Resting-State Beta-Band Recovery Network Related to Cognitive Improvement After Stroke

Abstract

Background
Stroke is the second leading cause of death worldwide and it causes important long-term cognitive and physical deficits that hamper patients' daily activity. Neuropsychological rehabilitation (NR) has increasingly become more important to recover from cognitive disability and to improve the functionality and quality of life of these patients. Since in most stroke cases, restoration of functional connectivity (FC) precedes or accompanies cognitive and behavioral recovery, understanding the electrophysiological signatures underlying stroke recovery mechanisms is a crucial scientific and clinical goal.

Methods
For this purpose, a longitudinal study was carried out with a sample of 10 stroke patients, who underwent two neuropsychological assessments and two resting-state magnetoencephalographic (MEG) recordings, before and after undergoing a NR program. Moreover, to understand the degree of cognitive and neurophysiological impairment after stroke and the mechanisms of recovery after cognitive rehabilitation, stroke patients were compared to 10 healthy controls matched for age, sex, and educational level.

Findings
After intra and inter group comparisons, we found the following results: (1) Within the stroke group who received cognitive rehabilitation, almost all cognitive domains improved relatively or totally; (2) They exhibit a pattern of widespread increased in FC within the beta band that was related to the recovery process (there were no significant differences between patients who underwent rehabilitation and controls); (3) These FC recovery changes were related with the enhanced of cognitive performance. Furthermore, we explored the capacity of the neuropsychological scores before rehabilitation, to predict the FC changes in the brain network. Significant correlations were found in global indexes from the WAIS-III: Performance IQ (PIQ) and Perceptual Organization index (POI) (i.e., Picture Completion, Matrix Reasoning, and Block Design).

Full text

ORIGINAL RESEARCH
published: 25 February 2022
doi: 10.3389/fneur.2022.838170
Frontiers in Neurology | www.frontiersin.org 1February 2022 | Volume 13 | Article 838170
Edited by:
Rafeed Alkawadri,
University of Pittsburgh Medical
Center, United States
Reviewed by:
Shennan Aibel Weiss,
SUNY Downstate Medical Center,
United States
Donna Clark Tippett,
Johns Hopkins Medicine,
United States
*Correspondence:
Sandra Pusil
spusil@ucm.es
†These authors have contributed
equally to this work and share first
authorship
Specialty section:
This article was submitted to
Applied Neuroimaging,
a section of the journal
Frontiers in Neurology
Received: 17 December 2021
Accepted: 03 February 2022
Published: 25 February 2022
Citation:
Pusil S, Torres-Simon L, Chino B,
López ME, Canuet L, Bilbao Á,
Maestú F and Paúl N (2022)
Resting-State Beta-Band Recovery
Network Related to Cognitive
Improvement After Stroke.
Front. Neurol. 13:838170.
doi: 10.3389/fneur.2022.838170
Resting-State Beta-Band Recovery
Network Related to Cognitive
Improvement After Stroke
Sandra Pusil1*†, Lucía Torres-Simon1†, Brenda Chino 2, María Eugenia López 1,
Leonides Canuet1, Álvaro Bilbao3, Fernando Maestú1and Nuria Paúl1
1Department of Experimental Psychology, Universidad Complutense de Madrid, Madrid, Spain, 2Institute of Neuroscience,
Autonomous University of Barcelona, Barcelona, Spain, 3National Centre for Brain Injury Treatment, Centro de Referencia
Estatal de Atención Al Daño Cerebral (CEADAC), Madrid, Spain
Background: Stroke is the second leading cause of death worldwide and it causes
important long-term cognitive and physical deficits that hamper patients’ daily activity.
Neuropsychological rehabilitation (NR) has increasingly become more important to
recover from cognitive disability and to improve the functionality and quality of life
of these patients. Since in most stroke cases, restoration of functional connectivity
(FC) precedes or accompanies cognitive and behavioral recovery, understanding the
electrophysiological signatures underlying stroke recovery mechanisms is a crucial
scientific and clinical goal.
Methods: For this purpose, a longitudinal study was carried out with a sample
of 10 stroke patients, who underwent two neuropsychological assessments and two
resting-state magnetoencephalographic (MEG) recordings, before and after undergoing
a NR program. Moreover, to understand the degree of cognitive and neurophysiological
impairment after stroke and the mechanisms of recovery after cognitive rehabilitation,
stroke patients were compared to 10 healthy controls matched for age, sex, and
educational level.
Findings: After intra and inter group comparisons, we found the following results:
(1) Within the stroke group who received cognitive rehabilitation, almost all cognitive
domains improved relatively or totally; (2) They exhibit a pattern of widespread increased
in FC within the beta band that was related to the recovery process (there were no
significant differences between patients who underwent rehabilitation and controls); (3)
These FC recovery changes were related with the enhanced of cognitive performance.
Furthermore, we explored the capacity of the neuropsychological scores before
rehabilitation, to predict the FC changes in the brain network. Significant correlations
were found in global indexes from the WAIS-III: Performance IQ (PIQ) and Perceptual
Organization index (POI) (i.e., Picture Completion, Matrix Reasoning, and Block Design).
Keywords: stroke, functional connectivity (FC), MEG (magnetoencephalography), cognitive performance,
neuropsychological rehabilitation

Pusil et al. Functional Recovery Network After Stroke
INTRODUCTION
Stroke is considered the second leading cause of death and
the third leading cause of disability worldwide (1). It is a
heterogeneous pathology with diverse clinical manifestations due
to its possible etiologies (i.e., hemorrhagic, or ischemic), locations
(i.e., different vascular vessels or arteries), and size of the lesion (2,
3). However, most stroke survivors suffer from different degrees
of cognitive disabilities (4–6). These patients may have damage
in general cognitive performance with important functional
disability, which has been broadly reported in the scientific
literature (7–9). Although stroke tends to impact on attention
and executive function compared with its impact on memory,
a malfunction in these cognitive domains could worsen the
performance in other cognitive areas (3,4,6,10). In any case,
it is important to highlight the role of non-pharmacological
rehabilitation especially neuropsychological rehabilitation (NR)
in order to improve cognitive abilities and daily functions (10,
11). Neuropsychological rehabilitation is a systematic therapeutic
activity oriented functionally based on the assessment and
understanding of the cognitive deficits, emotional disturbances,
disruptive behaviors, and functional disorders of patients (12,
13), and includes interventions that might be compensatory,
educational, or restorative (10).
According to some authors and approaches post-stroke
deficits have long been considered to be fundamentally associated
with the location of the lesion (14). This could be particularly true
for sensorimotor or language deficits, which are closely related to
the damage to the specific eloquent cortex. However, it has been
shown that although structural damage from stroke is usually
focal, remote disturbances may occur in brain distant regions
from the primary area of damage (15,16). This phenomenon
was previously associated with the concept of diaschisis, but
it is currently explained by the disruption of structural and
functional connectivity (FC) between brain areas (17). This way
of understanding the functioning of the brain gives a crucial
role to NR in the process of cognitive recovery, since it allows a
holistic management of cognitive impairment in contrast to other
more goal-oriented therapies.
In this context, as the restoration of FC precedes or
accompanies in most cases, cognitive and behavioral recovery in
stroke patients (18,19), understanding the electrophysiological
signatures underlying stroke recovery mechanisms is a crucial
scientific goal. This information could help the clinical
community to anticipate and modify NR programs to achieve
a more effective cognitive recovery, and consequently, improve
patients’ quality of life. With this purpose, in the present study we
used the magnetoencephalography (MEG), a neurophysiological
technique that allows a comprehensive analysis of brain dynamics
(20,21). While the functional magnetic resonance image (fMRI)
is intrinsically limited by the hemodynamic response, MEG
directly measures cortical neural activity. That means that the
modified vasomotor reactivity and neurovascular uncoupling in
stroke easily affects the blood oxygen level-dependent (BOLD)
response but leaves the MEG signal intact (22). The study of MEG
signatures is well-established for early detection and prognosis
in neurodegenerative disorders, such as multiple sclerosis (23)
or Alzheimer’s disease (24). Moreover, MEG has previously been
used to demonstrate the disruption and recovery of functional
networks, and even its relationship with cognitive improvement
after undergoing a NR program in acquired brain pathologies
such as stroke (22,25,26) or traumatic brain injury (TBI)
(13,27). Thanks to the relevance of the neurophysiological
changes found in previous literature, it seems plausible that MEG
may provide interesting information about how NR may induce
specific cognitive recovery in stroke patients.
According to the aforementioned antecedents, the aims of the
present exploratory study are: (1) to understand the cognitive
improvement achieved in stroke patients that received NR; (2) to
explore the possible neurophysiological mechanisms underlying
the recovery process, by using FC on frequency bands obtained
with MEG; and (3) to evaluate if these neurophysiological
changes are related with cognitive improvement. For this
purpose, we carried out a longitudinal study in which a sample
of 10 stroke patients (stroke patients) were examined at two
different time points. The first was before NR (from now on
we will say pre-condition), and the second was after NR (from
now on we will say post-condition). At both time points patients
were cognitively evaluated and underwent resting-state MEG
recordings. Moreover, to understand the degree of cognitive
and neurophysiological disruption after stroke, and the recovery
mechanisms in stroke patients who were enrolled on the NR, data
for a control group were included with 10 healthy controls paired
in age, sex, and educational level.
MATERIALS AND METHODS
Participants
The total dataset consisted of 20 subjects: 10 stroke patients (2
females/8 males; mean age 44.9 ±8.94; mean level of education
4.44 ±0.97) and 10 healthy controls (2 females/8 males; mean
age 43 ±12.72; mean level of education 4.78 ±0.67). The
mean time from the onset of the stroke to the start of the
study was 6.3 months, and the rehabilitation program lasted 7
months. The patient’s lesions were both ischemic (i.e., infarction;
n=5) and hemorrhagic (i.e., intracerebral hemorrhage; n=5)
and the stroke was located in different brain areas (for patient
detailed descriptive data see Table 1). To be enrolled in the study,
patients had to be diagnosed with a first-ever stroke, showing
a compatible lesion observed on computerized tomography
(CT) or magnetic resonance imaging (MRI). Although initially,
after the stroke some patients showed loss of consciousness [as
reported in Table 1 with the Glasgow Coma Scale (28)], at the
beginning of the study all patients were neurologically stable
without alterations in consciousness or alertness, and none of
them showed epileptiform discharges on MEG recordings.
Exclusion criteria were the following: a stroke involving
the brainstem or cerebellum, a diagnosis of neurological or
psychiatric diseases other than stroke, and a history of TBI, drug,
or alcohol abuse.
Patients were recruited from the National Brain Injury
Rehabilitation Center and from Lescer Brain Injury
Rehabilitation Center (Madrid, Spain), and all of them were
Frontiers in Neurology | www.frontiersin.org 2February 2022 | Volume 13 | Article 838170

