, W.-L., Hansen, P.C., Deb, S., Humphreys, G.W., Raemaekers, M.A.H.L.L., Prokop, M., Ramsey, N.F., on behalf of The Visual Brain
2010. Separating neural correlates of allocentric and egocentric neglect: distinct Group, 2021. Are visual working memory and episodic memory distinct processes?
cortical sites and common white matter disconnections. Cogn. Neuropsychol. 27 (3), Insight from stroke patients by lesion-symptom mapping. Brain Struct. Funct. 226
277–303. https://doi.org/10.1080/02643294.2010.519699. (6), 1713–1726. https://doi.org/10.1007/s00429-021-02281-0.
Chechlacz, M., Rotshtein, P., Humphreys, G.W., 2012. Neuroanatomical dissections of Mah, Y.-H., Husain, M., Rees, G., Nachev, P., 2014. Human brain lesion-deficit inference
unilateral visual neglect symptoms: ALE meta-analysis of lesion-symptom mapping. remapped. Brain 137 (9), 2522–2531. https://doi.org/10.1093/brain/awu164.
Front. Hum. Neurosci. 6. https://doi.org/10.3389/fnhum.2012.00230. Mah, Y.-H., Jha, A., Xu, T., Nachev, P., 2022. Multivariate lesion-behavior mapping. In:
Cho, S., Metcalfe, A.W.S., Young, C.B., Ryali, S., Geary, D.C., Menon, V., 2012. Pustina, D., Mirman, D. (Eds.), Lesion-to-Symptom Mapping: Principles and Tools.
Hippocampal–prefrontal engagement and dynamic causal interactions in the Springer US, pp. 199–218. https://doi.org/10.1007/978-1-0716-2225-4_11.
maturation of children’s fact retrieval. J. Cognit. Neurosci. 24 (9), 1849–1866. McCloskey, M., Caramazza, A., Basili, A., 1985. Cognitive mechanisms in number
https://doi.org/10.1162/jocn_a_00246. processing and calculation: evidence from dyscalculia. Brain Cognit. 4 (2), 171–196.
Cipolotti, L., Spanò, B., Healy, C., Tudor-Sfetea, C., Chan, E., White, M., Biondo, F., https://doi.org/10.1016/0278-2626(85)90069-7.
Duncan, J., Shallice, T., Bozzali, M., 2016. Inhibition processes are dissociable and Milosevich, E.T., Moore, M.J., Pendlebury, S.T., Demeyere, N., 2024. Domain-specific
lateralized in human prefrontal cortex. Neuropsychologia 93 (Pt A), 1–12. https:// cognitive impairment 6 months after stroke: the value of early cognitive screening.
doi.org/10.1016/j.neuropsychologia.2016.09.018. International Journal of Stroke 19 (3), 331–341.
Corbetta, M., Shulman, G.L., 2011. Spatial neglect and attention networks. Annu. Rev. Mirman, D., Chen, Q., Zhang, Y., Wang, Z., Faseyitan, O.K., Coslett, H.B., Schwartz, M.F.,
Neurosci. 34 (1), 569–599. https://doi.org/10.1146/annurev-neuro-061010- 2015. Neural organization of spoken language revealed by lesion–symptom
113731. mapping. Nat. Commun. 6 (1). https://doi.org/10.1038/ncomms7762. Article 1.
Corbetta, M., Siegel, J.S., Shulman, G.L., 2018. On the low dimensionality of behavioral Mirman, D., Landrigan, J.-F., Kokolis, S., Verillo, S., Ferrara, C., Pustina, D., 2018.
deficits and alterations of brain network connectivity after focal injury. Cortex 107, Corrections for multiple comparisons in voxel-based lesion-symptom mapping.
229–237. Neuropsychologia 115, 112–123. https://doi.org/10.1016/j.
de Haan, B., Karnath, H.-O., 2018. A hitchhiker’s guide to lesion-behaviour mapping. neuropsychologia.2017.08.025.
