ArticlePDF Available

Abstract and Figures

Sensorimotor rhythm (SMR) activity has been related to automaticity during skilled action execution. However, few studies have bridged the causal link between SMR activity and sports performance. This study investigated the effect of SMR neurofeedback training (SMR NFT) on golf putting performance. We hypothesized that pre-elite golfers would exhibit enhanced putting performance after SMR NFT. Sixteen pre-elite golfers were recruited and randomly assigned into either an SMR or a control group. Participants were asked to perform putting while electroencephalogram (EEG) was recorded, both before and after intervention. Our results showed that the SMR group performed more accurately when putting and exhibited greater SMR power than the control group after 8 intervention sessions. This study concludes that SMR NFT is effective for increasing SMR during action preparation and for enhancing golf putting performance. Moreover, greater SMR activity might be an EEG signature of improved attention processing, which induces superior putting performance.
Content may be subject to copyright.
626
ORIGINAL RESEARCH
Journal of Sport & Exercise Psychology, 2015, 37, 626 -636
http://dx.doi.org/10.1123/jsep.2015-0166
© 2015 Human Kinetics, Inc.
The quality of mental regulation can differentiate
superior from inferior performance in precision sports
activities such as golf putting. In golf, the putt is consid-
ered one of the most important parts of the game, represent-
ing on average 43% of all shots taken during a single round
(Pelz & Frank, 2000). From a technical perspective, putting
is the simplest skill used in golf. However, mentally, put-
ting is the most stressful and demanding activity in the
game (Nicholls, 2007). The mental challenge of putting
is reected by previous psychophysiological studies
showing complex brain processes during putting perfor-
mance (Babiloni et al., 2008). Hence, the maintenance
of a mental state conducive to skilled execution is critical
for ideal precision sports performance.
Superior performance in precision sports can be
characterized as an automatic process as opposed to a
controlled process, which is typically observed in less
skilled performers (Fitts & Posner, 1967). An automatic
process is by nature reexive, whereas a controlled pro-
cess is an intentionally initiated sequence of cognitive
activity (Schneider & Shiffrin, 1977). Achieving auto-
matic process in action execution is the primary goal of
mastery (Logan, Hockley, & Lewandowsky, 1991). Dif-
ferences between these two levels of cognitive processing
are reected at the neurophysiological level: participants
who were in the automatic stage exhibited weaker activity
of the bilateral cerebellum, presupplementary motor area,
premotor cortex, parietal cortex, and prefrontal cortex
compared with novices (Wu, Chan, & Hallett, 2008). In
addition, the somatosensory cortex has been related to
conscious perception of somatosensory stimuli (Nierhaus
et al., 2015), such that lower activity in the somatosensory
cortex might be a signature of reduced conscious involve-
ment in movement execution, as is frequently observed
in highly skilled performers.
Although previous studies of the brain function
underlying superior golf putting performance have
provided insights into adaptive mental states and their
cortical processes, few studies have examined the
cortical processes that are more directly associated
with somatosensory activity. For example, Babiloni
Ming-Yang Cheng is with Center of Excellence “Cognitive
Interaction Technology” (CITEC), Bielefeld University, Biele-
feld, Germany. Chung-Ju Huang is with the Graduate Institute
of Sport Pedagogy, University of Taipei, Taipei City, Taiwan,
Republic of China. Yu-Kai Chang is with the Graduate Institute
of Athletics and Coaching Science, National Taiwan Sport
University, Taoyuan County, Taiwan, Republic of China. Dirk
Koester is with the Center of Excellence “Cognitive Interaction
Technology” (CITEC), Bielefeld University, Bielefeld, Ger-
many. Thomas Schack is with the Center of Excellence “Cog-
nitive Interaction Technology” (CITEC), Bielefeld University,
Bielefeld, Germany. Tsung-Min Hung is with the Department of
Physical Education, National Taiwan Normal University, Taipei
City, Taiwan, Republic of China. Address author correspon-
dence to Tsung-Min Hung at ernesthungkimo@yahoo.com.tw.
Sensorimotor Rhythm Neurofeedback
Enhances Golf Putting Performance
Ming-Yang Cheng,1 Chung-Ju Huang,2 Yu-Kai Chang,3
Dirk Koester,1 Thomas Schack,1 and Tsung-Min Hung4
1Bielefeld University; 2University of Taipei; 3National Taiwan Sport University;
4National Taiwan Normal University
Sensorimotor rhythm (SMR) activity has been related to automaticity during skilled action execution. However,
few studies have bridged the causal link between SMR activity and sports performance. This study investigated
the effect of SMR neurofeedback training (SMR NFT) on golf putting performance. We hypothesized that
preelite golfers would exhibit enhanced putting performance after SMR NFT. Sixteen preelite golfers were
recruited and randomly assigned into either an SMR or a control group. Participants were asked to perform
putting while electroencephalogram (EEG) was recorded, both before and after intervention. Our results
showed that the SMR group performed more accurately when putting and exhibited greater SMR power than
the control group after 8 intervention sessions. This study concludes that SMR NFT is effective for increasing
SMR during action preparation and for enhancing golf putting performance. Moreover, greater SMR activity
might be an EEG signature of improved attention processing, which induces superior putting performance.