Pusil et al. Functional Recovery Network After Stroke
TABLE 1 | Clinical and sociodemographic characteristics of the patients.
Patient Age Sex Education GCS Stroke etiology Stroke lesion
1 44 M 3 12 Ischemia Right fronto-parietal
2 45 F 5 7 Ischemia Right middle cerebral artery
3 47 M 5 9 Ischemia Left middle cerebral artery
4 47 M 4 12 Ischemia Right middle and anterior cerebral arteries
5 60 M 3 12 Ischemia Left middle cerebral artery
6 28 F 4 7 Hemorrhage Thalamus and left basal ganglia
7 35 M 4 8 Hemorrhage Right intraparenchymal
8 41 M 5 7 Hemorrhage Left basal ganglia
9 49 M 6 9 Hemorrhage Right basal ganglia
10 53 M 5 9 Hemorrhage Left thalamus
N=10 44.9 ±8.9 8 M/2 F 4.4 ±0.9 9.2 ±2.1 5 isch/5 hem
Education (1, illiterate/functional illiterate; 2, elemental studies; 3, school graduate; 4, high school studies; 5, university graduate studies; 6, university post-graduate studies).
enrolled in a NR program. Healthy controls were matched with
patients for age, sex, and education level, and they did not have a
previous history of psychiatric or neurological disorders.
As previously mentioned, patients underwent MEG
recordings and neuropsychological evaluation in two different
moments: (1) Pre-condition (Pre): at the beginning of the
study, before NR program; and (2) Post-condition (Post): after
completing the NR program. In the case of healthy controls, both
data, neuropsychological and neurophysiological, were obtained
only once, at the beginning of the study.
Ethics Statement
Methods were carried out in accordance with approved
guidelines and regulations. The study was approved by the
National Brain Injury Rehabilitation Center Ethics Committee
(Madrid), and all participants or legal representatives signed a
written informed consent prior to participation.
Neuropsychological Assessment
All participants underwent a comprehensive neuropsychological
evaluation with the aim to identify their cognitive status
in multiple cognitive domains (attention, memory, language,
executive functions, and visuospatial abilities) as well as
their functional performance. The extensive neuropsychological
assessment included: the Wechsler Adult Intelligence Scale III
(WAIS III) (29), the Brief Test of Attention (BTA) (30), the
Trail Making Test (TMT) (31), the Stroop Color Word Test
(32,33), the Wisconsin Card Sorting Test (WCST) (34), the
Tower of Hanoi (35), the Zoo Map Test [from the Behavioral
Assessment of the Dysexecutive Syndrome (36)], the Boston
Naming Test (BNT) (37), the Digit Span Test [Wechsler Memory
Scale III (29)], the Visual Span Test [WMS-III; (29)], Logical
Memory and Visual Reproduction [WMS-III (29)], the Phonemic
and Semantic Fluency [Controlled Oral Word Association Test,
COWAT (38)], the Five Digit Test [FDT (39)], the Dysexecutive
Questionnaire [DEX (36)], and the Patient Competency Rating
Scale [PCRS (40)].
Neuropsychological Rehabilitation
Program
All stroke patients received an integrated treatment based on
the holistic-comprehensive model proposed by Ben-Yishay and
Diller (41). This program consists of 1 h/day of occupational
therapy, 1 h/2 days/week of neuropsychological therapy, and
2 h/day of group cognitive therapy (memory and executive
function/social skills). Neuropsychological therapy aimed to
improve attention, working memory, learning, memory and
problem solving/executive functions, and emotional-behavioral
problems, through evidence-based techniques that included both
restorative and compensatory strategies. Neuropsychological
treatment goals in each case were defined to achieve maximum
cognitive independence in daily living. In addition, patients
underwent 1 h of physiotherapy and half an hour of speech
therapy, in those cases that needed it. This rehabilitation plan
met the following requirements: (1) agreed by the family
and all professionals involved; (2) formulated in a specific
and operational manner (3) focused on meaningful goals
for the patients that allow them to achieve greater personal
autonomy, community integration, and adaptation to their
deficits; and (4) reviewed monthly. In addition, all patients
attended psychotherapy sessions to help them in the process of
accepting their new situation.
Magnetoencephalographic Recordings
Magnetic fields were recorded using a 148-channel whole-head
magnetometer (4D-MAGNES_2500 WH, 4-D Neuroimaging)
confined in a magnetically shielded room at the Universidad
Complutense of Madrid (Spain). Fields were measured during a
2-min resting-state eyes-closed condition and were sampled at a
frequency rate of 618.17 Hz. Ocular, cardiac, muscular, and jump
artifacts were identified first, by a visual inspection of an expert in
MEG, and then removed using ICA (42) in Brainstorm software
(43). Then, clean data were segmented into 4 s trial length, with
a minimum of 20 artifact-free segments for each subject. The
MEG data were filtered in the classical frequency bands: delta
(2–4 Hz), theta (4–8 Hz), alpha (8–12 Hz), beta (12–30 Hz), and
gamma (30–45 Hz) for further analysis.
Frontiers in Neurology | www.frontiersin.org 3February 2022 | Volume 13 | Article 838170