Neuropsychologia 115, 5–16. https://doi.org/10.1016/j. Mock, N., Balzer, C., Gutbrod, K., De Haan, B., Jäncke, L., Ettlin, T., Trost, W., 2022.
neuropsychologia.2017.10.021. Lesion-symptom mapping corroborates lateralization of verbal and nonverbal
Deary, I.J., Corley, J., Gow, A.J., Harris, S.E., Houlihan, L.M., Marioni, R.E., Penke, L., memory processes and identifies distributed brain networks responsible for memory
Rafnsson, S.B., Starr, J.M., 2009. Age-associated cognitive decline. Br. Med. Bull. 92 dysfunction. Cortex 153, 178–193. https://doi.org/10.1016/j.cortex.2022.04.017.
(1), 135–152. Moeller, K., Willmes, K., Klein, E., 2015. A review on functional and structural brain
Demeyere, N., Riddoch, M.J., Slavkova, E.D., Bickerton, W.-L., Humphreys, G.W., 2015. connectivity in numerical cognition. Front. Hum. Neurosci. 9. https://www.fronti
The Oxford Cognitive Screen (OCS): validation of a stroke-specific short cognitive ersin.org/articles/10.3389/fnhum.2015.00227.
screening tool. Psychol. Assess. 27 (3), 883–894. https://doi.org/10.1037/ Moore, M.J., 2022. A practical guide to lesion symptom mapping. PsyArXiv. https://doi.
pas0000082. org/10.31234/osf.io/2jxr9.
Demeyere, N., Riddoch, M.J., Slavkova, E.D., Jones, K., Reckless, I., Mathieson, P., Moore, M.J., Byrne, J., Gibson, E.C., Ford, L., Robinson, G.A., 2023a. Hayling and Stroop
Humphreys, G.W., 2016. Domain-specific versus generalized cognitive screening in Tests Tap Dissociable Deficits and Network-Level Neural Correlates. Brain Structure
acute stroke. J. Neurol. 263 (2), 306–315. https://doi.org/10.1007/s00415-015- and Function 229 (4), 879–896.
7964-4. Moore, M.J., Demeyere, N., 2022. Lesion symptom mapping of domain-specific cognitive
Demeyere, N., Sun, S., Milosevich, E., Vancleef, K., 2019. Post-stroke cognition with the impairments using routine imaging in stroke. Neuropsychologia 167, 108159.
Oxford cognitive screen vs Montreal cognitive assessment: a multi-site randomized https://doi.org/10.1016/j.neuropsychologia.2022.108159.
controlled study (OCS-care). AMRC Open Research 1 (12), 12. Moore, M.J., Demeyere, N., Rorden, C., Mattingley, J.B., 2023a. Lesion mapping in
Ghaleh, M., Lacey, E.H., Fama, M.E., Anbari, Z., DeMarco, A.T., Turkeltaub, P.E., 2020. neuropsychological research: a practical and conceptual guide. Cortex.
Dissociable mechanisms of verbal working memory revealed through multivariate Moore, M.J., Gillebert, C.R., Demeyere, N., 2021. Right and left neglect are not
lesion mapping. Cerebr. Cortex 30 (4), 2542–2554. anatomically homologous: a voxel-lesion symptom mapping study.
Gleichgerrcht, E., Fridriksson, J., Rorden, C., Bonilha, L., 2017. Connectome-based Neuropsychologia 162, 108024.
lesion-symptom mapping (CLSM): a novel approach to map neurological function. Moore, M.J., Hearne, L., Demeyere, N., Mattingley, J.B., 2023b. Comprehensive voxel-
Neuroimage: Clinical 16, 461–467. wise, tract-based and network lesion mapping reveals unique architectures of right
Griffis, J.C., Metcalf, N.V., Corbetta, M., Shulman, G.L., 2021. Lesion Quantification and left visuospatial neglect.
Toolkit: a MATLAB software tool for estimating grey matter damage and white Moore, M.J., Jenkinson, M., Griffanti, L., Huygelier, H., Gillebert, C.R., Demeyere, N.,
matter disconnections in patients with focal brain lesions. Neuroimage: Clinical 30, 2023c. A comparison of lesion mapping analyses based on CT versus MR imaging in
102639. stroke. Neuropsychologia 184, 108564.