Keywords: precision sports, attention, EEG, sensorimotor rhythm, automaticity
Neurofeedback Training and Golf Putting Performance 627
JSEP Vol. 37, No. 6, 2015
et al. (2008) demonstrated that successful putting was
preceded by higher high-frequency alpha (10–12 Hz)
event-related desynchronization over the frontal midline
and the right primary sensorimotor area compared with
unsuccessful putting performance. Similarly, studies
found that reduced (Kao, Huang, & Hung, 2013) and
stable (Chuang, Huang, & Hung, 2013) frontal midline
theta power was the precursor of superior performance
in precision sports. Since high-frequency alpha power
in these cortical areas reect only task-related attention
(Klimesch, Doppelmayr, Pachinger, & Ripper, 1997)
whereas frontal midline theta power indicates top-down
sustained attention (Sauseng, Hoppe, Klimesch, Gerloff,
& Hummel, 2007), these ndings support the importance
of specialized task-related attention on superior motor
performance. However, the information encoded during
automatic somatosensory processing during skilled
precision sport performance remains unexamined as yet.
Sensorimotor rhythm (SMR), the 12- to 15-Hz oscil-
lation of the sensorimotor cortex, has shown promising
as a link between adaptive mental states (e.g., automatic
process-related attention) and skilled visuomotor per-
formance. Sensorimotor rhythm is considered an indi-
cator of cortical activation, which is inversely related to
somatosensory processing (Mann, Sterman, & Kaiser,
1996). A recent study showed that skilled dart-throwing
players demonstrated higher SMR power before dart
release than novices in a dart-throwing task (Cheng
et al., 2015). This result suggests that lower cognitive
involvement in processing somatosensory information
as reected by higher SMR power is characteristic of
skilled performance. Furthermore, several lines of studies
pertaining to SMR power tuning for enhancing adaptive
cortical processing in motor performance have shown
promising results. Augmented SMR power resulting from
neurofeedback training (NFT) has been identied as a
relaxed focus state without somatosensory intervention
(Gruzelier, Foks, Steffert, Chen, & Ros, 2014). Similarly,
a reduced trait anxiety score and task-processing time
during microsurgery were observed after augmented
SMR NFT (Ros et al., 2009). Moreover, a facilitative
sense of control, condence, and feeling at-one with a
role was demonstrated after augmented SMR NFT before
acting performance (Gruzelier, Inoue, Smart, Steed, &
Steffert, 2010). Thus, increased SMR activity implies
the maintenance of a relaxed, focused state by reducing
motor perception (e.g., somatosensory processing) by
the sensorimotor cortex (Vernon et al., 2003). This inter-
pretation is similar to the mental characteristics of peak
performance in skilled athletes (Krane & Williams, 2006)
and is in agreement with the concept of automaticity
proposed by Fitts and Posner (1967). Hence, SMR power
not only might be a sensitive indicator of the activity of
sensorimotor cortex (Mann et al., 1996) but also shows
potential for a performance-enhancing intervention.
Although there is no direct evidence to support the
effectiveness of SMR NFT on performance enhancement
in precision sport, two lines of research lend support to
its potential use in sports. First, previous studies have
demonstrated the effectiveness of NFT on performance
enhancement in precision sports. For example, Land-
ers et al. (1991) demonstrated that correct NFT (i.e.,
augmented slow cortical potential at the left temporal
lobe) led to superior performance, whereas incorrect
NFT (i.e., augmented slow cortical potential at the right
temporal lobe) impaired performance in skilled archers.
Similarly, Kao, Huang, and Hung (2014) reported that
NFT targeting to reduce the frontal midline theta resulted
in improved performance in skilled golfers. These nd-
ings support the feasibility of tuning EEG to improve
behavioral outcome in precision sports. The second line
of evidence is the nding that SMR NFT has a benecial
effect on attention-related performance in various atten-
tional tasks. For example, an increased P300b amplitude
at frontal, central, and parietal sites during the auditory
oddball task and reduced commission errors, and a
reduction in reaction time variability during the Test
of Variables of Attention (TOVA) was observed after
augmented SMR NFT (Egner, Zech, & Gruzelier, 2004).
