Upper Alpha Based Neurofeedback Training in Chronic Stroke: Brain Plasticity Processes and Cognitive Effects

Abstract

In the present study, we investigated the effects of upper alpha based neurofeedback (NF) training on electrical brain activity and cognitive functions in stroke survivors. Therefore, two single chronic stroke patients with memory deficits (subject A with a bilateral subarachnoid hemorrhage; subject B with an ischemic stroke in the left arteria cerebri media) and a healthy elderly control group (N = 24) received up to ten NF training sessions. To evaluate NF training effects, all participants performed multichannel electroencephalogram (EEG) resting measurements and a neuropsychological test battery assessing different cognitive functions before and after NF training. Stroke patients showed improvements in memory functions after successful NF training compared to the pre-assessment. Subject B had a pathological delta (0.5-4 Hz) and upper alpha (10-12 Hz) power maximum over the unaffected hemisphere before NF training. After NF training, he showed a more bilateral and "normalized" topographical distribution of these EEG frequencies. Healthy participants as well as subject A did not show any abnormalities in EEG topography before the start of NF training. Consequently, no changes in the topographical distribution of EEG activity were observed in these participants when comparing the pre- and post-assessment. Hence, our results show that upper alpha based NF training had on the one hand positive effects on memory functions, and on the other hand led to cortical "normalization" in a stroke patient with pathological brain activation patterns, which underlines the potential usefulness of NF as neurological rehabilitation tool.

Full text

Vol.:(0123456789)
1 3
Appl Psychophysiol Biofeedback (2017) 42:69–83
DOI 10.1007/s10484-017-9353-5
Upper Alpha Based Neurofeedback Training inChronic Stroke:
Brain Plasticity Processes andCognitive Effects
SilviaErikaKober1,2· DanielaSchweiger1· JohannaLouiseReichert1·
ChristaNeuper1,2,3· GuilhermeWood1,2
Published online: 14 February 2017
© The Author(s) 2017. This article is published with open access at Springerlink.com
he showed a more bilateral and “normalized” topographical
distribution of these EEG frequencies. Healthy participants
as well as subject A did not show any abnormalities in EEG
topography before the start of NF training. Consequently,
no changes in the topographical distribution of EEG activ-
ity were observed in these participants when comparing
the pre- and post-assessment. Hence, our results show that
upper alpha based NF training had on the one hand positive
effects on memory functions, and on the other hand led to
cortical “normalization” in a stroke patient with pathologi-
cal brain activation patterns, which underlines the potential
usefulness of NF as neurological rehabilitation tool.
Keywords Cortical reorganization· Neurofeedback·
Memory· Stroke recovery
Background
Following stroke, changes in electrical brain activity as
well as cognitive impairment are often evident (Mel-
kas et al. 2014; Jordan 2004; Kaplan and Rossetti 2011;
Finnigan and van Putten 2013; Niedermeyer 2005). In this
context, EEG based neurofeedback (NF) might be a use-
ful rehabilitation tool. There is evidence that NF training
can lead to changes in electrical brain activity which goes
along with cognitive improvements (Kober et al. 2015a,
b; Reichert et al. 2016; Kropotov 2009; Gruzelier 2014).
Using NF, participants can learn to voluntarily modulate
their electrical brain activity. Specific parameters of the
EEG, such as power values in specific frequency bands,
can be extracted and analyzed in real-time and fed back to
the participants via auditory and/or visual feedback. Hence,
with the method of NF, the electrical activity of the brain
is modulated directly and, therefore, the cortical substrates
Abstract In the present study, we investigated the effects
of upper alpha based neurofeedback (NF) training on elec-
trical brain activity and cognitive functions in stroke sur-
vivors. Therefore, two single chronic stroke patients with
memory deficits (subject A with a bilateral subarachnoid
hemorrhage; subject B with an ischemic stroke in the left
arteria cerebri media) and a healthy elderly control group
(N = 24) received up to ten NF training sessions. To evalu-
ate NF training effects, all participants performed multi-
channel electroencephalogram (EEG) resting measure-
ments and a neuropsychological test battery assessing
different cognitive functions before and after NF training.
Stroke patients showed improvements in memory functions
after successful NF training compared to the pre-assess-
ment. Subject B had a pathological delta (0.5–4 Hz) and
upper alpha (10–12 Hz) power maximum over the unaf-
fected hemisphere before NF training. After NF training,
* Silvia Erika Kober
silvia.kober@uni-graz.at
Daniela Schweiger
daniela.hofer@uni-graz.at
Johanna Louise Reichert
johanna.reichert@uni-graz.at
Christa Neuper
neuper@tugraz.at
Guilherme Wood
guilherme.wood@uni-graz.at
1 Department ofPsychology, University ofGraz,
Universitaetsplatz 2/III, 8010Graz, Austria
2 BioTechMed-Graz, Mozartgasse 12/II, Graz8010, Austria
3 Institute ofNeural Engineering, Laboratory
ofBrain-Computer Interfaces, Graz University
ofTechnology, Stremayrgasse 16, Graz8010, Austria
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

70 Appl Psychophysiol Biofeedback (2017) 42:69–83
1 3
of cognitive functions. This direct access to neural activ-
ity by means of NF may alter or accelerate functional reor-
ganization in the brain following stroke. NF might speed up
functional recovery or even enable functional recovery that
otherwise would not have occurred (Nelson 2007). There-
fore, the aim of the present study was to evaluate the effects
of EEG based NF training on brain plasticity processes and
cognitive functions in stroke survivors.
It has been demonstrated that the electroencephalo-
gram (EEG) is a highly sensitive measure to detect cer-
ebral ischemic or hemorrhagic stroke. Stroke patients
show EEG abnormalities compared to healthy people,
which change over the course of the disease. The quan-
titative EEG during the acute and sub-acute state has a
high prognostic value concerning the outcome from
stroke (Finnigan et al. 2007; Finnigan and van Putten
2013; Tecchio et al. 2007; Sheorajpanday et al. 2011).
In this context, slow wave EEG activity in the delta
range (0.5–4Hz) as well as faster oscillatory activity in
the alpha range (8–12 Hz) turned out to play an essen-
tial role (Niedermeyer 2005). Delta power was found to
be negatively correlated with regional cerebral blood
flow (rCBF) while alpha power showed a relatively
strong positive correlation with rCBF (Tolonen and Sulg
1981; Finnigan and van Putten 2013). In stroke patients
with unilateral cerebral infarction delta activity is typi-
cally most pronounced over the affected hemisphere in
the acute state (Finnigan and van Putten 2013; Jordan
2004; Tecchio etal. 2006). Across a few hours during the
acute stroke period, the scalp topography of delta activ-
ity shifts from a maximum over the affected hemisphere
to a maximum over the healthy, unaffected hemisphere.
This interhemispheric shift of scalp delta power maxima
is associated with worsening of cerebral pathophysiology
and clinical state in stroke patients (Finnigan etal. 2008;
Tecchio etal. 2007; Zappasodi etal. 2007; Niedermeyer
2005; Rossini etal. 2003). There is evidence that patho-
logical asymmetry in EEG delta power decreases after
thrombolytic therapy, which in turn leads to improve-
ments in clinical symptoms (Finnigan etal. 2006; de Vos
etal. 2008). Alpha amplitude attenuation is also generally
indicative for cortical injury (Finnigan and van Putten
2013; Finnigan etal. 2007; Klimesch 1999). Alpha power
in the acute state is negatively related to the severity of
stroke symptoms in patients with unilateral lesions in
the arteria cerebri media (ACM) (Finnigan et al. 2007).
In stroke patients with acute subarachnoid hemorrhage
(SAH), EEG delta activity is increased and alpha activ-
ity is reduced, too (Vespa et al. 1997; Claassen et al.
2004; Niedermeyer 2005; Labar etal. 1991). Some EEG
studies also investigated changes in EEG activity in the
post-acute and chronic stage (Mattia etal. 2003). These
studies showed that the greatest improvement in EEG
activity occurred during the first 3 months after stroke
(Giaquinto et al. 1994; de Weerd et al. 1988; Jonkman
et al. 1984). Stroke patients with a unilateral insult in
the ACM showed decreases in delta power and increases
in alpha power levels over the injured hemisphere dur-
ing this time period. Alpha power levels also increased
over the healthy hemisphere. An overall increase in alpha
power was also partially associated with improvements in
motor functions and activities in daily living (Giaquinto
etal. 1994). Furthermore, delta and alpha power became
more symmetrically distributed over both hemispheres
with clinical recovery, which might be an indicator of
“normalization” of electrical brain activity (Giaquinto
etal. 1994; Tecchio etal. 2006).
In the present investigation, we evaluated whether EEG
based NF training can be used as therapeutic tool to evoke
changes in electrical brain activation patterns in chronic
stroke patients, which may be accompanied by cognitive
improvements. In NF training paradigms, participants can
learn to voluntarily increase or decrease the amplitude of
specific EEG frequencies. There is some empirical evidence
that voluntary modulation of EEG amplitudes determines
other aspects of electrical brain activity in healthy people,
which are responsible for improved cognitive performance
(Egner et al. 2004; Egner and Gruzelier 2004; Kropotov
et al. 2005; Kober et al. 2015a; Reichert et al. 2016). A
few single-case studies in stroke patients reported hetero-
geneous results. Some found positive effects of NF train-
ing on cognitive functions as well as a EEG normalization
after NF training (Rozelle and Budzynski 1995; Bearden
etal. 2003; Laibow etal. 2002; Putman 2002; Hofer etal.
2014), others could not find any significant effects (Dop-
pelmayr etal. 2007). However, the generalizability of these
prior findings is limited due to the incomplete descrip-
tion of training-specific EEG signal changes as well as the
absence of control groups. The majority of NF training
studies examined the effects of NF only on the behavioral
level (see Gruzelier 2014 for a review). Generally, suc-
cessful modulation of EEG band power is associated with
cognitive and behavioral improvements (Kropotov 2009;
Gruzelier 2014; Kober et al. 2015a, b; Hofer et al. 2014;
Reichert etal. 2016). For instance, voluntary up-regulation
of the upper alpha frequency band (UA, about 10–12Hz)
generally leads to improvements in working memory (WM)
and short-term memory performance (Escolano etal. 2011,
2012, 2013, 2014; Angelakis etal. 2007; Nan etal. 2012).
It is assumed that alpha activity inhibits unnecessary or
conflicting processes to the task being performed, thus
facilitating attention and memory by actively suppressing
distracting stimuli (Klimesch etal. 2007). Beside voluntary
modulation of the magnitude of EEG amplitudes, NF can
be also used to change the topographical distribution of
EEG activity. For instance, NF is used to tread depressive
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