Haupt, M., Gillebert, C.R., Demeyere, N., 2017. The zero effect: voxel-based lesion Moore, M.J., Milosevich, E., Beisteiner, R., Bowen, A., Checketts, M., Demeyere, N.,
symptom mapping of number transcoding errors following stroke. Sci. Rep. 7 (1), Fordell, H., Godefroy, O., Laczo, J., Rich, T., 2022. Rapid screening for neglect
9242. https://doi.org/10.1038/s41598-017-08728-x. following stroke: A systematic search and European Academy of Neurology
Heilman, K.M., Adams, D.J., 2003. Callosal neglect. Arch. Neurol. 60 (2), 276–279. recommendations.
https://doi.org/10.1001/archneur.60.2.276. Moore, M.J., Milosevich, E., Mattingley, J.B., Demeyere, N., 2023d. The neuroanatomy
Herbet, G., Lafargue, G., Duffau, H., 2015. Rethinking voxel-wise lesion-deficit analysis: of visuospatial neglect: a systematic review and analysis of lesion-mapping
a new challenge for computational neuropsychology. Cortex 64, 413–416. methodology. Neuropsychologia 108470.
Hobden, G., Moore, M.J., Colbourne, E., Pendlebury, S.T., Demeyere, N., 2023. Mori, S., Wakana, S., Van Zijl, P.C., Nagae-Poetscher, L.M., 2005. MRI Atlas of Human
Association of neuroimaging markers on clinical CT scans with domain-specific White Matter. Elsevier.
cognitive impairment in the early and later poststroke stages. Neurology 101 (17), Muir, Ryan T., Lam, Benjamin, Honjo Kie, Harry, Robin D., McNeely, Alicia A., Gao, Fu-
e1687–e1696. Qiang, Ramirez, Joel, Scott, Christopher J.M., Ganda, Anoop, Zhao, Jiali, Zhou, X.
Hobden, G., Moore, M., Mair, G., Pendlebury, S., Demeyere, N., 2022. Post-stroke Joe, Graham Simon, J., Rangwala, Novena, Gibson, Erin, Lobaugh, Nancy J.,
executive function in relation to white matter damage on clinically acquired CT Kiss, Alex, Stuss, Donald T., Nyenhuis, David L., Lee, Byung-Chul, Black, Sandra E.,
brain imaging. https://doi.org/10.1101/2021.11.12.21266247 (p. 2015. Trail making test elucidates neural substrates of specific poststroke executive
2021.11.12.21266247). medRxiv. dysfunctions. Stroke 46 (10), 2755–2761. https://doi.org/10.1161/
Hoeren, M., Kümmerer, D., Bormann, T., Beume, L., Ludwig, V.M., Vry, M.-S., Mader, I., STROKEAHA.115.009936.
Rijntjes, M., Kaller, C.P., Weiller, C., 2014. Neural bases of imitation and pantomime Park, J., Park, D.C., Polk, T.A., 2013. Parietal functional connectivity in numerical
in acute stroke patients: distinct streams for praxis. Brain 137 (10), 2796–2810. cognition. Cerebr. Cortex 23 (9), 2127–2135. https://doi.org/10.1093/cercor/
https://doi.org/10.1093/brain/awu203. bhs193.

14
M.J. Moore et al. Neuropsychologia 204 (2024) 109007

Peer, M., Salomon, R., Goldberg, I., Blanke, O., Arzy, S., 2015. Brain system for mental Sperber, C., Gallucci, L., Umarova, R., 2023. The low dimensionality of post-stroke
orientation in space, time, and person. Proc. Natl. Acad. Sci. USA 112 (35), cognitive deficits: it’s the lesion anatomy. Brain 146 (6), 2443–2452.
11072–11077. https://doi.org/10.1073/pnas.1504242112. Sperber, C., Wiesen, D., Karnath, H.-O., 2019. An empirical evaluation of multivariate
Philippi, C.L., Bruss, J., Boes, A.D., Albazron, F.M., Deifelt Streese, C., Ciaramelli, E., lesion behaviour mapping using support vector regression. Hum. Brain Mapp. 40 (5),
Rudrauf, D., Tranel, D., 2021. Lesion network mapping demonstrates that mind- 1381–1390.
wandering is associated with the default mode network. J. Neurosci. Res. 99 (1), Swienton, D.J., Thomas, A.G., 2014. The visual pathway—functional anatomy and
361–373. https://doi.org/10.1002/jnr.24648. pathology. Seminars Ultrasound, CT MRI 35 (5), 487–503.