These ndings suggest that augmenting SMR power
might improve attention-related processes by improving
impulse control and the ability to integrate relevant envi-
ronmental stimuli. Similarly, Ros et al. (2009) reported
that a shorter operation time and reduced trait anxiety
score were observed in surgeons following augmented
SMR NFT, suggesting that augmented SMR enhanced
the learning of a complex medical specialty by develop-
ing sustained attention and a relaxed attentional focus
as well as increasing working memory (Vernon et al.,
2003). Furthermore, Doppelmayr and Weber (Doppel-
mayr & Weber, 2011) revealed that augmented SMR
NFT not only resulted in a signicant SMR amplitude
increase accompanied by a signicant increase in reward
threshold, but also facilitated the performance of spatial-
rotation, simple, and choice-reaction time tasks. These
results indicate that visuospatial processing, semantic
memory regulation, and the integration of relevant stimuli
can be improved following augmented SMR NFT. Col-
lectively, the benets of augmented SMR NFT can be
attributed to an improved regulation of somatosensory
and sensorimotor pathways, which in turn leads to more
efcient attention allocation (Kober et al., 2014) that
results in an improved processing of task-relevant stimuli.
To the best of our knowledge, no study has directly
examined the effect of SMR NFT on precision sport
performance. Thus, this study investigated the effect of
SMR NFT on a golf putting task. We predicted that golf-
ers would be able to increase SMR power before putting
execution following augmented SMR NFT. More impor-
tantly, we predicted that increased SMR power improves
putting performance as a result of augmented SMR NFT.
Methods
Participants
Fourteen male and two female preelite and elite golf-
ers were recruited (mean handicap = 0, SD = 3.90).
628 Cheng et al.
JSEP Vol. 37, No. 6, 2015
Participants were matched based on performance history
supplemented by the assessment of a professional coach and
then randomly assigned into either an SMR neurofeedback
group (SMR NFT) or a control group (seven male and one
female for each group). The mean age of the SMR NFT and
control group were 20.6 (1.59) and 22.3 (2.07), respectively.
The years of experience in golf were 9.5 (2.67) for the SMR
NFT group and 9.2 (1.83) for the control group. An inde-
pendent t test showed no difference in age [t(14) = 1.895,
p = .079] or years of experience in golf [t(14) = 0.273, p
= .789] between the two groups. None of the participants
reported psychiatric and neurological disorders and had
never been hospitalized for general brain damage.
Procedures
For the pretest and posttest, we used the same procedure
to collect data. At pretest, after being informed of the
general purpose of the study, all participants were asked
to read and sign an informed consent form approved by
our institutional review board. They were then given the
opportunity to ask questions about the experiment. The
participants were individually tested in a sound-proof indoor
articial golf green, where they were initially required to
stand 3 m from a hole 10.8 cm in diameter to obtain an
individual putting distance (Arns, Kleinnijenhuis, Fallah-
pour, & Breteler, 2008). Participants performed a series of
10 putts, which were scored as successfully holed or not
holed. The percentage of successful putts in a series was
determined after each series. This process was repeated
until each participant achieved 50% accuracy.
After the individual putting distance was determined,
participants were tted with a Lycra electrode cap (Neu-
roscan, Charlotte, NC, USA). After a 10-min warm-up,
participants were rst asked to undergo a resting EEG
recording, including eye-closed and eye-opened condi-
tions, while assuming a normal putting stance for 1 min
each. Then, all participants performed golf putting tasks
consisting of 40 self-paced putting trials in four separate
recording blocks while EEGs were recorded. The partici-
pants performed the putting task in the standing position and
were allowed to take a brief rest between each putt. They
were also allowed to sit briey after each block of 10 putts.
The score was calculated based on the linear distance
from the edge of hole to the edge of the ball (cm). Put-
ting into the hole successfully was determined as score 0.
Putting trials in which the ball was deected by contact-
ing the edge of the hole were excluded, and participants
were asked to perform extra putting trials to complete the
forty trials. The experiment lasted approximately 2 hr in
total. After completing the pretest, all participants were
scheduled to go through 8 sessions of neurofeedback
training. Then the posttest, which was identical to the
pretest, followed the neurofeedback intervention.
Instrumentation
Electroencephalography. For the pretest and posttest,
EEGs were recorded at 32 electrode sites (FP1, FP2, F7,
F8, F3, F4, FZ, FT7, FT8, FC3, FC4, C3, C4, CZ, T3, T4,
T5, T6, TP7, TP8, CP3, CP4, CPZ, A1, A2, P3, P4, PZ,
O1, O2, OZ) corresponding to the International 10–10
system (Chatrian, Lettich, & Nelson, 1985). In addition,
four electrodes were attached to acquire horizontal and
vertical oculography (HEOL, HEOR, VEOU and VEOL).
All sites were initially referenced to A1 and then rerefer-
enced to linked ears ofine. A frontal midline site (FPz)
served as the ground. EEG data were collected and ampli-
ed using a Neuroscan Nuamps amplier (Neuroscan,
Charlotte, NC, USA) with a band-pass lter setting of
1–100 Hz and a 60-Hz notch lter. The EEG and EOG
signals were sampled at 500 Hz and recorded online with
NeuroScan 4.5 (Neuroscan, Charlotte, NC, USA) soft-
ware installed on a Lenovo R400 laptop (Lenovo, Taipei
City, R.O.C). Vertical and horizontal eye movement
artifacts were recorded via bipolar electro-oculographic
activity (EOG), in which vertical EOG was assessed by
electrodes placed above and below the left eye (VEOU
and VEOL), whereas horizontal EOG was assessed by
electrodes located at the outer canthi (HEOL, HEOR).