Pusil et al. Functional Recovery Network After Stroke
Source Reconstruction and Connectivity
Analysis
To reconstruct MEG sources, we used the default anatomy
(15,000 vertices) of the MNI/Colin27 brain (44) in Brainstorm.
This template was warped according to the polhemus points
(nasion and both preauricular) acquired during the head
digitalization to obtain a better approximation of the real
shape of the subject’s head. The overlapping sphere model
was calculated as the forward modeling of MEG measures.
Next, a noise covariance matrix was calculated to estimate
noise level in the MEG recordings. Sources were reconstructed
using the weighted Minimum Norm Estimation (wMNE) (45).
Weighted Minimum Norm Estimation is well-suited for the
estimation of large-scale FC networks, since it addresses the
problem of volume conduction, reducing the correlations of
spurious signals (46,47). Magnetoencephalography sources
were grouped into 68 anatomical regions of interest (ROI)
based on Brainstorm atlas Desikan-Killiany (48). For more
details about the brain areas used, referred to Supplementary
Material for Supplementary Table 1. We selected the mean as the
representative time series for each brain area delimited with the
aforementioned atlas.
Functional connectivity was assessed using the corrected
version of the imaginary phase locking value (ciPLV), a phase
synchronization measure that evaluates the distribution of phase
differences extracted from each of two sensor time series (49,50).
Corrected version of the imaginary phase locking value (Equation
1) was proposed by Bruña et al. (50) to remove the contribution
of the zero phase differences of PLV. Thus, this measure is
insensitive to zero-lag effects, and it is corrected to remove the
instantaneous phase contribution, which could be mainly due to
volume conduction.
ciPLVX,Y(t)=
1
TI{e−i(φX(t)−φY(t))}
q1−(1
TR{e−i(φX(t)−φY(t))})2(1)
where ϕxand ϕyrepresent the phases of each of the two-time
series and stands for the imaginary part of the numerator and
the real part in the denominator. See Figure 1 for the analysis
flow chart of the MEG data.
Statistical Analyses
This study aims to find the possible neurophysiological substrates
of the recovery network underlying the cognitive enhancement
found stroke patient’s sample in the post condition (after NR),
by using cognitive tests, functional scales, and FC measures.
In this context, we performed exploratory analyses with the
data obtained from stroke patients and healthy controls. The
analysis of demographic data showed that there were no
statistical differences in age, sex, and level of education between
patients and controls (p>0.05), so we did not include
them as confounding variables for the following explorations.
Non-parametric tests were used for all comparisons because
variables were non-normally distributed and because of the small
sample size. Specifically, the Mann-Whitney U test was used for
between groups analyses (stroke patients vs. healthy controls)
and Wilcoxon paired test for within-group comparison (Pre vs.
Post conditions in the stroke patients’ group). In the case of
neuropsychological variables, significant results were considered
with a p-value <0.05 after applying false discovery rate (FDR)
corrected for multiple comparisons. For FC data, a total of
10,000 permutations were used for each significant FC link,
and results were considered significant with a p-value <0.005
after applying FDR (51). Finally, with the aim to explore the
relationships between FC and cognition, Spearman’s correlation
analysis was employed. For all analyses the Matlab Statistical
Toolbox was used.
RESULTS
Cognitive Changes After
Neuropsychological Rehabilitation
As described before, the patients underwent a comprehensive
neuropsychological evaluation before and after the NR program.
From the total of battery tests, those scores with at least nine
reported patients (46 scores in total) were included for the
statistical analysis. Pre-condition results indicated that stroke
patients performed significantly worse compared with healthy
controls in all cognitive domains (p<0.05). Comparing pre
and post conditions in the stroke patients’ group, results showed
an important cognitive improvement with 33 scores (72% of
the 46 total scores) significantly different between conditions.
Of these, 21 scores (46% of the 46 total scores), could be
considered relatively enhanced since in the post-condition, they
were significantly different to those corresponding to the healthy
controls (see Figure 2). The remaining 12 scores (26% of the 46
total scores) from the post-condition did not show significant
differences with the healthy control group, indicating a total
improvement. To simplify the interpretation of these results, all
scores were clustered into several aggregated groups depending
on different cognitive domains: Functional performance, 4 scores
(of DEX and PCRS); Executive Functions, 10 scores (of WCST,
Tower of Hanoi, FDT and TMT); Attention, 1 score (of Brief
Test of Attention); Language, 2 scores (of BNT and FAS);
Episodic Memory, 4 scores (of WMS-III) and Working Memory,
9 scores (of WMS-III and WAIS-III). The last three columns
of Figure 2 correspond to the cognitive index of WAIS-III,
including all their subtests: Verbal Comprehension Index (VCI,
4 scores); Processing Speed Index (PSI, 3 scores), and Perceptual
Organization Index (POI, 4 scores). In addition, the three
general indices of WAIS-III were also included [Verbal IQ,
Performance IQ (PIQ), and Full-Scale IQ], as well as subtests
Picture Arrangement and Comprehension. In summary, we
found that all cognitive domains of the stroke patients’ group
were fully or partially enhanced in the post-condition (see
Figure 2). It is important to note that the VCI (that includes
the WAIS-III index and all its subtests), was a cognitive domain
without improvement in any of the four measures; and the four
scores maintained significant differences when comparing the
second scores of stroke patients with the healthy controls scores.
However, it is important to know that this index, already in the
pre-condition, shows a normal score (103.5) unlike the results of
Frontiers in Neurology | www.frontiersin.org 4February 2022 | Volume 13 | Article 838170

Pusil et al. Functional Recovery Network After Stroke
FIGURE 1 | (Top) Sample description (stroke patients and healthy control groups). Description of the protocol followed by each group and the variables collected for
this study. (Bottom) MEG Analysis flowchart. Sequential pipeline of the analysis performed on the MEG data.
the other cognitive index of the WAIS-III (WMI 91.9; PSI 80.3;
PRI 82.4) and the others all cognitive tests. In addition, their
result in the post-condition was 105, although it continues to be
statistically different from the healthy control group (117.6).
Functional Connectivity Disruption After
Stroke: Differences Between Stroke
Patients and Healthy Controls
In order to assess the possible disruption of the patients’
network due to the stroke, their FC in the pre-condition was
compared with the FC of the healthy controls. Stroke patients
exhibited significant FC reduction in the beta band (p<0.005,
FDR corrected) that comprised intra and inter-hemispheric
connections (Figure 3). No significant results were found in
other frequency bands.
Functional Connectivity After
Rehabilitation: The Recovery Network
When assessing the possible FC differences between stroke
patients’ conditions, a clear pattern of widespread increased
FC within the beta band was found. Stroke patients showed
significantly (p<0.005) higher FC in the post-condition
compared to the pre-condition in a variety of links comprising
intra and inter-hemispheric, and antero-posterior long-range
connections (Figure 4).
Moreover, when assessing the possible differences in FC
between groups (post-condition and healthy controls) we did not
find any statistically significant differences. This result indicated
that the original FC disruption in the beta band was restored in
stroke patients who went through the NR.
No significant results were found in other frequency bands in
the pre and post comparison after FDR correction (p<0.005).
Nevertheless, there is a clear pattern of enhanced connectivity
in low frequency bands (delta and theta) in the pre stage when
compared with the brain activity of stroke patients recorded
after the rehabilitation when a less restrictive statistical threshold
was used (p<0.05). Detailed description of these results
could be found in the Supplementary Material, Appendix 3.
These results were not included in the main findings of the
present study because we wanted to focus on the most reliable
FC signature, keeping the p<0.005 value as the go/no go
statistical limit.
Frontiers in Neurology | www.frontiersin.org 5February 2022 | Volume 13 | Article 838170

Pusil et al. Functional Recovery Network After Stroke
FIGURE 2 | Neuropsychological tests’ changes after NR. Functional Performance (4 scores: DEX and PCRS); Executive Functions (10 scores: WCST, Tower of Hanoi,
FDT, and TMT); Attention (1 score: Brief Test of Attention); Language (2 scores: BNT and FAS); Episodic Memory (4 scores: WMS-III); Working Memory (9 scores:
WAIS-III and WMS-III). The last three columns correspond to cognitive indices of WAIS-III, including all their subtests: VCI (verbal comprehension, 4 scores); PSI
(speed processing, 3 scores); and POI (perceptual organization, 4 scores). Percentages were calculated intra-domain. Test scores included. DEX, DEX-family,
DEX-difference; PCRS, PCRS-family, PCRS-difference; WCST, WCST-categories, WCST-conceptual, WCST-persevering; Tower of Hanoi, TH-3D-time,
TH-3D-movements, TH-4D-time, TH-4D-movements; Five Digit Test, FDT-switching, FDT-flexibility; Brief Test of Attention, BTA-total score; BNT, BNT-total score; FAS,
FAS-total score; WMS-III-episodic-memory, WMS-III-logical memory 1, WMS-III-logical memory 2, WMS-III-visual reproduction 1, WMS-III-visual reproduction 2;
WMS-working memory, WMS-III-forward digit span, WMS-III-backward digit span, WMS-III-forward visual span, WMS-III-backward visual span; WAIS-working
memory, WAIS-digit span, WAIS-arithmetic, WAIS-letter number sequencing, WAIS-working memory index; WAIS-VCI, WAIS-vocabulary, WAIS-information,
WAIS-similarities, WAIS-verbal comprehension index; WAIS-PSI, WAIS-symbol search, WAIS-digit symbol, WAIS-processing speed index; WAIS-POI, WAIS-block
design, WAIS-matrix reasoning, WAIS-picture completion; WAIS-perceptual organization index.
Correlations Between the Brain and
Cognitive Recovery Patterns
With the aim to explore if FC changes were related with the
enhanced cognitive performance in the stroke patients’ group,
we firstly calculated a ratio considering the strength of each
functional link that differed between both conditions (FC ratio
=Post/Pre). Then, we averaged these FC link ratios in just
one value for each stroke patient. This provided a unique
FC marker for each patient that condensed the information
obtained by the whole network and the two MEG sessions.
Next, for cognitive scores, we calculated the performance
differences (D) for the most representative tests of each
neuropsychological domain between pre and post conditions
in the stroke patients’ group (D =Post–Pre), with the aim
of finding the strongest cognitive improvement, to reduce the
redundancy of the information (since several scores measured
the similar aspects of the same cognitive domain) and to
avoid the statistical pitfall of multiple comparisons. Regarding
the selection of the most representative scores included for
the correlation analyses, the neuropsychological experts’ team
choose: Functional Performance (DEX-F), Executive functions
(WCST-Persevering, Tower of Hanoi-3D-T), Attention (BTA),
Episodic memory (WMS-III-LM1), Working memory (Digit span
test), and Language (BNT). Moreover, the WAIS-III general
indexes and some WAIS-III cognitive indexes were included
(FIQ, PIQ, VIQ, PSI, POI). We finally included for the
Spearman’s correlation analyses the average FC strength ratio
and 12 neuropsychological scores differences, illustrative of the
cognitive improvement, for each stroke patient.
In order to facilitate the understanding of each patient
cognitive improvement an extended material about individual
neuropsychological performance was added in Appendix 2
in Supplementary Material. There, the punctuations
for each patient before and after the rehabilitation are
described in detail for those tests included in the present
correlation analysis.
We found three positive and significant recovery signatures
correlations between the FC strength ratio and three cognitive
Frontiers in Neurology | www.frontiersin.org 6February 2022 | Volume 13 | Article 838170