71Appl Psychophysiol Biofeedback (2017) 42:69–83
1 3
symptoms by changing hemispheric asymmetry in alpha
band power (8–12Hz) in prefrontal brain areas (Kropotov
2009).
Summing up, we aimed at investigated the effects of NF
training on (i) electrical brain activity, such as power in
different EEG frequencies and the topographical distribu-
tion of EEG activity, and (ii) cognitive functions in chronic
stroke patients. Therefore, we present two cases, a stroke
patient with a unilateral middle cerebral artery (ACM)
stroke and a stroke patient with a bilateral subarachnoid
hemorrhage (SAH). We used an UA based NF training,
since the two chronic stroke patients showed deficits in
memory functions prior to the NF training. Based on the
literature, UA based NF training should have specific posi-
tive effects on memory functions (Escolano et al. 2011,
2012, 2013, 2014; Angelakis etal. 2007; Nan etal. 2012).
We compared the results of the two stroke patients to the
results of a healthy, elderly control group. We expected that
pathological EEG patterns in stroke patients would change
due to UA based NF training, which should be associated
with cognitive recovery.
Methods
Participants
Two stroke patients in a chronic state participated in the
NF training. Subject A was a 72-year-old right-handed man
who suffered a non-traumatic subarachnoid hemorrhage
(time since onset 27 months) at parieto-temporal regions
and a concomitant non-traumatic intracerebral hemorrhage
resulting in bilateral lesions in occipital–parietal and fron-
tal brain regions. He had no motor deficits. The neuropsy-
chological test-battery assessed before the start of the NF
training revealed cognitive deficits in different memory
functions (T-scores <40) (Fig. 1). Spatial, personal and
temporal orientation of subject A were intact. Subject B
was a 73-year-old right-handed man who suffered a mono-
hemispheric ischemic stroke (time since onset 70months).
The cerebral infarction was due to a thrombosis of the left
arteria cerebri media (ACM). The patient had no motor def-
icits or severe cognitive deficits. In the neuropsychological
pre-assessment he showed some deficits in short- and long-
term memory tasks (T-scores < 40) (Fig. 1). His spatial,
personal and temporal orientation were also intact.
Patient exclusion criteria were a visual hemi-neglect,
dementia (defined as a score of <24 in the mini-mental
state examination, MMSE Kessler etal. 1990), psychiatric
disorders such as depression or anxiety, other concomi-
tant neurological disorders (e.g. Parkinson disease; visual-
reflex epilepsy), aphasia, drug therapies that interfere with
the vigilance state, or insufficiently motivation and coop-
eration. Patients and healthy elderly controls had normal or
corrected-to-normal vision and hearing.
A neurologically healthy control group (CG) (N = 24,
33% male, mean age 63 years) was recruited as well. All
healthy participants disavowed any current or previous
Fig. 1 Test performance is expressed in T-scores with population
mean M = 50 and standard deviation SD = 10. Single subject test
scores and confidence intervals for measurements of attention, execu-
tive functions, short- and long-term memory, and working memory
(WM) performed during the pre- and post-assessment are depicted
separately for stroke patient A and B. Significant differences between
pre- and post-test (critical difference analysis on the group level,
Huber 1973) are marked with asterisks (*significant, +marginally
significant). CBTT corsi block tapping test, VVM visual and verbal
memory test
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

72 Appl Psychophysiol Biofeedback (2017) 42:69–83
1 3
psychiatric or neurological disorders. They received an
expense allowance of 7Euro per hour. The healthy CG
showed no deficits in any test parameter of the neuropsy-
chological assessment (all T-scores between 40 and 60).
All participants gave written informed consent to par-
ticipate prior to their inclusion in the study. We have
also obtained consent to publish from the participant
and to report individual patient data. The study was
approved by the local ethics committee of the University
of Graz (Reference No. GZ. 39/22/63 ex 2012/13 and
GZ. 39/11/63 ex 2013/14) and was in line with the code
of ethics of the World Medical Association, Declaration
of Helsinki.
Pre‑post‑assessment
Neuropsychological Assessment ofCognitive Functions
To evaluate possible effects of NF training on cognitive
functions, before and after the NF training all partici-
pants performed standardized neuropsychological tests
to assess attention, executive functions, short- and long-
term memory as well as working memory functions.
To assess the state of active attention, the subtest
Alertness of the Test of Attentional Performance (TAP)
test battery (Zimmermann and Fimm 2007–2012) was
administered. Divided attention was assessed with the
subtest Divided Attention of the TAP test battery (Zim-
mermann and Fimm 2007–2012) and inhibitory pro-
cesses were assessed with the subtest Go/NoGo of the
TAP test battery (Zimmermann and Fimm 2007–2012).
To investigate cognitive flexibility we used the sub-
test Flexibility of the TAP (Zimmermann and Fimm
2007–2012). Verbal long-term memory was measured
by using the Visual and Verbal Memory Test (VVM2)
subscale Construction 2 and non-verbal/visual–spatial
long-term memory was measured with the VVM2 sub-
scale City map 2 (Schelling and Schächtele 2001). Ver-
bal short-term memory was assessed with the VVM2
subscale Construction 1 (Schelling and Schächtele
2001) and with the forward task of the Digit Span test
of the Wechsler Memory Scale (WMS-R; Härting and
Wechsler 2000). Non-verbal or visual–spatial short-term
memory was assessed with the VVM2 subscale City map
1 (Schelling and Schächtele 2001) and the forward task
of the Corsi Block Tapping Test (CBTT; Härting and
Wechsler 2000). Additionally, the backwards tasks of the
Digit Span test and the CBTT were used to assess work-
ing memory performance (Härting and Wechsler 2000).
Parallel forms of the memory tests were used to avoid
learning effects.
EEG Resting Measurements
Before and after the NF training, all participants per-
formed EEG resting measurements with closed and open
eyes á 3min. For these multi-channel EEG recordings, a
BrainAmp Standard amplifier (Brain Products GmbH,
Munich, Germany) was used. EEG was recorded by Ag/
AgCl electrodes from 60 electrode positions according to
the extended 10–20 electrode placement system against a
linked mastoid reference, the ground was placed at FPz.
Vertical and horizontal EOG was recorded with three elec-
trodes in total, two were placed on the outer canthi of the
eyes and one was placed superior to the nasion. Electrode
impedances were kept below 5kOhms for the EEG record-
ing and below 10kOhms for the EOG recording. EEG sig-
nals were digitized at 500Hz and filtered with a 0.01 Hz
high-pass and a 100Hz low-pass.
Upper Alpha Neurofeedback Training
For the upper alpha NF training, EEG signal was recorded
using a 10-channel amplifier (NeXus-10 MKII, Mind Media
BV) with a sampling frequency of 256Hz, the ground was
located at the right mastoid, the reference was placed at
the left mastoid. The feedback electrode was placed over
Pz. One EOG channel was recorded at the left eye. The NF
paradigm was generated by using the software BioTrace+
(Mind Media BV, Kober etal. 2013). Up to ten NF training
sessions were carried out on different days 3–5 times per
week. Each session lasted approximately 45min and con-
sisted of a baseline run and six feedback runs á 3min each.
During the feedback runs, participants were rewarded by
getting points when they increased their upper alpha power
above an individually predefined threshold (mean power
of baseline run and previous runs), while keeping other
frequencies in the theta and beta range, which reflect eye
blinks/movements and muscle activity respectively, below
a certain threshold (mean power of baseline run + 1 SD)
(Weber etal. 2011; Doppelmayr and Weber 2011). Visual
feedback was provided by vertically moving bars depict-
ing the power values in the feedback frequency bands. The
upper alpha frequency range, which was used as feedback
frequency during NF training, was defined individually for
each single participant. Therefore, the EEG resting meas-
urements of the pre-assessment were used to calculate the
EEG power spectrum for each participant. EEG power
spectra were calculated using Fast Fourier Transforma-
tion (FFT). FFT was computed for the segmented resting
measurements (segment length 1 s) with maximum reso-
lution of ~0.98Hz. Furthermore, a 10% Hanning window
was applied including a variance correction to preserve
overall power. Afterwards, peak detection in the Alpha
frequency range was performed to identify the individual
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