Impedance values for all electrode sites were maintained
below 5 kΩ. An infrared ray sensor was set to detect the
swing for each putt. Once the back swing movement was
detected, an event mark was sent to the EEG data, which
served as the time point for analyzing the EEG activity
before putting. Twelve to fteen hertz of Cz was extracted
as the SMR (Babiloni et al., 2008).
Neurofeedback. Neurofeedback training was com-
pleted with a NeuroTek Peak Achievement Trainer (Neu-
roTek, Goshen, KY). The EEG data from the assessment
were band-pass ltered using the BioReview software
(NeuroTek, Goshen, KY). The active scalp electrode was
placed at Cz for SMR training, with the reference placed
on both mastoids. Signal was acquired at 256 Hz and
then A/D converted and band ltered to extract the SMR
(12–15 Hz). The amplitude of the SMR was transformed
online into graphical feedback representations including
the low-frequency audio-feedback tone by acoustic bass
(No. 33) in the BioReview software.
Neurofeedback Training Procedure
Participants underwent an eight-session training program
lasting 5 weeks. Each session was composed of neuro-
feedback training lasting from 30 to 45 min. On average,
a total of 12 training trials were performed in a single ses-
sion. Each training trial comprised 30 s. The total duration
of a single session was approximately 30 min. The SMR
NFT group aimed to increase absolute SMR amplitude
over the designated threshold, which was individually
determined by averaging 1.5 s of each participant’s suc-
cessful putting trials during the pretest. To enhance the
participants’ efcacy during NFT, a progressive adjust-
ment of the training threshold difculty was employed.
The standard for adjusting the training threshold was
based on the individualized standard deviation which
derived from the SMR power of the nal three 0.5-s
time windows before putting during the pretest. When
Neurofeedback Training and Golf Putting Performance 629
JSEP Vol. 37, No. 6, 2015
participants’ SMR power was higher than the threshold,
the acoustic bass sound was displayed. Participants were
instructed to perform based on their own putting routine
while receiving the auditory feedback. The successful
training ratio, dened as the time spent above threshold
divided by the total time of a single training trial (30 s),
was reported to participants following every training trial.
In the control group, the training protocol was similar
to that used by Egner, Strawson, and Gruzelier (2002)
to establish a mock feedback condition. This protocol
was designed to prevent study participants from learn-
ing to regulate SMR by using the randomly prerecorded
feedback tone during the training trials from SMR NFT
group. The total length of this prerecorded mock feedback
tone was 4 min that were derived from a randomly chosen
participant in the SMR NFT group during the Session
1 training. Researchers played the mock feedback tone
from a random starting point to guarantee a randomized
feedback tone was received by participants in the control
group. On average, a total of seven training trials were
performed in a single session and the total duration of a
single session was approximately 30 min.
To evaluate the neurofeedback learning effect,
the mean successful training ratio of each session was
recorded and computed for subsequent analysis. To
reduce the number of sessions necessary for statistical
evaluation of the learning efciency between the two
groups, we combined two consecutive sessions into one
section [e.g., Section 1 = (Session 1 + Session 2) / 2].
Data Reduction
The EEG data reduction was conducted ofine using the
Scan 4.5 software (Neuroscan, Charlotte, NC, USA).
EEG data were sampled 1.5 s before putting execution
and were triggered by the event-related marker from
infrared ray sensors. Trial preparation periods of less than
1.5 s were excluded to establish the common structure of
artifact-free data across trials and participants. EOG cor-
rection (Semlitsch, Anderer, Schuster, & Presslich, 1986)
was carried out on continuous EEG data to eliminate blink
artifacts. EEG segments with amplitudes exceeding ±100
μV from baseline were excluded from subsequent analysis.
After artifact-free EEG data were acquired, fast Fourier
transforms were calculated at 50% overlap on 256-sample
Hanning windows for all artifact-free segments to transform
to spectral power (μV2). Sensorimotor rhythm power was
computed as the mean of 12–15 Hz from Cz and then natura l
log transformed (Davidson, 1988). To compute a normal-
ized EEG power for each golfer, the relative power was
used, for which the ratio of power at 12–15 Hz to 1–30
Hz was computed (Niemarkt et al., 2011).
Statistical Analyses
The average putting score and standard deviation between
the two groups was analyzed by a 2 (Group: SMR
NFT, Control) × 2 (Test: pretest, posttest) ANOVA with
repeated measures on the test factor.
The difference score (posttest to pretest) for the
relative power of SMR was subjected to a 2 (Group:
SMR NFT, Control) × 3 [Time window: –1.5 to –1.0 s
(T1), –1.0 to –0.5 s (T2), –0.5 to 0 s (T3)] ANOVA with
repeated measures on the time window factor.