Pusil et al. Functional Recovery Network After Stroke
FIGURE 3 | (A) Significant FC results (p<0.005, corrected) in the beta band (12–30 Hz) when comparing healthy controls vs. stroke patients in the pre-condition.
Line thickness of significant links is proportional to FC values (a higher value corresponds to thicker lines, and vice versa); (B) Significant FC links in the beta band are
represented as bar graphs. Red color represents higher connectivity values for healthy controls compared to stroke patients and blue color illustrates lower
connectivity values for stroke patients compared to healthy controls. ROIs included: rmOrbG, Right Medial Orbito Frontal Gyrus; lmOrbG, Left Medial Orbito Frontal
Gyrus; rMFG, Right Middle Frontal Gyrus; rOrbG, Lateral Orbito Frontal Gyrus; rIFGtri, Right Inferior Frontal Parstriangularis; lITG, Left Inferior Temporal Gyrus; rITG,
Right Inferior Temporal Gyrus; lPHG, Left Parahippocampal Gyrus; rPHG, Right Parahippocampal Gyrus; rMCC, Right Posterior Cingulate Gyrus; rPoCG, Right
Postcentral Gyrus; rSMG, Right Supramarginal Gyrus; rLING, Right Lingual Cortex; lMOG, Left Lateral Occipital Gyrus.
measures: Full Scale IQ of WAIS-III (R=0.833; p=0.015), BNT
(R =0.756;p=0.035), and LM1 of WMS-III (R=0.854; p=
0.010) (Figure 5).
Prediction of Brain FC Recovery Based on
Cognitive Performance After Stroke
Lastly, with the aim of exploring the predictive capacity of the
neuropsychological test scores and the brain network recovery,
we correlated the cognitive scores of the pre-condition and the
FC strength ratio (by using Spearman correlation analyses).
Thus, we observed two markers for recovery prediction in
two global cognitive domains: (1) PIQ, with a significant positive
correlation between FC strength ratio and the PIQ scores (R=
0.850, p=0.011); (2) Perceptual Organization, with a significant
positive correlation between FC strength ratio and the POI scores
(R=0.874, p=0.007). Furthermore, within POI, we found a
positive association with Picture Completion (R=0.732; p=
0.048), Matrix Reasoning (R=0.795; p=0.023), and Block
Design (R=0.857; p=0.010) (Figure 6).
DISCUSSION
The present study aimed to provide evidence of the
neurophysiological mechanisms underlying cognitive deficits
and changes in brain function associated with the recovery of
cognitive processes in stroke patients who underwent a NR.
Additionally, this study is focused on the exploration of the
nature of the relationships between neurophysiological and
neuropsychological changes.
In this sense, our results indicate a positive effect in acute
stroke patients who received cognitive rehabilitation on both
levels, the cognitive system and brain functioning. Nevertheless,
the lack of a clinical control group (i.e., stroke patients without
rehabilitation) did not allow us to make causality assumptions,
assuring that cognitive improvement is due specifically or
uniquely to cognitive rehabilitation because it could also
represent some degree of spontaneous clinical recovery after
stroke. Then, according to the results of the present study, stroke
patients who undertook rehabilitation significantly improved
their performance in 72% of cognitive and functional scores.
But we also found than in neuropsychological scores related to
specific cognitive domains such as executive functions, attention,
language, episodic, and working memory, stroke patients showed
a relative improvement (46% improved but there were significant
differences between the post-condition and the control group),
or even a total enhancement (26% improved, and there were no
differences between post-condition and control group). These
two degrees of positive changes were also found in scores
related to global cognitive functioning such as Full-Scale IQ,
PIQ, Speed Processing Index, Working Memory Index, or POI.
In addition, some indicators related to functional performance,
such as DEX or PCRS (completed by relatives of patients),
also improved after rehabilitation. This trend could represent
Frontiers in Neurology | www.frontiersin.org 7February 2022 | Volume 13 | Article 838170

Pusil et al. Functional Recovery Network After Stroke
FIGURE 4 | (A) Significant FC results (p<0.005, corrected) in the beta band (12–30 Hz) when comparing within the stroke patients’ group, the pre-condition and the
post- condition. Line thickness of significant links is proportional to FC values (a higher value corresponds to thicker lines, and vice versa); (B) Significant functional
connectivity links in the beta band are represented as bar graphs. Red color represents higher connectivity values for pre-condition compared to post-condition and
blue color illustrates lower connectivity values for pre-condition compared to post-condition. ROIs included: lMFG, Left Middle Frontal Gyrus; rMFG, Right Middle
Frontal Gyrus; lIFG, Left Inferior Frontal Parstriangularis; rIFG, Right Inferior Frontal Parstriangularis; rmOrbG, Right Medial Orbito Frontal Gyrus; lACCr, Left Rostral
Anterior Cingulate; rACCr, Right Rostral Anterior Cingulate; lOrbG, Left Lateral Orbito Frontal Gyrus; rIFGorb, Right Inferior Frontal Orbital; lIFGop, Left Inferior Frontal
Gyrus Opercular; lSTG, Left Superior Temporal Gyrus; rPreCG, Right Precentral Gyrus; rSTG, Right Superior Temporal Gyrus; lMTG, Left Middle Temporal Gyrus;
rPoCG, Right Poscentral Gyrus; lPHG, Left Parahipocampal Gyrus; rSMG, Right Supramarginal Gyrus; lSMG, Left Supramarginal Gyrus; lPCUN, Left Precuneus;
lstroke patientsL, Left Superior Parietal Lobule; rstroke patientsL, Right Superior Parietal Lobule; rLING, Right Lingual Cortex; rCAL, Right Calcarine; lMOG, Left
Lateral Occipital Gyrus; rMOG, Right Lateral Occipital Gyrus.
a partial cognitive and functional improvement and may have
clinically relevant implications, since it may be considered as
an indicator of recovery. To rule out the possibility that the
learning effect could be influencing the improvement of some
cognitive scores, we have other complementary data related to
changes in brain function. Specifically, the stroke patients of this
study exhibited a widespread increased FC pattern within the
beta band, indicating that their original disruption was restored
in the recording performed after NR in that frequency band.
We also obtained two very important results associated with the
relationships between cognitive scores and changes in FC. On
the one hand, we found three positive and significant recovery
signatures correlations between the FC strength ratio and three
cognitive measures changes (in Full-Scale IQ, BNT, and LM1),
and on the other hand, we observed the predictive capacity of
some neuropsychological test scores (in the pre-condition) and
the recovery of the brain network (in the FC strength ratio). In
this sense, we found two predictive markers of brain recovery
related to two global cognitive domains, PIQ and Perceptual
Organization (both from the WAIS-III scale).
Based on these data, we can affirm that these stroke patients
experienced at least some recovery in their global cognitive
capacity, despite the different etiology and location of their
lesions. Nonetheless the previous literature about the effect of
NR on specific cognitive domains remains unclear. Low to
moderate effects of rehabilitation in executive functions (6),
attention (4,52), or memory (53) have been reported. These
results could have low consistency for different reasons: (1)
the low methodological quality or insufficient description (2,4,
52) the use of small samples; (3) the absence of comparisons
between intervention and no intervention or placebo conditions
(4,6) the deficit of randomized control trials (4,5,52) the
need for standardized definition and outcome measures (53,54);
and (6) the lack of inclusion of functional ability measures in
the rehabilitation outcome evaluation (52). It seems important
to find the most effective procedures to try recovering the
cognitive deficits associated with stroke, considering that the
prevalence of post-stroke cognitive impairment is 53.4% (55),
since it causes an increase in the institutionalization rate and
costs of care (56,57) and a decrease in the quality of life (58).
Frontiers in Neurology | www.frontiersin.org 8February 2022 | Volume 13 | Article 838170