73Appl Psychophysiol Biofeedback (2017) 42:69–83
1 3
alpha frequency (IAF). The upper and lower alpha band
were defined in the following way (Klimesch 1999):
lower alpha = (IAF −2 Hz) to IAF; upper alpha = IAF to
(IAF + 2 Hz). Both stroke patients showed a clear alpha
peak at 10Hz during the EEG resting measurements of the
pre-assessment (IAF = 10Hz). The CG showed a mean IAF
of 9.25Hz (SE = 0.20). No significant changes in IAF val-
ues due to NF training could be observed.
For statistical analysis, UA power values of the NF train-
ing sessions were z-transformed to ensure comparability
across sessions and subjects. To analyze more closely the
time course of UA power over the training runs averaged
over all sessions, we conducted linear regression analyses
(predictor variable = run; dependent variable = UA power).
As during successful NF, a within-session increase of the
feedback frequency power is expected, the slope of the
regression line was used as an indicator of NF performance
(see Escolano et al. 2012; Witte et al. 2013; Zoefel et al.
2011; Kober etal. 2015a).
EEG Data Preprocessing andAnalysis
Data analysis of all EEG recordings (EEG resting meas-
urements during pre- and post-assessment and NF train-
ing data) was performed offline using the Brain Vision
Analyzer software (version 2.01, Brain Products GmbH,
Munich, Germany). Artefacts (e.g. muscle activity) were
rejected by means of a semi-automatic artefact rejection
(criteria for rejection: > 50.00 µV voltage step per sam-
pling point, absolute voltage value > ± 120.00 µV). Ocular
artefacts (eye blinks) were automatically corrected using
the algorithm developed by Gratton, Coles and Donchin
(1983) (Gratton et al. 1983). All epochs with artefacts
were excluded from the EEG analysis. To analyze the feed-
back training data, absolute values of upper alpha (IAF to
(IAF + 2 Hz)) power were calculated and averaged sepa-
rately for each 3-min run of each sessions using the Brain
Vision Analyzer’s built-in method of complex demodula-
tion (method to calculate power, Brain Products GmbH
2009). The complex demodulation is based on the complex
(analytical) signal of a time series, where all frequencies
except the one of interest are filtered out (Draganova and
Popivanov 1999). To analyze the EEG resting measure-
ments, absolute values of delta 0.5-4Hz, theta 4–8Hz, low
alpha 8–10Hz, upper alpha 10–12Hz, low beta 12–15Hz,
mid beta 15–21 Hz, high beta 21–35 Hz, and gamma
35–45Hz power in a fixed range were calculated and aver-
aged separately for the eyes-open and eyes-closed condition
using the Brain Vision Analyzer’s built-in method of com-
plex demodulation. Note that no NF training effects were
found in theta, low alpha, beta, and gamma bands for the
resting measurements. Therefore, only the analysis of EEG
power in the delta and upper alpha band will be reported.
Description ofStatistical Analysis
In order to analyse the NF training performance, we
determined the time course of UA power averaged over
the NF training sessions across the six feedback runs
using linear regression analysis (see 2.3). In addition,
one-sample t tests against 0 were calculated for the CG to
verify the consistency of the learning effects. The slopes
of the regression line of the single stroke patients were
statistically compared to the slopes of the CG by apply-
ing single-case analysis methods (Crawford and Garth-
waite 2002). These methods enable the assessment of the
probability that test scores and test score discrepancies of
a single patient and a modest-sized control sample belong
to the same distribution (Crawford and Garthwaite 2002).
To analyze the results of the neuropsychological
assessment of cognitive functions, T-scores of the single
neuropsychological test parameters were used. To inves-
tigate the effects of NF on cognitive functions, we con-
ducted intra-individual comparisons between cognitive
performance assessed during pre- and post-assessment
by using critical difference analysis (Huber 1973). To
identify significant improvement or decline for each par-
ticipant, the critical difference of the relevant test param-
eter was compared with the test score difference obtained
during the post-assessment minus the pre-assessment.
To establish the critical difference for a pair of test
scores, a correction for measurement error based on the
test–retest reliability of the test is performed. The differ-
ence between pre- and post-assessment shown by the sin-
gle participants is considered significant when it is larger
than the critical difference, which can be detected by
each test (error probability α < 5%) and only occurs in the
population with a probability lower than α < 10%. Differ-
ences in T-scores between pre- and post-assessment for
each cognitive test parameter were compared with critical
differences on the single subject level as well as on the
group level.
To analyze the EEG activity of the CG during rest,
2 × 3 × 3 univariate repeated-measures analyses of variance
(ANOVA) with the within-subjects factors TIME (pre- vs.
post-assessment), ACP (anterior vs. central vs. posterior
electrode positions) and LATERALITY (left vs. midline
vs. right electrode positions) were calculated separately
for delta and upper alpha power and for the eyes-closed
and eyes-open resting condition. In the CG, delta power
was largest over FCz (see Fig.2). Therefore, the following
electrodes were chosen for statistical analysis of the topo-
graphical distribution of delta power in the CG: FC5, FCz,
FC6, C5, Cz, C6, CP5, CPz, and CP6. Upper alpha power
was most pronounced at parieto–occipital sites in healthy
controls (see Fig.3). For the statistical analysis of the topo-
graphical distribution of upper alpha power in the CG we
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

74 Appl Psychophysiol Biofeedback (2017) 42:69–83
1 3
Fig. 2 Topographical maps of
delta power during the eyes-
open (upper two rows) and
eyes-closed (lower two rows)
conditions assessed during the
pre- and post-test, presented
separately for the two single
stroke patients and the healthy
CG
Fig. 3 Topographical maps of
upper alpha power during the
eyes-open (upper two rows) and
eyes-closed (lower two rows)
conditions assessed during the
pre- and post-test, presented
separately for the two single
stroke patients and the healthy
CG. Note the different scaling
of EEG upper alpha power
between subjects in the eyes-
closed condition
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

75Appl Psychophysiol Biofeedback (2017) 42:69–83
1 3
chose the following electrodes: CP3, CPz, CP4, P3, Pz, P4,
PO3, POz, and PO4.
For the single stroke patients, we calculated difference
values (laterality values) in delta and upper alpha power
between electrodes of the left and right hemisphere. These
laterality values of the single stroke patients were statisti-
cally compared to the laterality values of the CG by apply-
ing single-case analysis methods (Crawford and Garthwaite
2002). Difference values in EEG power were also calcu-
lated between the pre- and post-test to reveal any alterations
in absolute EEG power due to NF training.
For all analyses, the probability of a Type I error was
maintained at 0.05. For the ANOVA’s, Mauchly’s tests of
sphericity were carried out on the repeated-measures vari-
ables, and where violated, Greenhous–Geisser correction
was applied. For post-hoc analyses, Bonferroni corrections
for multiple comparisons were applied.
Results
NF performance
Both stroke patients as well as the CG showed a linear
increase in UA power across the training runs within the
NF training sessions as indicated by positive regression
slopes (regression slope: subject A 0.19; subject B 0.18;
CG 0.05). The regression model was significant for subject
A and the CG (p < 0.05), while the regression model was
by trend significant for subject B (p = 0.08). For the CG, a
one-sample t-test revealed that the linear regression slopes
of healthy controls were significantly larger than zero (t
(23) = 2.15, p < 0.05). When analyzing the time course of
UA power over the training runs separately for each par-
ticipant of the CG, 15 out of 24 participants (i.e. 62.5%)
showed a positive gradient of the learning curve. One of
the remaining participants showed a flat learning curve gra-
dient and eight a negative gradient. A single-case analysis
(Crawford and Garthwaite 2004) indicated that the regres-
sion slope of subject A (t (23) = 1.15, p = 0.26, effect size
for difference between case and controls = 1.18) and subject
B (t (23) = 1.04, p = 0.31, effect size for difference between
case and controls = 1.06) did not differ significantly from
the healthy participants’ slopes, indicating that the abil-
ity of stroke patients to alter UA power was not different
from that of healthy controls. The increase in UA power
observed in stroke patients and controls across training
runs indicated successful NF training.
Effects onCognitive Functions
After the NF training, subject A showed significant
improvements in short-term memory and long-term
memory tasks (Digit Span forward task, VVM2 sub-
scale City map 2). His working memory performance also
slightly increased although this pre-post difference did not
exceed the critical difference value (Fig.1).
Subject B also showed performance improvements after
NF training compared to the pre-assessment. The strongest
increase in performance could be observed in a long-term
memory task, the subscale City map 2 of the VVM2. Fur-
thermore, subject B showed an improved performance in
the subtest Flexibility of the TAP after NF compared to the
pre-assessment (Fig.1).
The CG showed no significant changes in cognitive
functions when comparing the pre- and post-assessment.
Effects onEEG Activity During Rest
Delta Power
Figure 2 illustrates the topographical distribution of delta
power during the pre- and post-measurements separately
for the eyes-open and eyes-closed condition for the two
single stroke patients and the healthy CG. In the CG, delta
power showed a maximum over fronto-central midline elec-
trodes. The 2 × 3 × 3 ANOVA for the eyes-open condition
revealed a significant main effect ACP (F (2,46) = 17.87,
p < 0.01, η2 = 0.44). Delta power was higher at fronto–cen-
tral sites (FC: M = 6.72 µV2, SE = 0.48) than at central (C:
M = 6.23 µV2, SE = 0.46) and centro-parietal sites (CP:
M = 6.01 µV2, SE = 0.44). The main effect LATERALITY
was significant, too (F (2,46) = 96.53, p < 0.01, η2 = 0.81).
Posttests revealed that delta power was highest at midline
electrode positions (M = 9.49 µV2, SE = 0.73) than over
the left (M = 4.66 µV2, SE = 0.34) or right (M = 4.81 µV2,
SE = 0.35) hemisphere. Hence, the CG showed a bilateral
distribution of delta power. Furthermore, the interaction
effect ACP*LATERALITY was significant (F(4,92) = 5.97,
p < 0.01, η2=0.21). Posttests indicated a lower delta power
over CPz than over FCz and Cz.
For the eyes-closed condition, the ANOVA revealed
comparable results than for the eyes-open condition. The
main effect ACP (F (2,46) = 13.59, p < 0.01, η2=0.37) was
significant. Delta power was higher at fronto-central sites
(FC: M = 7.74 µV2, SE = 0.73) than at central (C: M = 7.08
µV2, SE = 0.67) and centro-parietal sites (CP: M = 6.74µV2,
SE = 0.61). The main effect LATERALITY was significant,
too (F(2,46) = 57.15, p < 0.01, η2 = 0.71). Posttests revealed
that delta power was highest at midline electrode positions
(M = 10.97µV2, SE = 1.13) than over the left (M = 5.19µV2,
SE = 0.44) or right (M = 5.40 µV2, SE = 0.47) hemisphere.
The interaction effect ACP*LATERALITY was also sig-
nificant (F (4,92) = 7.58, p < 0.01, η2 = 0.25). Posttests indi-
cated a lower delta power over CPz than over FCz and Cz.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