In addition, we ran several control analyses to
provide additional evidence to support our conclusions.
The success of the training ratio was tested by a 2
(Group: SMR NFT, Control) × 4 (Training section: Sec-
tion 1: sessions 1–2; Section 2: sessions 3–4; Section
3: sessions 5–6; Section 4: sessions 7–8) ANOVA with
repeated measures on the training section.
To characterize the within-session learning effect,
we compared the successful training ratio of the rst and
last trials of each session across all eight sessions. A 2
(Group: SMR NFT, Control) × 8 (Session: session 1, 2,
3, 4, 5, 6, 7, 8) × 2 (Trial: rst trial, last trial) three-way
ANOVA with repeated measures on the session, and trial
was used to examine this issue.
To ensure control of neurofeedback in the SMR
NFT group within the training program, we employed a
one-way ANOVA with training section (Training section:
Section 1: sessions 1–2; Section 2: sessions 3–4; Section
3: sessions 5–6; Section 4: sessions 7–8) as a variable to
detect the threshold uctuation within the four training
sections.
To examine the regional uctuation of 12–15 Hz
power before and after training, we carried out a 2 (Group:
SMR NFT, Control) × 4 (Region: frontal, central, parietal,
occipital) two-way ANOVA with repeated measures on
the region.
The examination of concurrent changes in neigh-
boring frequency bands was conducted by analyzing the
pre-to-post difference scores for theta (4–7 Hz), alpha
(8–12 Hz), low beta (13–20 Hz), high beta (21–30 Hz),
and broad beta (13–30 Hz) frequency bands with a 2
(Group: SMR NFT, Control) × 3 [Time window: –1.5 to
–1.0 s (T1), –1.0 to –0.5 s (T2), –0.5 to 0 s (T3)] two-
way ANOVA.
Mauchly’s test was used to assess the validity of
the ANOVA sphericity assumption whenever neces-
sary. The degrees of freedom were corrected using the
Greenhouse–Geisser procedure, and least signicant
difference analysis was used for post hoc comparisons
(p < .05). The partial eta square was used to estimate the
effect size, with values of .02, .12, and .26 suggesting
relatively small, medium, and large effect sizes, respec-
tively (Cohen, 1992).
Results
Putting Performance
The mean distance of the SMR group in the pretest
and posttest was 29.62 cm (5.59) and 16.59 cm (8.92),
respectively. The control group distance was 20.17 cm
(12.07) and 18.80 cm (5.58), respectively. An indepen-
dent t test showed no difference in the mean distance in
the pretest between two groups [t(14) = 2.008, p = .073,
630 Cheng et al.
JSEP Vol. 37, No. 6, 2015
η2p = .224]. The 2 (Group: SMR NFT, Control) × 2
(Test: pretest, posttest) mixed-model ANOVA revealed
a signicant interaction effect on putting performance
[F(1, 14) = 5.029, p = .042, η2p = .264]. The SMR neu-
rofeedback group exhibited a shorter distance from the
hole in posttest than pretest [t(7) = 3.417, p = .011, η2p
= .625]. No signicant difference was observed for other
comparisons.
Putting Performance
in Standard Deviation
A marginal interaction effect was observed in the 2
(Group: SMR NFT, Control) × 2 (Test: pretest, posttest)
ANOVA [F(1, 14) = 4.121, p = .062, η2p = .227]. We
did not observe an effect on Group factor [F(1, 14) =
0.136, p = .717, η2p = .010]. The SMR group exhibited
a signicantly lower SD in the posttest (16.11 cm) than
in the pretest (24.70 cm) [t(7) = 4.408, p = .003, η2p =
.735], whereas the control group showed no signicant
variation in SD (21.03 cm to 18.38 cm) [t(7) = 1.208, p
= .266, η2p = .173].
SMR Relative Power
The difference scores of the SMR group members for T1,
T2, and T3 was 0.481 (0.588), 0.186 (0.378), and 0.040
(0.268), respectively. For the control group, the difference
scores was –0.200 (0.424), –0.143 (0.440), and 0.009
(0.444), respectively. We compared the difference scores
with a 2 (Group: SMR NFT, Control) × 3 [Time window:
–1.5 to –1.0 s (T1), –1.0 to –0.5 s (T2), –0.5 to 0 s (T3)]
two-way ANOVA and observed a marginally signicant
two-way interaction effect [F(2, 28) = 3.315, p = .051,
η2p = .191]. To explore this marginal interaction effect
and examine the training effect before and after NFT, a
subsequent simple main effect analysis was performed
and revealed a marginal Time effect [F(2, 14) = 3.470,
p = .060, η2p = .331] in the SMR NFT group. Post hoc
analysis showed that the SMR power was signicantly
greater in T1 than in T3 [t(7) = 2.925, p = .022, η2p =
.550]. No signicant simple main effect was observed in
the control group [F(2, 14) = .671, p = .567, η2 = .141]. In
addition, a simple main effect analysis revealed that the
SMR NFT group exhibited a relatively higher SMR power
than that of the control at T1 [t(14) = 2.657, p = .019, η2p
= 335]. The signicant group main effect revealed that
the SMR NFT group had a higher SMR power than that
of the control group [F(1, 14) = 4.665, p = .049, η2p =
.250]. The difference scores between the two groups are
depicted in Figure 1.