Pusil et al. Functional Recovery Network After Stroke
FIGURE 5 | Recovery signatures correlations. Statistically significant
correlations between cognitive differences and the mean beta strength ratio in
the stroke patients’ group. (A) Full Scale IQ (of WAIS-III); (B) BNT (Boston
Naming Test); (C) WMS-LM1 (Logical Memory 1 of WMS-III).
In addition, if stroke is a central factor in the development of
cognitive impairment, or if this depends on the severity, subtype,
location, or its recurrence, it becomes essential to understand the
brain mechanisms that produce both deficits and their recovery.
There is agreement around the idea that cognitive rehabilitation
interventions aim to improve the impaired brain functions in
stroke patients, and that it must be related to the damaged
anatomical substrate (10). Usually, rehabilitation facilitates the
development of behavioral and cognitive strategies that have a
positive impact on the structural and functional recovery of the
brain (53,59). In this sense, it seems worthy to have in mind other
promising complementary interventions including for example
non-invasive transcranial magnetic stimulation to enhance some
cognitive recoveries in stroke patients (60).
Restoration interventions aim to regain the cognitive abilities
of stroke patients, including domain-specific interventions and
treatments for generalized cognitive impairment (10). The
patients of our study received an integrated treatment based on
the holistic-comprehensive model proposed by Ben-Yishay and
Diller (41), which is consistent with the interventions suggested
by some experts in post-stroke cognitive rehabilitation in terms
of their global treatment approach. This type of treatment could
be very successful for this type of patients, insofar as it produces
more clearly a pattern of overall improvement, both at behavioral
and brain level. Other types of cognitive interventions, such as
computer-assisted cognitive rehabilitation that has increased in
recent years, although show some efficacy in improving attention,
memory, executive function, or visuo-spatial neglect in stroke
patients (61,62), present very limited effects on working memory
and even no effects on cognitive function compared to healthy
controls (63).
As discussed above, overall review studies on the effectiveness
of cognitive intervention with stroke patients do not provide
clear conclusions. However, we must know that one of the most
important issues regarding the functioning of the human mind
has to do with the factor of interdependence between the different
cognitive domains. This aspect is often overlooked in cognitive
performance studies, and review studies of the effectiveness of
cognitive treatments do not usually consider it. For example,
there are studies that focus on improving attention after having
specifically trained it, and thus with the rest of cognitive domains,
without evaluating the impact of attention deficit or executive
deficit in other domains such as memory or language. However,
cognitive interdependence makes it very exceptional for patients
who have brain injuries to suffer a specific cognitive deficit
in a specific cognitive domain. The cognitive deficit of brain
injury patients usually affects several domains, for example,
visuo-spatial attention, working memory, executive functions,
and episodic memory. Thus, trying to understand functioning
of human cognition from independent cognitive domains, is
probably an incorrect approach that hinders the interpretation of
the results in neuropsychology. Usual intervention in the clinical
setting is not as domain specific as studies suggest, since isolating
cognitive processes in habitual actions is not easy. However, the
neuropsychological literature continues to try to understand the
effect of rehabilitation on each cognitive domain individually.
This discrepancy requires a revision and a paradigm shift.
Furthermore, our intention was to go one step further
trying to understand whether this recovery process seen at a
cognitive and behavioral level could have some reflection in
brain functioning. Cognitive functions depend on the integrated
functioning of large-scale distributed brain networks (64).
Specifically, recent evidence suggests that FC between brain
regions may play an important role when difficulties arise from
deficits in attention, memory, or other cognitive functions (65).
In this context, we firstly looked for a FC pattern related to stroke.
We found a disruption in the pattern of brain functioning, with
a significant decrease in the beta band FC for intra and inter-
hemispheric connections in stroke patients before rehabilitation
Frontiers in Neurology | www.frontiersin.org 9February 2022 | Volume 13 | Article 838170

Pusil et al. Functional Recovery Network After Stroke
FIGURE 6 | Recovery prediction. Statistically significant differences between the cognitive scores in the pre-condition and the mean beta strength ratio (post/pre) in
the stroke patients’ group. (A) Performance IQ (WAIS-III); (B) POI (Perceptual Organization Index of WAIS-III); (C) Block Design, Matrix Reasoning, and Picture
Completion (of WAIS-III). Note that the tests represented in this figure are related to each other, since Block Design, Matrix Reasoning, and Picture Completion are the
tests included in Perceptual Organization Index, and POI (Perceptual Organization Index) is included in Performance IQ (on the WAIS-III scale).
compared to healthy controls. General disruption of dynamic
networks after stroke have been previously reported in MEG
studies (25). Alterations in beta band activity have been especially
related to stroke compared to healthy controls (26), supporting
our results.
On the other hand, understanding the interaction between
brain regions within a network (i.e., their FC), and the
interactions among networks are both important for efficient
cognitive function (66). Therefore, exploring the possible
neurophysiological mechanisms underlying the recovery process
after stroke seems to be a crucial point in understanding the
effect of cognitive rehabilitation on the brain. By observing
stroke patients before and after the rehabilitation, a specific
brain recovery pattern emerged, characterized by a widespread
increased FC in the beta in the post-condition. The functioning
of the beta frequency band has previously been related, in
healthy population, to different cognitive tasks such as working
memory (67–69), attention (70), and motor performance (71).
On the contrary, our neurophysiological data were acquired
during resting state (RS) which has been shown that is the most
stable condition across patients with different symptomatology
and it also has been considered a hallmark for clinical diagnosis
and monitoring the recovery of patients that underwent a
rehabilitation, both in MRI (18,72) and MEG studies (22,73).
Previous results have described the role of RS FC as a predictor
of motor learning ability in beta-band for healthy participants
(74) or as a predictor of post-stroke motor recovery in alpha-
band (75). Furthermore, the reorganization in FC in the beta-
band during resting-state has previously been associated with the
success of cognitive and physical interventions (13,76,77).
Up to this point, two independent markers of functional
recovery (i.e., cognitive, and neurophysiological) were found
in our stroke patients who went through the NR, but we
developed further analyses to discover the possible relation
between them. As mentioned before, trying to understand
complex systems such as human cognition or brain functioning
focusing on only some of their components gives partial
information of the entire process. Therefore, with the aim
to simplify the dimensionality of the data and to explore
relationships between cognitive and neurophysiological findings,
the difference (D =Post–Pre) of the most representative
cognitive scores and the ratio of change (post/pre) of the total
FC beta network strength were used. Three positive correlations
between recovery signatures (i.e., cognitive and neurophysiology)
were observed for stroke patients, corresponding to Full
Scale IQ of WAIS-III, Boston Naming Test, and Logical
Memory 1 of WMS-III. These results showed that the beta
connectivity changes after the NR, compared with the data
Frontiers in Neurology | www.frontiersin.org 10 February 2022 | Volume 13 | Article 838170