76 Appl Psychophysiol Biofeedback (2017) 42:69–83
1 3
As illustrated in Fig.2, subject B showed an increased
delta power over the right hemisphere during the pre-
assessment. During the post-assessment, subject B showed
a central maximum of delta power, which was compara-
ble to the bilateral distribution of delta power in the CG.
In contrast, subject A showed no such shift in delta power
when comparing the pre- and post-assessment. Delta
power was maximal over central electrode positions in
subject A during both time points. To test whether these
hemispheric differences in subject B were statistically sig-
nificant, we performed single-case analysis (Crawford and
Garthwaite 2002). Therefore, we calculated the difference
value of delta power over the right (C6) minus the left (C5)
hemisphere (laterality values). The mean laterality values
in delta power between the right and left hemisphere are
summarized in Table 1. One-sample t tests revealed that
the laterality values of the CG did not differ significantly
from zero (all p > 0.05). These results indicate that the CG
showed no laterality effects in delta power. During the pre-
assessment, subject B showed a significant larger laterality
value than the CG during the eyes-open and eyes-closed
condition. During the post-assessment, no significant dif-
ferences in laterality values between subject B and the CG
could be observed. Subject A did not differ in his laterality
values from the CG during all conditions.
Upper Alpha Power
Figure 3 illustrates the topographical distribution of UA
power during the different conditions for the two sin-
gle stroke patients and the healthy CG. In the CG, UA
power showed a maximum over parieto–occipital elec-
trodes. The 2 × 3 × 3 ANOVA for the eyes-open condition
of the CG revealed no significant results. After NF train-
ing, UA power values were numerically higher (Post:
M = 4.53 µV2, SE = 1.42) compared to the pre-assessment
(Pre: M = 2.62 µV2, SE = 0.39), although the main effect of
TIME (F (1,23) = 2.16, p = 0.15, η2 = 0.09) did not reach
significance.
For the eyes-closed condition, the ANOVA revealed
a significant main effect ACP (F (2,46) = 9.47, p < 0.01,
η2 = 0.29). UA power was higher at parietal (P:
M = 9.14 µV2, SE = 2.05) and parieto–occipital (PO:
M = 10.34µV2, SE = 2.17) sites compared to centro-pari-
etal sites (CP: M = 6.65µV2, SE = 1.32). No hemispher ic
differences could be observed in the CG.
In the eyes-open condition, subject B showed no
prominent changes in the topographical distribution of
UA power when comparing the pre- and post-assessment.
However, after NF training UA power was stronger over
parieto-occipital sites compared to the pre-test. For statis-
tical comparison, we compared the post minus pre differ-
ences in UA power of subject B with the same difference
values of the CG using single-case analysis (Crawford
and Garthwaite 2002). Subject B showed a statistically
stronger increase in UA power between pre- and post-
test compared to controls over parieto-occipital electrode
positions (Table 2). Subject A showed no changes in
absolute UA power values when comparing the pre- and
post-assessment.
In the eyes-closed condition, UA power of subject B
was more pronounced over the right compared to the left
hemisphere during the pre-assessment, while during the
post-assessment the topographical distribution became
Table 1 Laterality values in
delta power between the right
and left hemisphere (difference
value: C6-C5), presented
separately for the two single
stroke patients and the CG
Significant differences between the CG and single subjects (single-case analysis, Crawford and Garthwaite
2002) are marked with asterisks (*p < 0.05)
Difference in delta power between right and left hemi-
sphere (C6-C5)
CG (Mean and SE) Subject A Subject B
Pre assessment—eyes-open condition 0.17 (0.13) 0.85 2.21*
Post assessment—eyes-open condition 0.15 (0.12) 0.37 0.06
Pre assessment—eyes-closed condition 0.12 (0.16) 0.86 2.43*
Post assessment—eyes-closed condition 0.35 (0.12) 0.19 0.23
Table 2 Difference values in upper alpha power during the eyes-
open condition between the post- and pre-assessment for parietal and
parieto–occipital electrode positions, presented separately for the two
single stroke patients and the CG
Significant differences between the CG and single subjects (single-
case analysis, Crawford and Garthwaite 2002) are marked with aster-
isks (*p < 0.05; +p < 0.10)
Difference in upper alpha power between post- and pre-
assessment
CG (Mean and SE) Subject A Subject B
P3 1.92 (0.43) 0.21 4.22
Pz 2.67 (0.30) 0.23 4.64
P4 1.98 (0.43) 0.06 5.24
PO3 2.16 (0.64) 0.10 8.19+
POz 1.95 (0.49) 0.10 7.78*
PO4 2.37 (0.67) 0.02 8.30+
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

77Appl Psychophysiol Biofeedback (2017) 42:69–83
1 3
more bilateral (Fig.3). Subject A showed no prominent
UA maximum, neither during the eyes-open nor during
the eyes-closed condition. The CG showed a bilateral
distribution of UA power during all conditions. Single-
case analysis (Crawford and Garthwaite 2002) revealed
a stronger hemispheric difference in subject B compared
to the CG at different posterior electrode positions during
the pre-assessment. During the post-assessment, these
hemispheric differences in subject B did not differ signifi-
cantly from the laterality values of healthy controls. The
mean laterality values in UA power between the right and
left hemisphere are summarized in Table3. One-sample t
tests revealed that the laterality values of the CG did not
differ significantly from zero (all p > 0.05).
Discussion
Here we show that EEG based NF training led to cortical
reorganization in a stroke patient with pathological EEG
activation patterns. Furthermore, UA based NF training led
to improvements in memory functions in stroke survivors
that showed memory deficits prior to the training. In the
following, these results are discussed in more detail.
NF Performance
In a first step, we could show that the two stroke patients
were able to voluntarily increase their upper alpha power
through NF training. NF performance of stroke patients
and healthy controls was comparable. The CG as well as
the stroke patients showed a linear increase in the trained
frequency band over the feedback runs, indicating success-
ful NF training (Kober etal. 2015a). These results support
previous single-case studies showing that stroke patients
are able to control their electrical brain activity during NF
training (Bearden etal. 2003; Cannon etal. 2010; Reichert
et al. 2016) and extend the evidence by showing that the
stroke patients were equally able to up-regulate upper alpha
activity as age-matched healthy controls. In line with pre-
vious findings in healthy participants, about 1/3 of healthy
controls were not able to modulate their EEG activity (Alli-
son and Neuper 2010). The inability to control the own
brain activity may be attributed to different factors such as
differences in brain structure (Allison and Neuper 2010),
inter-individual differences in neurophysiological and
psychological factors, or cognitive strategies (Kober etal.
2013; Witte etal. 2013; Wood etal. 2014).
Effects oncognitive functions
Upper alpha based NF training had positive effects on
memory functions in stroke patients, which is in line with
previous findings in healthy people (Escolano etal. 2011,
2012, 2013, 2014; Angelakis etal. 2007; Nan etal. 2012).
Subject A, who suffered from a bilateral SAH, showed
the strongest deficits in memory functions prior to NF
training. After NF training, his performance significantly
increased in short- and long-term memory tasks com-
pared to the pre-assessment. Before the start of the NF
training, subject A’s performance in the Digit Span for-
ward task and the VVM2 subscale City map 2 was below
average (T-scores < 40). During the post-assessment his
performance in these two tasks was in a normal range.
He also showed marginal performance improvements in
working memory. Subject B, with lesions in the left hem-
isphere due to an ischemic stroke, also showed perfor-
mance improvements in different memory tasks, although
only the pre-post difference in the VVM2 subscale City
map 2 reached significance. Generally, upper alpha activ-
ity is associated with improved stimulus processing and
inhibiting unnecessary or conflicting processes to the
task being performed, which should foster the storage and
retrieval of information and might explain the improve-
ments in memory functions (Klimesch etal. 2007). A few
prior single-case studies also found positive effects of NF
Table 3 Laterality values in
upper alpha power between
right and left hemisphere
(difference values: CP4−CP3,
P4−P3, PO4−PO3), presented
separately for the two single
stroke patients and the CG
Significant differences between the CG and single subjects (single-case analysis, Crawford and Garthwaite
2002) are marked with asterisks (*p < 0.05)
CG (Mean and SE) Subject A Subject B
Difference in upper alpha power between right and left hemisphere (CP4-CP3)
Pre assessment—eyes-closed condition −0.41 (0.49) 0.76 6.66*
Post assessment—eyes-closed condition −0.03 (0.57) 0.44 4.31
Difference in upper alpha power between right and left hemisphere (P4-P3)
Pre assessment—eyes-closed condition 0.24 (0.59) 0.66 12.93*
Post assessment—eyes-closed condition 1.22 (0.78) 0.37 6.87
Difference in upper alpha power between right and left hemisphere (PO4-PO3)
Pre assessment—eyes-closed condition 0.38 (0.71) 0.40 14.96*
Post assessment—eyes-closed condition 0.95 (0.78) 0.31 7.64
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