Control Analyses
Successful Training Ratio. The overall mean of the
golfers’ successful training ratio was 62.39 (8.88) %
for the SMR training group and 22.27 (22.28) % for the
control group. The 2 (Group: SMR NFT, Control) × 4
Figure 1 — The difference scores of SMR relative power between the SMR NFT and control groups at T1 (–1.5 to –1.0 s), T2
(–1.0 to –0.5 s), and T3 (–0.5 to 0 s).
Neurofeedback Training and Golf Putting Performance 631
JSEP Vol. 37, No. 6, 2015
(Training section: Section 1: sessions 1–2; Section 2:
sessions 3–4; Section 3: sessions 5–6; Section 4: sessions
7–8) ANOVA showed no interaction effect [F(3,42) =
0.694, p = .497, η2p = .047], but a signicant group main
effect was observed [F(1,14) = 22.188, p = .001, η2p =
.613]. The SMR group showed a signicantly higher per-
centage of successful training ratios than did the control
group. Table 1 lists the successful training ratio for each
group during the training sections.
Within-Session Learning.
The results of NFT can be affected by day-to-day uc-
tuations in arousal level (Gruzelier et al., 2014). Thus,
in addition to comparing the average successful training
ratios of the eight sessions between these two groups, we
compared the successful training ratios of the rst and
last trials of each session for all eight sessions between
the two groups to determine whether participants in the
NFT group improved within each training session. We
hypothesized that the successful training ratio would be
greater in the last trial than in the rst trial for the SMR
NFT group but not for control group. A 2 (Group: SMR
NFT, Control) × 8 (Session: sessions 1, 2, 3, 4, 5, 6, 7,
8) × 2 (Trial: rst trial, last trial) three-way ANOVA
was employed to test this hypothesis. The result
showed that although the 3-way interaction effect was
not signicant [F(7, 98) = 2.063, p = .082, η2p = .128], a
2 (Group: SMR NFT, Control) × 2 (Trial: rst trial, last
trial) interaction effect [F(1, 14) = 33.192, p = .001, η2p
= .703] was revealed. Post hoc analysis was consistent with
our prediction; only the SMR NFT group demonstrated
a greater successful training ratio in the last trial (M =
77.65, SD = 7.84) than in the rst trial (M = 50.58, SD =
10.65) for all sessions [t(7)= 8.344, p = .001, η2p = 909].
The control group did not show a signicant difference
between the rst trial (M = 12.19, SD = 11.86) and last
trial (M = 16.32, SD = 17.00) [t(7) = 1.784, p = .118, η2p
= 313]. In addition, the SMR NFT group demonstrated a
signicantly higher training ratio on the rst trial [t(7) =
6.810, p = .001, η2p = 768] and last trial [t(7) = 9.267, p
= .001, η2p = .860] than did the control group (Figure 2).
Threshold Increments Within SMR Training Sessions.
Although our control analyses provided supportive evi-
dence for the learning progress made by the SMR NFT
group, we further analyzed the change in threshold during
each session of SMR NFT. In our study, threshold level
was used as a difculty index in the SMR NFT group, in
which golfers were instructed to increase the SMR above
designated level to meet our training demand. Thus, an
improvement in the successful training ratio from the
two previous control analysis was meaningful only
when the threshold for each session was also examined.
Previous studies evaluated the threshold variation within
day-to-day sessions and suggested that the increased
threshold could serve as a marker for improvement of
the controllability due to neurofeedback training (Dop-
pelmayr & Weber, 2011). Thus, we converted the eight
training sessions into four sections as described in the
methods section and examined the training threshold
variation by employing an one-way ANOVA to examine
the effect of Training section (Section 1: sessions 1–2;
Section 2: sessions 3–4; Section 3: sessions 5–6; Section
4: sessions 7–8) in the SMR group. We hypothesized that
the threshold value would increase after the rst training
section, which supports an improvement in controllability
due to SMR neurofeedback training. The average train-
ing thresholds for sections one to four in the SMR NFT
group were 5.862 (2.781), 7.636 (3.368), 8.214 (3.718),
and 7.750 (3.816), respectively. As predicted, a signicant
difference was detected by the one-way ANOVA [F(3,
18) = 9.945, p = .001, η2p = .624]. Post hoc analysis
demonstrated that the training thresholds in the second,
third, and fourth sections were signicantly higher than
that of the rst section.