Pustina, D., Avants, B., Faseyitan, O.K., Medaglia, J.D., Coslett, H.B., 2018. Improved Thiebaut de Schotten, M., Kinkingnéhun, S., Delmaire, C., Lehéricy, S., Duffau, H.,
accuracy of lesion to symptom mapping with multivariate sparse canonical Thivard, L., Volle, E., Levy, R., Dubois, B., Bartolomeo, P., 2008. Visualization of
correlations. Neuropsychologia 115, 154–166. disconnection syndromes in humans. Cortex 44 (8), 1097–1103. https://doi.org/
Quinn, T.J., Richard, E., Teuschl, Y., Gattringer, T., Hafdi, M., O’Brien, J.T., 10.1016/j.cortex.2008.02.003.
Merriman, N., Gillebert, C., Huygelier, H., Verdelho, A., Schmidt, R., Ghaziani, E., Tsuchida, A., Fellows, L.K., 2013. Are core component processes of executive function
Forchammer, H., Pendlebury, S.T., Bruffaerts, R., Mijajlovic, M., Drozdowska, B.A., dissociable within the frontal lobes? Evidence from humans with focal prefrontal
Ball, E., Markus, H.S., 2021. European Stroke Organisation and European Academy damage. Cortex 49 (7), 1790–1800. https://doi.org/10.1016/j.cortex.2012.10.014.
of Neurology joint guidelines on post-stroke cognitive impairment. Eur. J. Neurol. 28 Tzourio-Mazoyer, N., Landeau, B., Papathanassiou, D., Crivello, F., Etard, O.,
(12), 3883–3920. https://doi.org/10.1111/ene.15068. Delcroix, N., Mazoyer, B., Joliot, M., 2002. Automated anatomical labeling of
Robotham, R.J., Riis, J.O., Demeyere, N., 2020. A Danish version of the Oxford cognitive activations in SPM using a macroscopic anatomical parcellation of the MNI MRI
screen: a stroke-specific screening test as an alternative to the MoCA. Aging single-subject brain. Neuroimage 15 (1), 273–289. https://doi.org/10.1006/
Neuropsychol. Cognit. 27 (1), 52–65. https://doi.org/10.1080/ nimg.2001.0978.
13825585.2019.1577352. Varjacic, A., Mantini, D., Demeyere, N., Gillebert, C.R., 2018. Neural signatures of Trail
Rorden, C., Bonilha, L., Fridriksson, J., Bender, B., Karnath, H.-O., 2012. Age-specific CT Making Test performance: evidence from lesion-mapping and neuroimaging studies.
and MRI templates for spatial normalization. Neuroimage 61 (4), 957–965. https:// Neuropsychologia 115, 78–87. https://doi.org/10.1016/j.
doi.org/10.1016/j.neuroimage.2012.03.020. neuropsychologia.2018.03.031.
Rorden, C., Karnath, H.-O., 2010. A simple measure of neglect severity. Varjačić, A., Mantini, D., Levenstein, J., Slavkova, E.D., Demeyere, N., Gillebert, C.R.,
Neuropsychologia 48 (9), 2758–2763. https://doi.org/10.1016/j. 2018. The role of left insula in executive set-switching: lesion evidence from an acute
neuropsychologia.2010.04.018. stroke cohort. Cortex; a Journal Devoted to the Study of the Nervous System and
Rosenzopf, H., Wiesen, D., Basilakos, A., Yourganov, G., Bonilha, L., Rorden, C., Behavior 107, 92–101. https://doi.org/10.1016/j.cortex.2017.11.009.