78 Appl Psychophysiol Biofeedback (2017) 42:69–83
1 3
training on cognitive functions in stroke patients (Rozelle
and Budzynski 1995; Bearden etal. 2003; Putman 2002;
Hofer et al. 2014; Cannon et al. 2010; Reichert et al.
2016). However, only a few studies explicitly investigated
the effects of upper alpha based NF training on cognitive
functions in stroke survivors (Doppelmayr et al. 2007;
Kober et al. 2015b). Doppelmayr et al. (2007) reported
the results of two NF training studies. In their first exper-
iment, they found stronger positive effects of upper alpha
based NF training on memory functions in a group of
stroke patients compared to a control group of stroke
patients that received relaxation training. In a second
study, they could not replicate their findings. Upper alpha
based NF training had comparable effects as theta based
NF training and a control intervention (Doppelmayr etal.
2007). Nevertheless, in both experiments, they found an
improved memory performance in stroke patients after
upper alpha NF compared to the pre-test. The descrip-
tion of the stroke patients that participated in the study
of Doppelmayer et al. (2007) is lacking, which dimin-
ishes the comparability of the present findings in stroke
patients to the results of Doppelmayer etal. (2007). In
line with the present results, Kober et al. (2015b) also
reported on positive effects of upper alpha based NF
training on memory functions. The present findings indi-
cate that NF might be a suitable cognitive rehabilitation
tool for stroke patients and should be investigated further
in future studies. Research on NF for stroke rehabilitation
is of special importance as positive effects of traditional
cognitive trainings in this domain remain disputed (Nair
and Lincoln 2007; Hoffmann etal. 2010).
Concerning the specificity of upper alpha based NF
training, we found the strongest effects of UA NF on
memory functions, except for subject B who also showed
an improved performance in cognitive flexibility after NF
compared to the pre-assessment. The majority of prior
NF studies also linked upper alpha to memory functions,
whereas performance improvements in other cognitive
functions such as attention, inhibitory control, or cognitive
flexibility were mainly found in theta/beta based NF train-
ing studies (Hofer etal. 2014; Kropotov 2009; Arns etal.
2009; Fox etal. 2005; Gruzelier 2014). In the NF litera-
ture, there are issues concerning specificity of NF training
(Gruzelier 2014). To a large extent, our results indicate that
there is specificity and independence regarding cognitive
outcome such that performance enhancement in memory
functions is specific to changes in UA frequency band
while leaving other cognitive functions unchanged, except
for cognitive flexibility.
Healthy controls, who showed no deficits in cognitive
functions prior to the NF training, showed no significant
changes in cognitive performance when comparing the pre-
and post-assessment.
Effects onEEG Activity During Rest
Upper alpha based NF training led to a topographical nor-
malization in EEG resting delta and upper alpha power in
subject B, who showed pathological deviations in EEG
topography prior to NF training compared to healthy con-
trols. The healthy control group showed a fronto-central
maximum of delta power during the pre- as well as during
the post-test. This is in line with prior findings in healthy
people showing that delta activity is largest over fronto-cen-
tral brain regions (Niedermeyer and Lopes da Silva 2005;
Emek-Savaş etal. 2015; Schmiedt-Fehr and Basar-Eroglu
2011). Upper alpha power was largest over bilateral parieto-
occipital electrode positions in the CG (Klimesch 1999).
Subject B, who had lesions in the left hemisphere due to
a stroke in the left ACM, showed an increased delta power
over the healthy, right hemisphere compared to the affected,
left hemisphere during the pre-test. Furthermore, before the
start of the NF training upper alpha power was more pro-
nounced over right than over left parieto–occipital brain
regions in subject B. Single-case analysis revealed that this
hemispheric asymmetry in delta and upper alpha power
was significantly larger in subject B compared to healthy
controls during the pre-assessment. There is evidence that
an increased delta activity over the unaffected hemisphere
is associated with poor recovery, a bad health status and
even with earlier death (Finnigan etal. 2008; Sheorajpan-
day etal. 2011; Tecchio etal. 2007; Zappasodi etal. 2007;
Niedermeyer 2005; Rossini et al. 2003). A higher alpha
power over the unaffected compared to the affected hemi-
sphere in stroke patients with unilateral lesions was also
linked to poor clinical recovery (Giaquinto etal. 1994; Tec-
chio etal. 2006). Generally, an increased activity over the
unaffected hemisphere in the chronic state after a stroke
is associated with “maladaptive plasticity” (Rossini etal.
2003). An increased delta power over the unaffected hemi-
sphere in subject B might be explained by a reduced blood
flow in the unaffected hemisphere (Giaquinto etal. 1994)
since there is evidence that delta power is negatively cor-
related with rCBF (Tolonen and Sulg 1981; Finnigan and
van Putten 2013). In this context, studies in patients with
infarctions of the ACM showed distant effects of some
strokes such as reduced cerebral blood flow and reduced
metabolic rates in contra-lesional regions, which might be
related to the spill-over of EEG slowing into the healthy
hemisphere (Niedermeyer 2005). Tecchio etal. (2007) also
found increased delta power values in the unaffected hemi-
sphere in patients with unilateral stroke in the ACM, which
was related to a bad clinical long-term outcome prognosis.
They also mentioned that a loss of excitatory input from the
affected hemisphere might cause reduced metabolism in
remote connected, healthy areas. Consequently, the authors
concluded that clinical recovery might be paired with shifts
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