Electrode Specificity. Although the current study dem-
onstrated that the relative SMR power of the SMR NFT
group was signicantly higher than that of the control
group following SMR NFT, it remained unknown whether
the greater 12–15 Hz EEG relative power after training
was limited to the sensorimotor cortex or there was a
spillover to other regions, such as the frontal, parietal
and occipital cortices. Thus, we compared the difference
scores at 12–15 Hz EEG relative power among Fz, Cz, Pz,
and Oz between pre- and posttest sessions. Previous work
has shown that the SMR originated in the centro-parietal
region (Grosse-Wentrup, Schölkopf, & Hill, 2011). Thus,
we hypothesized that the difference score of 12–15 Hz at
Cz would be greater than that of the frontal and occipital
regions for SMR group participants after training. A 2
(Group: SMR NFT, Control) × 4 (Region: Frontal, Cen-
tral, Parietal, Occipital) two-way ANOVA between the
two groups was performed to test this hypothesis.
The difference scores at Fz, Cz, Pz, and Oz were
0.035 (0.200), 0.212 (0.178), 0.135 (0.298), and 0.003
(0.241), respectively, for the SMR NFT group. For the
Table 1 The Successful Training Ratios Between the SMR NFT
and Control Groups Across the Four Training Sections
(Every Two Consecutive Sessions Were Folded Resulting In Four Sections)
Section 1 Section 2 Section 3 Section 4 Total
SMR 53.82 (19.71) 63.85 (12.53) 65.63 (9.52) 66.27 (17.91) 62.39 (5.08)
Control 20.51 (24.11) 23.02 (26.31) 22.94 (21.58) 22.62 (19.61) 22.27 (1.09)
Note. The unit is the percentage of increasing time for successfully controlling SMR power.
632 Cheng et al.
JSEP Vol. 37, No. 6, 2015
control group, the difference scores at Fz, Cz, Pz, and Oz
were –0.056 (0.309), –0.438 (0.169), –0.150 (0.268),
and –0.168 (0.640), respectively. This result yielded
a marginally signicant interaction effect [F(3, 42) =
2.680, p = .089, η2p = .161]. Because of the exploratory
nature of this study, we conducted a follow-up analysis
of this interaction effect. The independent t tests of
the four regions between the two groups showed that
signicance was only observed at a difference score
of Cz [t(14) = 5.159, p = .001, η2p = 655], in which
the SMR NFT group exhibited a signicantly higher
difference score than the control group. Moreover, one-
way ANOVA of four regions in the SMR NFT group
reached marginal signicance [F(3, 21) = 2.644, p =
.076, η2p = .274]. The follow-up pairwise t tests found
that the difference score of Cz was higher than that of Fz
[t(7) = 3.740, p = .007, η2p = 666] and Oz [t(7) = 2.530,
p = .039, η2p = .478]. These lines of evidence provide
preliminary support for the electrode specicity of SMR
NFT in this study.
Frequency Specificity. Previous studies have shown
that neurofeedback training may generate concurrent
changes in anking frequency bands (Enriquez-Geppert
et al., 2014). The aim of this analysis was to investigate
whether SMR NFT resulted in a change in frequency
bands close to SMR. We compared the relative power
difference scores of theta (4–7 Hz), alpha (8–12 Hz),
low beta (13–20 Hz), high beta (21–30 Hz), and broad
beta (13–30 Hz) frequency bands before golf putting
from pretest and posttest between the two groups. The 2
(Group: SMR NFT, Control) × 3 [Time window: –1.5 to
–1.0 s (T1), –1.0 to –0.5 s (T2), –0.5 to 0 s (T3)] two-
way ANOVA showed that neither interaction effects on
theta power [F(2, 28) = 0.550, p = .583, η2p = .038],
alpha power [F(2, 28) = 0.113, p = .802, η2p = .011],
low beta power [F(2, 28) = 0.052, p = .949, η2p = .004],
high beta power [F(2, 28) = 0.503, p = .496, η2p = .035],
and broad beta band [F(2, 28) = 0.883, p = .425, η2p =
.059] nor group main effects on theta power [F(1, 14) =
0.032, p = .860, η2p = .002], alpha power [F(1, 14) =
0.070, p = .795, η2p = .005], low beta power [F(1, 14)
= 0.764, p = .397, η2p = .052], high beta power [F(1,
14) = 0.677, p = .424, η2p = .046], and broad beta power
[F(1, 14) = 0.023, p = .881, η2p = .002] were observed.
The difference scores among these ve frequency bands
are listed in Table 2.
Discussion
The aim of this study was to investigate the effect of SMR
neurofeedback training on golf putting performance. Our
results showed that golfers receiving SMR neurofeedback
training demonstrated enhanced SMR activity during the
nal 1.5 s before golf putting, resulting in better putting
performance compared with the control group. This nd-
ing lends preliminary support to the hypothesis that SMR
NFT is effective for increasing SMR power, and leads to
superior putting performance.
Figure 2 — The mean successful training ratio for the rst and last trial between the SMR NFT and control groups across the
eight training sessions.