Pusil et al. Functional Recovery Network After Stroke
obtained in the first recording are, in fact, the reflection of
the cognitive improvement in the brain. The measures related
with improvement represent global neuropsychological indices,
which contain different cognitive domains such as sustained
and switching attention, visuo-spatial attention, visuo-spatial
working memory, planning, flexibility, or processing speed,
but also episodic memory or verbal denomination. Initially,
the neuropsychological deficit observed in stroke patients was
global and their subsequent cognitive recovery, although not
complete, was also general. The scope of cognitive changes
after NR was really wide, probably due to the type of
intervention, which was holistic and not only focused on
specific cognitive functions, including global and interdependent
domains, and focusing on individual cognitive, functional,
emotional, and behavioral imbalances. All these evidences are
consistent with the brain network global changes observed in
stroke patients.
While addressing brain and cognitive changes in stroke
patients is important to understand the underlying mechanisms
of stroke and brain plasticity, the early detection of patterns
or biomarkers is also relevant to predict which subjects are
more likely to improve and benefit from neurorehabilitation.
In this regard, adjustments can be made for those patients
who will not benefit from this option. Thus, we performed
an additional correlation analysis in which the pre-condition
cognitive measures and the ratio of change (post/pre) of the
total FC beta network strength were taken into account. Two
global cognitive markers were stated as predictors of brain
functioning recovery, PIQ and POI. Furthermore, within POI,
we found a significant association for every subtest included in
the global index (i.e., Picture Completion, Matrix Reasoning,
and Block Design). In the clinical setting, the role of prediction
in terms of the degree of future recovery has always been
important, however it is a complex and complicated issue.
Until now we only had clinical, cognitive, behavioral, and
social variables, but these results indicate that the relationship
between behavior and the brain can contribute to this topic.
In this case, the results indicate that the initial state around
some cognitive domains such as visuo-spatial attention, visuo-
spatial working memory, or planning capacity, could have
a very relevant role in the evolution and recovery of the
brain network of stroke patients. This information is not only
important to predict the patients who will improve the most,
but it can also serve to think about more powerful intervention
procedures for those patients who have a more serious deficit
around these cognitive domains. This result has very relevant
clinical implications.
In conclusion, if NR aims to improve people’s cognitive
function in order to restore their general performance and
independence in functional activities, the results of the
present study are in line with this objective, showing a clear
improvement pattern in stroke patients who received NR both
cognitively and brain function. We are also sure that this
study points out the importance of including neuropsychological
and neurophysiological variables in the assessment of the
outcome and effectiveness of psychological interventions of
stroke patients.
LIMITATIONS AND FUTURE RESEARCH
An important contribution of this study may be that, unlike
most studies, functional brain connectivity was measured with
MEG. Despite the intrinsic limitations of BOLD fMRI, MEG
is a measure of brain activity with incredibly high temporal
resolution (ms). Despite its advantages, MEG is an underused
neuroimaging tool in clinical and research contexts. Although
the sample of this exploratory study size was small, we were able
to identify a pattern of recovery of FC in the beta band related
to cognitive enhancement in stroke patients who underwent a
NR. Additionally, we were able to make predictions based on
the cognitive performance of stroke patients before rehabilitation
about the future functional restoration of the brain network.
Thus, larger MEG studies with stroke patients are needed to
demonstrate the power of this neurophysiological tool within
this neurological field. Another limitation that this study faces
is the heterogeneity of stroke patients, in terms of etiology and
location of the injury. However, we believe that this heterogeneity
has provided an interesting approach to the study since it
has allowed us to explore the effect of cognitive intervention
both at cognitive and neurophysiological level in stroke patients
with different etiology. Furthermore, another limitation of the
study is the absence of a clinical control group (i.e., stroke
patients without rehabilitation). Considering this limitation, we
cannot assure that cognitive improvement is due specifically
or uniquely to cognitive rehabilitation because it could also
represent some degree of spontaneous clinical recovery after
stroke. In accordance with the aforementioned obstacles, future
studies should include a group of stroke patients without
cognitive intervention (e.g., on the waiting list), and larger
samples of patients that allow comparisons based on the etiology
and the location of the lesion. Finally, it would be interesting to
focus on different networks of the brain (78) such as the default
mode, the salience, or the executive control networks (79), to
explore specific changes in each network recovery pattern, and
its possible relationship to particular improvements in cognition
after stroke.
DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be
made available by the authors, without undue reservation.
ETHICS STATEMENT
The studies involving human participants were reviewed and
approved by the National Brain Injury Rehabilitation Center
Ethics Committee (Madrid). The patients/participants provided
their written informed consent to participate in this study.
AUTHOR CONTRIBUTIONS
SP, LT-S, ML, NP, and FM designed research. SP and LT-S
performed main calculations of the study and prepared figures.
SP, LT-S, ML, BC, LC, AB, FM, and NP collaborated actively in
writing the manuscript. All authors contributed to the article and
approved the submitted version.
Frontiers in Neurology | www.frontiersin.org 11 February 2022 | Volume 13 | Article 838170

Pusil et al. Functional Recovery Network After Stroke
FUNDING
Financial support of the project was provided by IMSERSO (07-
2008) and the Spanish MICINN (PSI2011-28388). Research by SP
was supported by the Spanish MINECO post-doctoral fellowship
(FJC2019-041205-I). Additionally, this work was supported by a
predoctoral researcher grant from Universidad Complutense de
Madrid (CT42/18-CT43/18) and co-founded by Santander Bank
to LT-S, and by the National Council of Science, Technology
and Technological Innovation (CONCYTEC, Perú) through the
National Fund for Scientific and Technological Development
(FONDECYT, Perú) to BC.
SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found
online at: https://www.frontiersin.org/articles/10.3389/fneur.
2022.838170/full#supplementary-material
REFERENCES
1. Feigin VL, Norrving B, Mensah GA. Global burden of stroke. Circ Res. (2017)
120:439–48. doi: 10.1161/CIRCRESAHA.116.308413
2. Aam S, Einstad MS, Munthe-Kaas R, Lydersen S, Ihle-Hansen H, Knapskog
AB, et al. Post-stroke cognitive impairment—impact of follow-up time and
stroke subtype on severity and cognitive profile: the Nor-COAST study. Front
Neurol. (2020) 11:699. doi: 10.3389/fneur.2020.00699
3. Al-Qazzaz NK, Ali SH, Ahmad SA, Islam S, Mohamad K. Cognitive
impairment and memory dysfunction after a stroke diagnosis: a post-
stroke memory assessment. Neuropsychiatr Dis Treat. (2014) 10:1677–
91. doi: 10.2147/NDT.S67184
4. Loetscher T, Potter KJ, Wong D, das Nair R. Cognitive rehabilitation for
attention deficits following stroke. Cochrane Database Syst Rev. (2019)
2019:CD002842. doi: 10.1002/14651858.CD002842.pub3
5. Feigin VL, Roth GA, Naghavi M, Parmar P, Krishnamurthi R, Chugh S, et al.
Global burden of stroke and risk factors in 188 countries, during 1990–2013:
a systematic analysis for the Global Burden of Disease Study 2013. Lancet
Neurol. (2016) 15:913–24. doi: 10.1016/S1474-4422(16)30073-4
6. Chung CSY, Pollock A, Campbell T, Durward BR, Hagen S. Cognitive
rehabilitation for executive dysfunction in adults with stroke or other adult
non-progressive acquired brain damage. Cochrane Database Syst Rev. (2013)
2013:CD008391. doi: 10.1002/14651858.CD008391.PUB2
7. Rasquin SMC, Lodder J, Ponds RWHM, Winkens I, Jolles J, Verhey FRJ.
Cognitive functioning after stroke: a one-year follow-up study. Dement
Geriatr Cogn Disord. (2004) 18:138–44. doi: 10.1159/000079193
8. Katz N, Hartman-Maeir A, Ring H, Soroker N. Relationships of cognitive
performance and daily function of clients following right hemisphere stroke:
predictive and ecological validity of the LOTCA battery. Occup Ther J Res.
(2016) 20:3–17. doi: 10.1177/153944920002000101
9. Ferreira MGR, Moro CHC, Franco SC. Desempenho cognitivo após
acidente vascular cerebral isquêmico. Dement Neuropsychol. (2015) 9:165–
75. doi: 10.1590/1980-57642015DN92000011
10. Zhao Q, Wang X, Wang T, Dmytriw AA, Zhang X, Yang K, et al.
Cognitive rehabilitation interventions after stroke: protocol for a systematic
review and meta-analysis of randomized controlled trials. Syst Rev. (2021)
10:66. doi: 10.1186/s13643-021-01607-7
11. Canuet L. Neurorehabilitation in stroke: the role of functional connectivity.
Int J Neurorehabil. (2015) 2:172. doi: 10.4172/2376-0281.1000172
12. Cicerone KD, Mott T, Azulay J, Sharlow-Galella MA, Ellmo WJ, Paradise S,
et al. Randomized controlled trial of holistic neuropsychologic rehabilitation
after traumatic brain injury. Arch Phys Med Rehabil. (2008) 89:2239–
49. doi: 10.1016/j.apmr.2008.06.017
13. Castellanos NP, Paúl N, Ordóñ Ez VE, Demuynck O, Bajo R, Campo
P, et al. Reorganization of functional connectivity as a correlate of
cognitive recovery in acquired brain injury. Brain. (2010) 133:2365–
81. doi: 10.1093/brain/awq174
14. Alexander LD, Black SE, Gao F, Szilagyi G, Danells CJ, McIlroy WE.
Correlating lesion size and location to deficits after ischemic stroke: the
influence of accounting for altered peri-necrotic tissue and incidental silent
infarcts. Behav Brain Funct. (2010) 6:6. doi: 10.1186/1744-9081-6-6
15. Ovadia-Caro S, Villringer K, Fiebach J, Jungehulsing GJ, van der Meer
E, Margulies DS, et al. Longitudinal effects of lesions on functional
networks after stroke. J Cereb Blood Flow Metab. (2013) 33:1279–
85. doi: 10.1038/jcbfm.2013.80
16. Bayrak S, Khalil AA, Villringer K, Fiebach JB, Villringer A, Margulies DS, et
al. The impact of ischemic stroke on connectivity gradients. NeuroImage Clin.
(2019) 24:101947. doi: 10.1016/j.nicl.2019.101947
17. Carrera E, Tononi G. Diaschisis: Past, present, future. Brain. (2014) 137:2408–
22. doi: 10.1093/brain/awu101
18. Baldassarre A, Ramsey LE, Siegel JS, Shulman GL, Corbetta M. Brain
connectivity and neurological disorders after stroke. Curr Opin Neurol. (2016)
29:706–13. doi: 10.1097/WCO.0000000000000396
19. Ramsey LE, Siegel JS, Baldassarre A, Metcalf NV, Zinn K, Shulman GL, et al.
Normalization of network connectivity in hemispatial neglect recovery. Ann
Neurol. (2016) 80:127–41. doi: 10.1002/ana.24690
20. Maestú F, Peña JM, Garcés P, González S, Bajo R, Bagic A, et al. A multicenter
study of the early detection of synaptic dysfunction in mild cognitive
impairment using magnetoencephalography-derived functional connectivity.
Neuroimage Clin. (2015) 9:103–9. doi: 10.1016/j.nicl.2015.07.011
21. Stam CJ. Use of magnetoencephalography (MEG) to study functional brain
networks in neurodegenerative disorders. J Neurol Sci. (2010) 289:128–
34. doi: 10.1016/j.jns.2009.08.028
22. Paggiaro A, Birbaumer N, Cavinato M, Turco C, Formaggio E, del Felice A,
et al. Magnetoencephalography in stroke recovery and rehabilitation. Front
Neurol. (2016) 7:35. doi: 10.3389/fneur.2016.00035
23. Tewarie P, Steenwijk MD, Tijms BM, Daams M, Balk LJ, Stam CJ, et al.
Disruption of structural and functional networks in long-standing multiple
sclerosis. Hum Brain Mapp. (2014) 35:5946–61. doi: 10.1002/HBM.22596
24. Pusil S, López ME, Cuesta P, Bruña R, Pereda E, Maestú F.
Hypersynchronization in mild cognitive impairment: the ‘X’ model. Brain.
(2019) 142:3936–50. doi: 10.1093/brain/awz320
25. Marsh EB, Brodbeck C, Llinas RH, Mallick D, Kulasingham JP, Simon JZ, et
al. Poststroke acute dysexecutive syndrome, a disorder resulting from minor
stroke due to disruption of network dynamics. Proc Natl Acad Sci USA. (2020)
117:33578–585. doi: 10.1073/pnas.2013231117/
26. Kulasingham JP, Brodbeck C, Khan S, Marsh EB, Simon JZ. Bilaterally
reduced rolandic beta band activity in minor stroke patients. bioRxiv Preprint.
(2021). doi: 10.1101/2021.10.15.464457
27. Castellanos NP, Leyva I, Buldú JM, Bajo R, Paúl N, Cuesta P, et al. Principles
of recovery from traumatic brain injury: reorganization of functional
networks. Neuroimage. (2011) 55:1189–99. doi: 10.1016/j.neuroimage.201
0.12.046
28. Teasdale G, Jennett B. Assessment of coma and impaired consciousness a
practical scale. Lancet. (1974) 2:81–4. doi: 10.1016/s0140-6736(74)91639-0
29. Wechsler D. Wechsler Adult Intelligence Scale (WAIS-III). San Antonio, TX:
Psychological Corporation (1997).
30. Schretlen D. Brief Test of Attention. Baltimore, MD: Psychological Assessment
Resources (1989).
31. Reitan RM, Wolfson D. The Halstead-Reitan Neuropsychological Test
Battery e Theory and Clinical Interpretation. Tucson, AZ: Neuropsychology
Press (1993)
32. Golden CJ. A Manual for the Clinical and Experimental Use of the Stroop Color
and Word Test. Chicago, IL: Stoelting (1978).
33. Jensen AR, Rohwer WD. The Stroop color-word test: a review. Acta Psychol.
(1966) 25:36–93. doi: 10.1016/0001-6918(66)90004-7
Frontiers in Neurology | www.frontiersin.org 12 February 2022 | Volume 13 | Article 838170