79Appl Psychophysiol Biofeedback (2017) 42:69–83
1 3
of excitability balance between the affected and unaffected
hemisphere (Tecchio etal. 2007). In line with this assump-
tion, subject B showed a more bilateral distribution of
delta power, with a fronto-central maximum after NF train-
ing, which was comparable to the topographical distribu-
tion of delta in the CG. Upper alpha power also showed a
shift from a maximum over the unaffected healthy hemi-
sphere during the pre-assessment to a more central, bilat-
eral maximum during the post-assessment. The changes in
topography of delta and upper alpha power after NF train-
ing in subject B compared to the pre-test might be a sign
of “normalization” of neural activity. Single-case analysis
revealed that the difference in EEG power between the left
and right hemisphere in subject B regained values similar
to healthy elderly controls after NF training. This reduction
in pathological delta and upper alpha asymmetry in sub-
ject B was accompanied by cognitive improvements. The
return toward homeostatic conditions of pathological brain
mechanisms is in line with previous findings that showed
that clinical recovery after stroke is accompanied by a
reduction of inter-hemispheric asymmetries (Tecchio etal.
2006; Giaquinto etal. 1994; Rossini et al. 2003; Finnigan
etal. 2006; de Vos etal. 2008). Rossini etal. (2003) also
mentioned that an effective recovery after stroke is associ-
ated with a gradual normalization of the initially excessive
intensity as well as with a normalization of the balance of
hemispheric activation away from predominant activation
in the unaffected, healthy hemisphere.
The observed topographical changes in delta and upper
alpha power might be a sign of brain plasticity processes
due to NF training in chronic stroke. Functional recovery
after stroke is generally sustained by brain plasticity involv-
ing synaptogenesis, dendritic arborisation, as well as syn-
aptic and axonal recruitment (Rossini etal. 2003; Sheora-
jpanday etal. 2011). Such brain plasticity processes might
be fostered by NF training, since NF enables a direct access
to neural activity, which might alter or accelerate functional
reorganization in the brain following stroke (Nelson 2007).
Our results indicate that NF can lead to brain plasticity
processes in chronic stroke patients. According to Ross-
ini etal. (2003), reorganization in response to therapeutic
interventions can be studied best in patients in the chronic
stage of stroke. Any recovery is likely to be the result of
the specific intervention, since the probability of spontane-
ous recovery is negligible at this stage (Rossini etal. 2003).
Furthermore, the probability that topographical changes
in EEG power found in subject B between pre- and post-
assessment were due to random fluctuations in the EEG
signal is relatively low. Sheorajpanday etal. (2009) inves-
tigated the general reproducibility of EEG spectral power
and asymmetry of EEG activity in stroke patients. They
found relatively high Cronbach alpha values ranging from
0.95 to 0.99 for both EEG power and asymmetry indices
(Sheorajpanday etal. 2009). These results show that EEG
measures are robust parameters with a high reproducibility
in stroke patients.
Beside changes in EEG topography, subject B showed
an increased upper alpha power over parieto–occipital sites
after NF training compared to the pre-test. The healthy con-
trols also showed numerically increased upper alpha power
values when comparing the pre- and post-resting meas-
urements, however, these differences were not statistically
significant. Single-case analysis revealed that subject B
showed a significantly higher increase in upper alpha power
after NF training compared to the pre-test than the CG.
There is evidence that alpha power increases with clinical
recovery in stroke victims (Giaquinto etal. 1994; Finnigan
etal. 2007). An alpha amplitude attenuation is also gener-
ally indicative for cortical injury (Finnigan and van Putten
2013; Finnigan et al. 2007; Klimesch 1999). In healthy
people, an increased tonic alpha power is associated with
good cognitive performance, mainly memory performance
(Klimesch 1999). Hence, the increase in upper alpha power
due to NF training in subject B might be related to the
improvement of cognitive functions.
Subject A, who suffered bilateral lesions due to a SAH,
showed no hemispheric asymmetries in EEG power, neither
before nor after NF training. This indicates that pathologic
hemispheric asymmetries in EEG activity, which could
only be observed in subject B with a left ACM stroke,
might be specific for mono-hemispheric brain lesions. This
is in line with prior assumptions that lateralization of slow
activity such as delta power may be indicative of the pri-
marily involved hemisphere (Niedermeyer 2005). Subject
A also showed no prominent topographical alpha power
maximum during rest. In this context, Niedermayer (2005)
mentioned that the EEG of patients with SAH shows dif-
fuse changes, disorganization, and a disruption of the pos-
terior alpha rhythm, which might explain the missing topo-
graphical alpha power maximum over parieto–occipital
sites in subject A.
Subject A showed cognitive improvements although no
changes in EEG resting activity could be observed when
comparing the pre- and post-assessment. Probably because
of the diffuse and disorganized EEG and the disruption of
the posterior alpha rhythm, which might be caused by the
SAH, the method of EEG might have been not sensitive
enough to disclose brain plasticity processes in subject A
caused by NF (Niedermeyer 2005). Nevertheless, subject A
could learn from NF training. He was able to voluntarily
increase his upper alpha power within the training sessions
and showed cognitive improvements.
We observed differences in EEG activity during resting
conditions between stroke patients and controls only in the
delta and upper alpha frequency range. For instance, sub-
ject B showed no hemispheric asymmetries in the theta,
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

80 Appl Psychophysiol Biofeedback (2017) 42:69–83
1 3
beta or gamma frequency range before the start of the NF
training. Prior studies that showed deviant EEG activa-
tion patterns in stroke patients constantly reported abnor-
malities in the delta frequency range (Giaquinto etal. 1994;
Finnigan etal. 2007, 2008; Finnigan and van Putten 2013;
Fernandez-Bouzas etal. 2000; Claassen etal. 2004; Sheo-
rajpanday etal. 2011; Tecchio etal. 2006). The majority
of prior studies also demonstrated pathological alpha activ-
ity after stroke (Giaquinto etal. 1994; Finnigan etal. 2007;
Finnigan and van Putten 2013; Sheorajpanday etal. 2011;
Claassen etal. 2004; Tecchio etal. 2006; Fernandez-Bou-
zas et al. 2000). However, a few studies also reported on
pathological theta or beta activity after stroke, which could
not be observed in the present investigation (Giaquinto
et al. 1994; Tecchio et al. 2006, 2007; Finnigan and van
Putten 2013).
An important difference between prior studies and the
present investigation is that in prior studies alpha power was
defined within a broader frequency range of about 8–12Hz
(Giaquinto etal. 1994; Sheorajpanday etal. 2011; Claassen
etal. 2004; Tecchio et al. 2006; Finnigan et al. 2007). In
the present study, we split up the alpha frequency band in a
lower (8–10Hz) and upper (10–12Hz) alpha power range,
as suggested by Klimesch (1999). Upper and lower alpha
power are linked to different cognitive processes. While
lower alpha activity is associated with attentional processes
that are relatively task- and stimulus-non-specific, upper
alpha activity is specifically related to (semantic) long-term
memory performance. Especially, search and retrieval pro-
cesses in (semantic) long-term memory are reflected by
upper alpha oscillations (Klimesch 1999). Furthermore,
we chose upper alpha as feedback frequency during NF
training, since prior NF training studies indicated positive
effects of upper alpha based NF training on memory func-
tions (Escolano et al. 2011, 2012, 2013, 2014; Angelakis
etal. 2007; Nan etal. 2012). Hence, the increased resting
state upper alpha power, which we observed in subject B
and also marginally in the CG when comparing the multi-
channel EEG measurements performed during pre- and
post-assessment, might be indicative for specific improve-
ments in memory functions due to upper alpha based
NF, such as enhanced memory retrieval. In contrast, the
absence of changes in lower alpha power might be a sign
that upper alpha based NF had no effects on general atten-
tional functions.
Changes in delta power due to NF training were com-
parable between the eyes-open and eyes-closed resting
condition. In upper alpha power, topographical shifts
between pre- and post-test could only be observed in the
eyes-closed condition. Generally, alpha power is more
pronounced when they eyes are closed than during eyes-
open conditions. When participants open their eyes, EEG
oscillations in the alpha band decrease in amplitude or
disappear completely. This well-known and robust phe-
nomenon in called “alpha blocking” (Niedermeyer and
Lopes da Silva 2005; Kirschfeld 2005; Pfurtscheller and
Lopes da Silva 1999). Probably, topographical shifts in
upper alpha power could only be observed in subject B
during the eyes-closed condition because alpha oscilla-
tions were more pronounced during this resting meas-
urement compared to the eyes-open condition. Subject
B also showed larger upper alpha amplitudes during the
eyes-closed condition compared to subject A and the
healthy CG. However, it is well known that alpha fre-
quency shows large inter-individual differences (Klime-
sch 1999).
Conclusions
Here we showed that upper alpha based NF training had (i)
positive effects on memory functions and (ii) led to neu-
ronal plasticity processes in chronic stroke victims. Ini-
tial pathological EEG activation patterns normalized after
repeated NF training in a patient with unilateral stroke.
Although changes in EEG activity during rest could only
be observed in a stroke patient with unilateral stroke,
both stroke patients could benefit from NF training. The
NF training protocol was feasible for stroke patients with
memory deficits and may represent a new rehabilitation
strategy suitable to overcome some of the usual pitfalls
of traditional cognitive rehabilitation. NF seems to be an
alternative, innovative and easy-to-use cognitive rehabilita-
tion tool since the electrical activity of the brain is affected
directly and, therefore, the cortical substrates of cognitive
functions themselves (Nelson 2007).
Acknowledgements This work was partially supported by the Euro-
pean Community’s Seventh-Framework-Programme FP7/2007-2013,
CONTRAST project, under Grant agreement no. 287320. Possible
inaccuracies of information are under the responsibility of the project
team. The text reflects solely the views of its authors. The European
Commission is not liable for any use that may be made of the infor-
mation contained therein. The authors are grateful to Julia Schobel
and Jasmin Wiesler for data acquisition.
Compliance with Ethical Standards
Conflict of Interest Statement The authors declare that they have
no conflict of interest.
Ethical standards All participants gave written informed consent to
participate prior to their inclusion in the study. We have also obtained
consent to publish from the participant and to report individual patient
data. The study was approved by the local ethics committee of the Uni-
versity of Graz (reference number GZ. 39/22/63 ex 2012/13 and GZ.
39/11/63 ex 2013/14) and was in line with the code of ethics of the
World Medical Association, Declaration of Helsinki.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