Fridriksson, J., Karnath, H.-O., Sperber, C., 2022. Mapping the human praxis Weaver, N.A., Kuijf, H.J., Aben, H.P., Abrigo, J., Bae, H.-J., Barbay, M., Best, J.G.,
network: an investigation of white matter disconnection in limb apraxia of gesture Bordet, R., Chappell, F.M., Chen, C.P.L.H., Dondaine, T., van der Giessen, R.S.,
production. Brain Communications 4 (1), fcac004. https://doi.org/10.1093/ Godefroy, O., Gyanwali, B., Hamilton, O.K.L., Hilal, S., Huenges Wajer, I.M.C.,
braincomms/fcac004. Kang, Y., Kappelle, L.J., et al., 2021. Strategic infarct locations for post-stroke
Rost, N.S., Brodtmann, A., Pase, M.P., van Veluw, S.J., Biffi, A., Duering, M., Hinman, J. cognitive impairment: a pooled analysis of individual patient data from 12 acute
D., Dichgans, M., 2022. Post-stroke cognitive impairment and dementia. Circ. Res. ischaemic stroke cohorts. Lancet Neurol. 20 (6), 448–459. https://doi.org/10.1016/
130 (8), 1252–1271. https://doi.org/10.1161/CIRCRESAHA.122.319951. S1474-4422(21)00060-0.
Sagnier, S., Munsch, F., Bigourdan, A., Debruxelles, S., Poli, M., Renou, P., Olindo, S., Wiesen, D., Sperber, C., Yourganov, G., Rorden, C., Karnath, H.-O., 2019. Using machine
Rouanet, F., Dousset, V., Tourdias, T., Sibon, I., 2019. The influence of stroke learning-based lesion behavior mapping to identify anatomical networks of cognitive
location on cognitive and mood impairment. A voxel-based lesion-symptom mapping dysfunction: spatial neglect and attention. Neuroimage 201, 116000. https://doi.
study. J. Stroke Cerebrovasc. Dis.: The Official Journal of National Stroke org/10.1016/j.neuroimage.2019.07.013.
Association 28 (5), 1236–1242. https://doi.org/10.1016/j. Yeh, F.-C., Panesar, S., Fernandes, D., Meola, A., Yoshino, M., Fernandez-Miranda, J.C.,
jstrokecerebrovasdis.2019.01.010. Vettel, J.M., Verstynen, T., 2018. Population-averaged atlas of the macroscale
Salvalaggio, A., De Filippo De Grazia, M., Zorzi, M., Thiebaut de Schotten, M., human structural connectome and its network topology. Neuroimage 178, 57–68.
Corbetta, M., 2020. Post-stroke deficit prediction from lesion and indirect structural https://doi.org/10.1016/j.neuroimage.2018.05.027.
and functional disconnection. Brain 143 (7), 2173–2188. https://doi.org/10.1093/ Yourganov, G., Fridriksson, J., Rorden, C., Gleichgerrcht, E., Bonilha, L., 2016.
brain/awaa156. Multivariate connectome-based symptom mapping in post-stroke patients: networks
Saxena, S., Keser, Z., Rorden, C., Bonilha, L., Fridriksson, J., Walker, A., Hillis, A.E., supporting language and speech. J. Neurosci. 36 (25), 6668–6679.
2022. Disruptions of the human connectome associated with hemispatial neglect. Zhang, X., Kedar, S., Lynn, M., Newman, N., Biousse, V., 2006. Homonymous hemianopia
Neurology 98 (2), e107–e114. https://doi.org/10.1212/WNL.0000000000013050. in stroke. J. Neuro Ophthalmol. 26 (3), 180–183. https://doi.org/10.1097/01.
Schaefer, A., Kong, R., Gordon, E.M., Laumann, T.O., Zuo, X.-N., Holmes, A.J., wno.0000235587.41040.39.
Eickhoff, S.B., Yeo, B.T.T., 2018. Local-global parcellation of the human cerebral Zhang, Y., Kimberg, D.Y., Coslett, H.B., Schwartz, M.F., Wang, Z., 2014. Multivariate
cortex from intrinsic functional connectivity MRI. Cerebr. Cortex 28 (9), 3095–3114. lesion-symptom mapping using support vector regression. Hum. Brain Mapp. 35
https://doi.org/10.1093/cercor/bhx179. (12), 5861–5876.
Schumacher, R., Halai, A.D., Lambon Ralph, M.A., 2019. Assessing and mapping
language, attention and executive multidimensional deficits in stroke aphasia. Brain
142 (10), 3202–3216. https://doi.org/10.1093/brain/awz258.

15