Pusil et al. Functional Recovery Network After Stroke
34. Heaton RK, Grant I, Matthews CG. Comprehensive Norms for an Expanded
Halstead-Reitan Battery: Demographic Corrections, Research Findings, and
Clinical Applications. Odessa, FL: Psychological Assessment Resources. (1991).
35. Édouard L. Récréations Mathématiques. Vol III. Paris: Reprinted
several times by Albert Blanchard (1983). Available online at: https://
books.google.es/books?hl=en&lr=&id=hwYAAAAAQAAJ&oi=fnd&
pg=PA3&dq=%C3%89douard$+$L.$+$R%C3%A9cr%C3%A9ations$+
$Math%C3%A9matiques.$+$vol.$+$III&ots=6q1xOXhLC4&sig=
bZgxKMrYvQEczRsMbW7m9ovm5WM&redir_esc=y#v=onepage&q=
%C3%89douard%20L.%20R%C3%A9cr%C3%A9ations%20Math%C3
%A9matiques.%20vol.%20III&f=false (accessed December 15, 2021).
36. Wilson BA, Alderman N, Burgess PW, Emslie H, Evans JJ. BADS: Behavioural
Assessment of the Dysexecutive Syndrome : Test Manual. London: Thames
Valley Test Company (1996).
37. Kaplan E. Boston Naming Test. Philadelphia, PA: Lea & Febiger (1983).
38. Benton AL, Hamsher K. Controlled Oral Word Association Test – COWAT.
Iowa City, IA: University of Iowa Press (1976)
39. Sedó M. FDT. Test de los Cinco Dígitos. TEA Ediciones (2007). Available
online at: https://web.teaediciones.com/fdt--test-de-los-cinco-digitos.aspx
(accessed December 15, 2021).
40. Prigatano GP, Schacter DL. Awareness of Deficit After Brain Injury. New York,
NY: Oxford University Press (1991).
41. Ben-Yishay Y, Diller L. Handbook of Holistic Neuropsychological
Rehabilitation: Outpatient Rehabilitation of Traumatic Brain Injury.
Oxford University Press (2011). Available online at: https://books.
google.es/books?hl=es&lr=&id=Yzg_zb94ZHcC&oi=fnd&pg=PP1&dq=
Handbook$+$of$+$Holistic$+$Neuropsychological$+$Rehabilitation:$+
$Outpatient$+$Rehabilitation$+$of$+$Traumatic$+$Brain$+$Injury&ots=
KzE1fPTgBK&sig=0neZLIlQmPk85DCW3BVyrj5qVPQ#v=onepage&q=
Handbook%20of%20Holistic%20Neuropsychological%20Rehabilitation%3A
%20Outpatient%20Rehabilitation%20of%20Traumatic%20Brain%20Injury&
f=false (accessed December 15, 2021).
42. Bell AJ, Sejnowski TJ. An information-maximization approach to blind
separation and blind deconvolution. Neural Comput. (1995) 7:1129–
59. doi: 10.1162/NECO.1995.7.6.1129
43. Tadel F, Baillet S, Mosher JC, Pantazis D, Leahy RM. Brainstorm: a user-
friendly application for MEG/EEG analysis. Comput Intell Neurosci. (2011)
2011:879716. doi: 10.1155/2011/879716
44. Holmes CJ, Hoge R, Collins L, Woods R, Toga AW, Evans AC.
Enhancement of MR images using registration for signal averaging. J
Comput Assist Tomogr. (1998) 22:324–33. doi: 10.1097/00004728-19980300
0-00032
45. Mosher JC, Baillet S, Leahy RM. Equivalence of linear approaches
in bioelectromagnetic inverse solutions. In: IEEE Workshop on
Statistical Signal Processing Proceedings. St. Louis, MO (2003) p.
294–7. doi: 10.1109/SSP.2003.1289402
46. Palva S, Palva JM. Discovering oscillatory interaction networks with
M/EEG: challenges and breakthroughs. Trends Cogn Sci. (2012) 16:219–
30. doi: 10.1016/j.tics.2012.02.004
47. Hassan M, Dufor O, Merlet I, Berrou C, Wendling F. EEG source connectivity
analysis: from dense array recordings to brain networks. PLoS ONE. (2014)
9:e105041. doi: 10.1371/JOURNAL.PONE.0105041
48. Desikan RS, Ségonne F, Fischl B, Quinn BT, Dickerson BC, Blacker D, et al.
An automated labeling system for subdividing the human cerebral cortex on
MRI scans into gyral based regions of interest. Neuroimage. (2006) 31:968–
80. doi: 10.1016/j.neuroimage.2006.01.021
49. Mormann F, Lehnertz K, David P, Elger CE. Mean phase coherence
as a measure for phase synchronization and its application to the
EEG of epilepsy patients. Phys D Nonlin Phenom. (2000) 144:358–
69. doi: 10.1016/S0167-2789(00)00087-7
50. Bruña R, Maestú F,Pereda E. Phase locking value revisited: teaching new tricks
to an old dog. J Neural Eng. (2018) 15:056011. doi: 10.1088/1741-2552/aacfe4
51. Benjamini Y, Yekutieli D. The control of the false discovery rate
in multiple testing under dependency. Ann Statist. (2001) 29:1165–
88. doi: 10.1214/aos/1013699998
52. Hazelton C. Can cognitive rehabilitation improve attention deficits
following stroke? - A Cochrane Review summary with commentary
NeuroRehabilitation. (2020) 47:355–7. doi: 10.3233/NRE-209007
53. das Nair R, Cogger H, Worthington E, Lincoln NB. Cognitive rehabilitation
for memory deficits after stroke. Cochrane Database Syst Rev. (2016)
9:CD002293. doi: 10.1002/14651858.CD002293.pub3
54. Bernhardt J, Hayward KS, Kwakkel G, Ward NS, Wolf SL, Borschmann K, et
al. Agreed definitions and a shared vision for new standards in stroke recovery
research: the Stroke Recovery and Rehabilitation Roundtable taskforce. Int J
Stroke. (2017) 12:444–50. doi: 10.1177/1747493017711816
55. Barbay M, Diouf M, Roussel M, Godefroy O. Systematic review and
meta-analysis of prevalence in post-stroke neurocognitive disorders in
hospital-based studies. Dement Geriatr Cogn Disord. (2019) 46:322–
34. doi: 10.1159/000492920
56. Pasquini M, Leys D, Rousseaux M, Pasquier F, Hénon H. Influence of cognitive
impairment on the institutionalisation rate 3 years after a stroke. J Neurol
Neurosurg Psychiatry. (2007) 78:56–9. doi: 10.1136/jnnp.2006.102533
57. Claesson L, Lindén T, Skoog I, Blomstrand C. Cognitive impairment after
stroke - Impact on activities of daily living and costs of care for elderly
people: the Göteborg 70+stroke study. Cerebrovasc Dis. (2005) 19:102–
9. doi: 10.1159/000082787
58. Park JH, Kim BJ, Bae H-J, Lee J, Lee J, Han M-K, et al. Impact of post-stroke
cognitive impairment with no dementia on health-related quality of life. J
Stroke. (2013) 15:49–56. doi: 10.5853/JOS.2013.15.1.49
59. Robertson S. The efficacy of oro-facial and articulation exercises in
dysarthria following stroke Int J Lang Commun Disord. (2001) 36(Suppl):292–
7. doi: 10.3109/13682820109177900
60. Draaisma LR, Wessel MJ, Hummel FC. Non-invasive brain stimulation
to enhance cognitive rehabilitation after stroke. Neurosci Lett. (2020)
719:133678. doi: 10.1016/j.neulet.2018.06.047
61. Yoo C, Yong MH, Chung J, Yang Y. Effect of computerized cognitive
rehabilitation program on cognitive function and activities of living in stroke
patients. J Phys Ther Sci. (2015) 27:2487–9. doi: 10.1589/jpts.27.2487
62. Svaerke K, Niemeijer M, Mogensen J, Christensen H. The effects of
computer-based cognitive rehabilitation in patients with visuospatial neglect
following stroke: a systematic review. Top Stroke Rehabil. (2019) 26:214–
25. doi: 10.1080/10749357.2018.1556963
63. Wentink MM, Berger MAM, de Kloet AJ, Meesters J, Band GPH, Wolterbeek
R, et al. The effects of an 8-week computer-based brain training programme
on cognitive functioning, QoL and self-efficacy after stroke. Neuropsychol
Rehabil. (2016) 26:847–65. doi: 10.1080/09602011.2016.1162175
64. Mesulam MM. From sensation to cognition. Brain. (1998) 121(Pt 6):1013–
52. doi: 10.1093/brain/121.6.1013
65. Siegel JS, Ramsey LE, Snyder AZ, Metcalf NV, Chacko RV, Weinberger K,
et al. Disruptions of network connectivity predict impairment in multiple
behavioral domains after stroke. Proc Natl Acad Sci USA. (2016) 113:E4367–
76. doi: 10.1073/pnas.1521083113
66. Leech R, Kamourieh S, Beckmann CF, Sharp DJ. Fractionating the
default mode network: distinct contributions of the ventral and dorsal
posterior cingulate cortex to cognitive control. J Neurosci. (2011) 31:3217–
24. doi: 10.1523/JNEUROSCI.5626-10.2011
67. Koshy SM, Wiesman AI, Proskovec AL, Embury CM, Schantell MD, Eastman
JA, et al. Numerical working memory alters alpha-beta oscillations and
connectivity in the parietal cortices. Hum Brain Mapp. (2020) 41:3709–
19. doi: 10.1002/hbm.25043
68. Proskovec AL, Heinrichs-Graham E, Wilson TW. Load modulates the alpha
and beta oscillatory dynamics serving verbal working memory. Neuroimage.
(2019) 184:256–65. doi: 10.1016/j.neuroimage.2018.09.022
69. Schmidt R, Ruiz MH, Kilavik BE, Lundqvist M, Starr PA, Aron
AR. Beta oscillations in working memory, executive control of
movement and thought, and sensorimotor function. J Neurosci.
39:8231–8. doi: 10.1523/JNEUROSCI.1163-19.2019
70. Kami´
nski J, Brzezicka A, Gola M, Wróbel A. Beta band oscillations
engagement in human alertness process. Int J Psychophysiol. (2012) 85:125–
8. doi: 10.1016/j.ijpsycho.2011.11.006
71. Chung JW, Ofori E, Misra G, Hess CW, Vaillancourt DE. Beta-
band activity and connectivity in sensorimotor and parietal cortex are
important for accurate motor performance. Neuroimage. (2017) 144:164–
73. doi: 10.1016/j.neuroimage.2016.10.008
72. Dacosta-Aguayo R, Graña M, Savio A, Fernández-Andújar M, Millán M,
López-Cancio E, et al. Prognostic value of changes in resting-state functional
Frontiers in Neurology | www.frontiersin.org 13 February 2022 | Volume 13 | Article 838170