81Appl Psychophysiol Biofeedback (2017) 42:69–83
1 3
Open Access This article is distributed under the terms of the
Creative Commons Attribution 4.0 International License (http://
creativecommons.org/licenses/by/4.0/), which permits unrestricted
use, distribution, and reproduction in any medium, provided you give
appropriate credit to the original author(s) and the source, provide a
link to the Creative Commons license, and indicate if changes were
made.
References
Allison, B., & Neuper, C. (2010). Could anyone use a BCI? In D. Tan
& A. Nijholt (Eds.), Brain–computer interfaces: Human–Com-
puter interaction series (pp.35–54). London: Springer.
Angelakis, E., Stathopoulou, S., Frymiare, J. L., Green, D. L., Lubar,
J. F., & Kounios, J. (2007). EEG neurofeedback: A brief over-
view and an example of peak alpha frequency Training for cog-
nitive enhancement in the elderly. The Clinical Neuropsycholo-
gist, 21(1), 110–129.
Arns, M., de Ridder, S., Strehl, U., Breteler, M., & Coenen, T. (2009).
Efficacy of neurofeedback treatment in ADHD: The effects on
inattention, impulsivity and hyperactivity: A meta-analysis. Clin-
ical EEG and Neuroscience: Official Journal of the EEG and
Clinical Neuroscience Society (ENCS), 40(3), 180–189.
Bearden, T., Cassisi, J., & Pineda, M. (2003). Neurofeedback training
for a patient with thalamic and cortical infarctions. Applied Psy-
chophysiology and Biofeedback, 28(3), 241–253.
Brain Products GmbH (2009). Brainvision Analyzer 2.0.1 user man-
ual, (3rd ed.). Munich.
Cannon, K. B., Sherlin, L., & Lyle, R. R. (2010). Neurofeedback Effi-
cacy in the treatment of a 43-year-old female stroke victim: A
case study. Journal of Neurotherapy, 14(2), 107–121.
Claassen, J., Hirsch, L. J., Kreiter, K. T., Du, E. Y., Sander Connolly,
E., Emerson, R. G., & Mayer, S. A. (2004). Quantitative continu-
ous EEG for detecting delayed cerebral ischemia in patients with
poor-grade subarachnoid hemorrhage. Clinical Neurophysiology,
115(12), 2699–2710.
Crawford, J., & Garthwaite, P. (2004). Statistical methods for sin-
gle-case studies in neuropsychology: comparing the slope of a
patient’s regression line with those of a control sample. Cor-
tex: A Journal Devoted to the Study of the Nervous System and
Behavior, 40, 533–548.
Crawford, J., & Garthwaite, P. H. (2002). Investigation of the single
case in neuropsychology: Confidence limits on the abnormality
of test scores and test score differences. Neuropsychologia, 40,
1196–1208.
de Vos, C. C., van Maarseveen, S. M., Brouwers, P., & Putten, M.
(2008). Continuous EEG monitoring during thrombolysis in
acute hemispheric stroke patients using the brain symmetry
index. Journal of Clinical Neurophysiology, 25(2), 77–82.
de Weerd, A. W., Veldhuizen, R. J., Veering, M. M., Poortvliet, D.
C. J., & Jonkman, E. J. (1988). Recovery from cerebral ischae-
mia. EEG, cerebral blood flow and clinical symptomatology in
the first three years after a stroke. Electroencephalography and
Clinical Neurophysiology, 70(3), 197–204.
Doppelmayr, M., Nosko, H., Pecherstorfer, T., & Fink, A. (2007). An
Attempt to Increase Cognitive Performance After Stroke With
Neurofeedback. Biofeedback and Self-Regulation, 3, 126–130.
Doppelmayr, M., & Weber, E. (2011). Effects of SMR and Theta/Beta
Neurofeedback on Reaction Times, Spatial Abilities, and Crea-
tivity. Journal of Neurotherapy, 15(2), 115–129.
Draganova, R., & Popivanov, D. (1999). Assessment of EEG fre-
quency dynamics using complex demodulation. Physiologi-
cal research/Academia Scientiarum Bohemoslovaca, 48 (2),
157–165.
Egner, T., & Gruzelier, J. (2004). EEG biofeedback of low beta band
components: frequency-specific effects on variables of attention
and event-related brain potentials. Clinical Neurophysiology,
115(1), 131–139.
Egner, T., Zech, T., & Gruzelier, J. (2004). The effects of neurofeed-
back training on the spectral topography of the electroencephalo-
gram. Clinical Neurophysiology, 115(11), 2452–2460.
Emek-Savaş, D. D., Güntekin, B., Yener, G. G., & Başar, E. (2015).
Decrease of delta oscillatory responses is associated with
increased age in healthy elderly. International Journal of
Psychophysiology.
Escolano, C., Aguilar Herrero, M., & Minguez, J. (2011). EEG-based
upper alpha neurofeedback training improves working memory
performance. 33rd Annual International IEEE EMBS Confer-
ence, August 30–September 3, 2011, Boston Marriott Copley
Place, Boston, MA, 2327–2330.
Escolano, C., Navarro-Gil, M., Garcia-Campayo, J., Congedo, M.,
De_Ridder, D., & Minguez, J. (2014). A controlled study on the
cognitive effect of alpha neurofeedback training in patients with
major depressive disorder. Frontiers in Behavioral Neuroscience,
8, 1–12.
Escolano, C., Navarro-Gil, M., Garcia-Campayo, J., & Minguez,
J. (2013). EEG-based upper-alpha neurofeedback for cogni-
tive enhancement in major depressive disorder: A preliminary,
uncontrolled study. Conference proceedings, 2013, 6293–6296.
Escolano, C., Olivan, B., Lopez-del-Hoyo, Y., Garcia-Campayo, J., &
Minguez, J. (2012). Double-blind single-session neurofeedback
training in upper-alpha for cognitive enhancement of healthy
subjects. Conference proceedings, 2012, 4643–4647.
Fernandez-Bouzas, A., Harmony, T., Fernandez, T., Silva-Pereyra, J.,
Valdés, P., Bosch, J., Aubert, E., Casián, G., Ojeda, G., Ricardo,
J., Hernández-Ballesteros, A., & Santiago, E. (2000). Sources of
abnormal EEG activtiy in brain infarctions. Clinical EEG (elec-
troencephalography), 31 (4), 165–169.
Finnigan, S., & van Putten, M. J. (2013). EEG in ischaemic stroke:
Quantitative EEG can uniquely inform (sub-)acute prognoses
and clinical management. Clinical Neurophysiology, 124(1),
10–19.
Finnigan, S. P., Rose, S. E., & Chalk, J. B. (2006). Rapid EEG
changes indicate reperfusion after tissue plasminogen activator
injection in acute ischaemic stroke. Clinical Neurophysiology,
117(10), 2338–2339.
Finnigan, S. P., Rose, S. E., & Chalk, J. B. (2008). Contralateral hemi-
sphere delta EEG in acute stroke precedes worsening of symp-
toms and death. Clinical Neurophysiology, 119(7), 1689–1695.
Finnigan, S. P., Walsh, M., Rose, S. E., & Chalk, J. B. (2007). Quan-
titative EEG indices of sub-acute ischaemic stroke correlate
with clinical outcomes. Clinical Neurophysiology, 118(11),
2525–2532.
Fox, D. J., Tharp, D. F., & Fox, L. C. (2005). Neurofeedback: An
alternative and efficacious treatment for attention deficit hyper-
activity disorder. Applied Psychophysiology and Biofeedback,
30(4), 365–373.
Giaquinto, S., Cobianchi, A., Macera, F., & Nolfe, G. (1994). EEG
recordings in the course of recovery from stroke. Stroke: A Jour-
nal of Cerebral Circulation, 25, 2204–2209.
Gratton, G., Coles, M. G., & Donchin, E. (1983). A new method for
off-line removal of ocular artifact. Electroencephalography and
Clinical Neurophysiology, 55(4), 468–484.
Gruzelier, J. H. (2014). EEG-neurofeedback for optimising per-
formance. I: A review of cognitive and affective outcome in
healthy participants. Neuroscience & Biobehavioral Reviews, 44,
124–141.
Härting, C., & Wechsler, D. (2000). Wechsler-Gedächtnistest revidi-
erte Fassung: WMS-R: Manual; deutsche Adaptation der revidi-
erten Fassung der Wechsler Memory Scale. Bern: Huber.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