Pusil et al. Functional Recovery Network After Stroke
connectivity patterns in cognitive recovery after stroke: a 3T fMRI pilot study.
Hum Brain Mapp. (2014) 35:3819–31. doi: 10.1002/hbm.22439
73. Martín-Buro MC, Garcés P, Maestú F. Test-retest reliability of resting-state
magnetoencephalography power in sensor and source space. Hum Brain
Mapp. (2016) 37:179–90. doi: 10.1002/hbm.23027
74. Sugata H, Yagi K, Yazawa S, Nagase Y, Tsuruta K, Ikeda T,
et al. Role of beta-band resting-state functional connectivity
as a predictor of motor learning ability. NeuroImage. (2020)
210:116562. doi: 10.1016/j.neuroimage.2020.116562
75. Westlake KP, Hinkley LB, Bucci M, Guggisberg AG, Findlay AM, Henry RG, et
al. Resting state alpha-band functional connectivity and recovery after stroke.
Exp Neurol. (2012) 237:160–9. doi: 10.1016/j.expneurol.2012.06.020
76. Weiss SA, Bassett DS, Rubinstein D, Holroyd T, Apud J, Dickinson D, et al.
Functional brain network characterization and adaptivity during task practice
in healthy volunteers and people with schizophrenia. Front Hum Neurosci.
(2011) 5:81. doi: 10.3389/fnhum.2011.00081
77. Klados MA, Styliadis C, Frantzidis CA, Paraskevopoulos E, Bamidis PD. Beta-
band functional connectivity is reorganized in mild cognitive impairment
after combined computerized physical and cognitive training. Front Neurosci.
(2016) 10:55. doi: 10.3389/fnins.2016.00055
78. Power JD, Cohen AL, Nelson SM, Wig GS, Barnes KA, Church JA, et al.
Functional network organization of the human brain. Neuron. (2011) 72:665–
78. doi: 10.1016/j.neuron.2011.09.006
79. Raichle ME, MacLeod AM, Snyder AZ, Powers WJ, Gusnard
DA, Shulman GL, et al. A default mode of brain function.
Proc Nat Acad Sci USA. (2001) 98:676–82. doi: 10.1073/PNAS.9
8.2.676
Conflict of Interest: The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could be construed as a
potential conflict of interest.
Publisher’s Note: All claims expressed in this article are solely those of the authors
and do not necessarily represent those of their affiliated organizations, or those of
the publisher, the editors and the reviewers. Any product that may be evaluated in
this article, or claim that may be made by its manufacturer, is not guaranteed or
endorsed by the publisher.
Copyright © 2022 Pusil, Torres-Simon, Chino, López, Canuet, Bilbao, Maestú and
Paúl. This is an open-access article distributed under the terms of the Creative
Commons Attribution License (CC BY). The use, distribution or reproduction in
other forums is permitted, provided the original author(s) and the copyright owner(s)
are credited and that the original publication in this journal is cited, in accordance
with accepted academic practice. No use, distribution or reproduction is permitted
which does not comply with these terms.
Frontiers in Neurology | www.frontiersin.org 14 February 2022 | Volume 13 | Article 838170