82 Appl Psychophysiol Biofeedback (2017) 42:69–83
1 3
Hofer, D., Kober, S. E., Reichert, J., Krenn, M., Farveleder, K.,
Grieshofer, P., Neuper, C., & Wood, G. (2014). Spezifische
Effekte von EEG basiertem Neurofeedbacktraining auf kog-
nitive Leistungen nach einem Schlaganfall: Ein nutzvolles
Werkzeug für die Rehabilitation? Lernen und Lernstörungen,
3, 1–19.
Hoffmann, T., Bennett, S., Koh, C.-L., & McKenna, K. (2010). A
systematic review of cognitive interventions to improve func-
tional ability in people who have cognitive impairment following
stroke. Topics in Stroke Rehabilitation, 17(2), 99–107.
Huber, H. P. (1973). Psychometrische Einzelfalldiagnostik. Wein-
heim: Beltz.
Jonkman, E. J., van Dieren, A., Veering, M. M., Ponsen, L., Da Silva,
F. H. L., & Tulleken, C. A. F. (1984). EEG and CBF in cerebral
Ischemia. Follow-up studies in humans and monkeys. In E. J. a.
F. L. D. S. G. Pfurtscheller (Ed.), Progress in Brain Research
Brain Ischemia: Quantitative Eeg and Imaging Techniques
(pp.145–171): Elsevier.
Jordan, K. (2004). Emergency EEG and continuous EEG monitoring
in acute ischemic stroke. Journal of Clinical Neurophysiology,
21(5), 341–352.
Kaplan, P., & Rossetti, A. (2011). EEG patterns and imaging cor-
relations in encephalopathy: encephalopathy Part II. Journal of
Clinical Neurophysiology, 28(3), 233–251.
Kessler, J., Markowitsch, H. J., & Denzler, P. (1990). Mini mental sta-
tus examination MMSE: German version. Weinheim: Beltz.
Kirschfeld, K. (2005). The physical basis of alpha waves in the elec-
troencephalogram and the origin of the ?Berger effect? Biologi-
cal Cybernetics, 92(3), 177–185.
Klimesch, W. (1999). EEG alpha and theta oscillations reflect cog-
nitive and memory performance: A review and analysis. Brain
Research, 29(2–3), 169–195.
Klimesch, W., Sauseng, P., & Hanslmayr, S. (2007). EEG alpha
oscillations: The inhibition-timing hypothesis. Brain Research
Reviews, 53(1), 63–88.
Kober, S. E., Schweiger, D., Witte, M., Reichert, J. L., Grieshofer, P.,
Neuper, C., & Wood, G. (2015b). Specific effects of EEG based
neurofeedback training on memory functions in post-stroke vic-
tims. Journal of Neuroengineering and Rehabilitation, 12(107),
1–13.
Kober, S. E., Witte, M., Ninaus, M., Neuper, C., & Wood, G. (2013).
Learning to modulate one’s own brain activity: the effect of
spontaneous mental strategies. Frontiers in Human Neurosci-
ence. doi:10.3389/fnhum.2013.00695
Kober, S. E., Witte, M., Stangl, M., Valjamae, A., Neuper, C., &
Wood, G. (2015a). Shutting down sensorimotor interference
unblocks the networks for stimulus processing: An SMR neuro-
feedback training study. Clinical Neurophysiology Official Jour-
nal of the International Federation of Clinical Neurophysiology,
126(1), 82–95.
Kropotov, J. D. (2009). Quantitative EEG, event-related potentials
and neurotherapy (1stedn.). Amsterdam :Elsevier/Academic.
Kropotov, J. D., Grin-Yatsenko, V. A., Ponomarev, V. A., Chutko, L.
S., Yakovenko, E. A., & Nikishena, I. S. (2005). ERPs correlates
of EEG relative beta training in ADHD children. International
Journal of Psychophysiology, 55(1), 23–34.
Labar, D. R., Fisch, B. J., Pedley, T. A., Fink, M. E., & Solomon, R.
A. (1991). Quantitative EEG monitoring for patients with suba-
rachnoid hemorrhage. Electroencephalography and Clinical
Neurophysiology, 78(5), 325–332.
Laibow, R. E., Stubblebine, A. N., Sandground, H., & Bounias, M.
(2002). EEG-NeuroBioFeedback treatment of patients with brain
injury: Part 2: Changes in EEG parameters versus rehabilitation.
Journal of Neurotherapy, 5(4), 45–71.
Mattia, D., Spanedda, F., Babiloni, F., Romigi, A., & Marciani, M.
G. (2003). Quantitative EEG patterns following unilateral stroke:
a study in chronic stage. The International Journal of Neurosci-
ence, 113(4), 465–482.
Melkas, S., Jokinen, H., Hietanen, M., & Erkinjuntti, T. (2014). Post-
stroke cognitive impairment and dementia: prevalence, diagno-
sis, and treatment. Degenerative Neurological and Neuromuscu-
lar Disease, 4, 21–27.
Nair, R. D., & Lincoln, N. B. (2007). Cognitive rehabilitation for
memory deficits following stroke. The Cochrane database of sys-
tematic reviews (3), CD002293.
Nan, W., Rodrigues, J. P., Ma, J., Qu, X., Wan, F., Mak, P.-I., Mak,
P. U., Vai, M. I., & Rosa, A. (2012). Individual alpha neurofeed-
back training effect on short term memory. International Journal
of Psychophysiology, 86(1), 83–87.
Nelson, L. (2007). The Role of biofeedback in stroke rehabilitation:
Past and future directions. Topics in Stroke Rehabilitation, 14(4),
59–66.
Niedermeyer, E. (2005). Cerebrovascular disorders and EEG. In E.
Niedermeyer, & F. H. Lopes da Silva (Eds.), Electroencephalog-
raphy: Basic principles, Clinical applictions, and related fields.
(5 th ed.) (pp.339–362). Baltimore, MD: Lippincott Williams &
Wilkins.
Niedermeyer, E., & Lopes da Silva, F. H. (Eds.) (2005). Electroen-
cephalography: Basic principles, clinical applictions, and related
fields. (5 th ed.). Baltimore, MD: Lippincott Williams & Wilkins.
Pfurtscheller, G., & Lopes da Silva, F. H. (1999). Event-related EEG/
MEG synchronization and desynchronization: Basic principles.
Clinical Neurophysiology, 110(11), 1842–1857.
Putman, J. (2002). EEG Biofeedback on a female stroke patient with
depression: A case study. Journal of Neurotherapy, 5(3), 27–38.
Reichert, J. L., Kober, S. E., Schweiger, D., Grieshofer, P., Neuper, C.,
& Wood, G. (2016). Shutting down sensorimotor interferences
after stroke. A proof-of-principle SMR neurofeedback study.
Frontiers in Human Neuroscience, 10(348), 1–14.
Rossini, P. M., Calautti, C., Pauri, F., & Baron, J.-C. (2003). Post-
stroke plastic reorganisation in the adult brain. The Lancet Neu-
rology, 2(8), 493–502.
Rozelle, G. R., & Budzynski, T. H. (1995). Neurotherapy for stroke
rehabilitation: A single case study. Biofeedback and self-regula-
tion, 20(3), 211–228.
Schelling, D., & Schächtele, B. (2001). VVM. Visueller und Verbaler
Merkfähigkeitstest. Göttingen: Hogrefe.
Schmiedt-Fehr, C., & Basar-Eroglu, C. (2011). Event-related delta
and theta brain oscillations reflect age-related changes in both a
general and a specific neuronal inhibitory mechanism. Clinical
Neurophysiology, 122(6), 1156–1167.
Sheorajpanday, R. V., Nagels, G., Weeren, A. J., van Putten, M. J.,
& de Deyn, P. P. (2009). Reproducibility and clinical relevance
of quantitative EEG parameters in cerebral ischemia: A basic
approach. Clinical Neurophysiology, 120(5), 845–855.
Sheorajpanday, R. V., Nagels, G., Weeren, A. J., van Putten, M. J.,
& de Deyn, P. P. (2011). Quantitative EEG in ischemic stroke:
Correlation with functional status after 6 months. Clinical Neu-
rophysiology, 122(5), 874–883.
Tecchio, F., Pasqualetti, P., Zappasodi, F., Tombini, M., Lupoi, D.,
Vernieri, F., & Rossini, P. M. (2007). Outcome prediction in
acute monohemispheric stroke via magnetoencephalography.
Journal of Neurology, 254(3), 296–305.
Tecchio, F., Zappasodi, F., Tombini, M., Oliviero, A., Pasqualetti, P.,
Vernieri, F., Ercolani, M., Pizzella, V., & Rossini, P. M. (2006).
Brain plasticity in recovery from stroke: An MEG assessment.
NeuroImage, 32 (3), 1326–1334.
Tolonen, U., & Sulg, I. A. (1981). Comparison of quantitative EEG
parameters from four different analysis techniques in evalua-
tion of relationships between EEG and CBF in brain infarction.
Electroencephalography and Clinical Neurophysiology, 51(2),
177–185.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

83Appl Psychophysiol Biofeedback (2017) 42:69–83
1 3
Vespa, P. M., Nuwer, M. R., Juhasz, C., Alexander, M., Nenov, V.,
Martin, N., & Becker, D. P. (1997). Early detection of vasospasm
after acute subarachnoid hemorrhage using continuous EEG ICU
monitoring. Electroencephalography and Clinical Neurophysiol-
ogy, 103(6), 607–615.
Weber, E., Köberl, A., Frank, S., & Doppelmayr, M. (2011). Predict-
ing successful learning of SMR neurofeedback in healthy partici-
pants: Methodological considerations. Applied Psychophysiology
and Biofeedback, 36(1), 37–45.
Witte, M., Kober, S. E., Ninaus, M., Neuper, C., & Wood, G. (2013).
Control beliefs can predict the ability to up-regulate sensorimo-
tor rhythm during neurofeedback training. Frontiers in Human
Neuroscience, 7(478), 1–8.
Wood, G., Kober, S. E., Witte, M., & Neuper, C. (2014). On the need
to better specify the concept of “control” in brain–computer–
interfaces/neurofeedback research. Frontiers in Systems Neuro-
science. doi: 10.3389/fnsys.2014.00171
Zappasodi, F., Tombini, M., Milazzo, D., Rossini, P. M., & Tecchio,
F. (2007). Delta dipole density and strength in acute monohemi-
spheric stroke. Neuroscience Letters, 416(3), 310–314.
Zimmermann, P., & Fimm, B. (2007–2012). Testbatterie zur
Aufmerksamkeitsprüfung (TAP): Herzogenrath: Psytest.
Zoefel, B., Huster, R. J., & Herrmann, C. S. (2011). Neurofeedback
training of the upper alpha frequency band in EEG improves
cognitive performance. NeuroImage, 54(2), 1427–1431.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com