Neurofeedback Training and Golf Putting Performance 633
JSEP Vol. 37, No. 6, 2015
Increased SMR power by NFT results in better visuo-
motor performance. For behavioral data, we observed that
SMR neurofeedback training improved skilled golfers’
putting performance, as indicated by the reduced average
distance from the hole and the variability of the score. No
signicant change in putting performance was observed
in the control group. Previous studies have demonstrated
that augmenting SMR by NFT improved visual motor
performance (Ros et al., 2009) and increased self-rating
scores of subjective ow state in dancers (Gruzelier
et al., 2010). Furthermore, augmenting SMR by NFT
was related to an improved attention-related mental
state (Vernon et al., 2003) and memory performance
(Hoedlmoser et al., 2008). In addition, converging lines
of evidence support the effectiveness of NFT based
on non-SMR variables enhancing performance in the
sport domain (Arns et al., 2008; Gruzelier et al., 2010;
Kao et al., 2014; Landers et al., 1991; Raymond, Sajid,
Parkinson, & Gruzelier, 2005; Ring, Cooke, Kavussanu,
McIntyre, & Masters, 2015). Nevertheless, the current
study is the rst, to our best knowledge, to use the SMR
protocol to investigate the effectiveness of NFT on sport
performance. Our results support the nding of the aug-
mented SMR power which is linked with more adaptive
ne-motor performance (Cheng et al., 2015) and extend
the potential facilitation effects of SMR training to the
sport domain.
Less task-irrelevant interference of somatosensory
and sensorimotor processing, as reected in augmented
SMR power after training, leads to improved putting
performance. A previous study has indicated that par-
ticipants in the automatic stage showed weaker activity
in the presupplementary motor area, premotor cortex,
parietal cortex, and prefrontal cortex compared with
novices in a self-paced sequential nger movement task
(Wu et al., 2008). A negative relationship between SMR
power and sensorimotor activity has been suggested
(Mann et al., 1996). The drop in sensorimotor activity,
as reected by increased SMR power, may indicate a
greater adaptive task-related attention allocation that
facilitates the execution of sport performance (Gruzelier
et al., 2010). Increasing SMR power through NFT is
also related to more efcient and modulated visuomotor
performance (Gruzelier et al., 2010; Ros et al., 2009).
These results suggest that augmenting SMR power led
to an improved adjustment of somatosensory and sen-
sorimotor pathways (Kober et al., 2014), which resulted
in increased task-related attention toward specic tasks
(Egner & Gruzelier, 2001). Moreover, previous studies
have suggested that enhanced SMR power leads to a
relatively higher ow state (Gruzelier et al., 2010) and
calming mood (Gruzelier, 2014a). Based on the functional
role of SMR, these ndings imply that a reduction in
sensorimotor activity may lessen the conscious process-
ing involved in motor execution, which would lead to a
more conceptual automatic process (Cheng et al., 2015).
This interpretation is in line with converging evidence
supporting a benecial effect of augmented SMR on
focusing and sustaining attention, working memory, and
psychomotor skills (Egner & Gruzelier, 2001; Ros et al.,
2009). Collectively, the superior golf putting performance
observed in the present SMR NFT group might be the
result of reduced somatosensory information processing
before the back swing, which leads to rened golf put-
ting performance. The interpretation that a reduction in
conscious interference facilitates motor operation is in
line with the concept of automatic processing proposed
by Fitts and Posner (1967). However, given the relatively
small sample size, future research should verify the causal
relationship between augmented SMR power and ne-
motor performance.
Reduced cortical activity in the sensorimotor area, as
reected by the higher power of 12–15 Hz, is sensitive to
superior putting performance. First, the electrode speci-
city of SMR NFT was demonstrated. Although electrode
specicity has been suggested to be an important step in
support of the NFT training effect on the corresponding
EEG component at a specic brain region (Gruzelier,
2014b), this is the rst study in the area of NFT and
sport performance to provide such preliminary evidence
for the localized training effects. The lack of difference
between Cz and Pz might suggest that this region is
also part of a network associated with SMR activity in
motor performance. This speculation is in line with the
evidence that the parietal region is involved in processing
visual-spatial information during motor performance (Del
Percio et al., 2011).
Second, frequency specicity was analyzed. One
might argue that enhanced putting performance was
caused by variation in another frequency band at the Cz
Table 2 Difference Scores (%) of Relative Power for Theta, Alpha, Low Beta, High Beta,
and Beta Frequency Bands in Three Time Windows Between the Two Groups, SMR and Control
T1 (–1.5 to –1.0 s) T2 (–1.0 to –0.5 s) T3 (–0.5 to 0 s)
Relative Power SMR Control SMR Control SMR Control
Theta .025 (.621) .338 (.493) –.234 (.172) –.186 (.528) .311 (1.071) .085 (.452)
Alpha .006 (.134) .048 (.221) .052 (.177) .017 (.216) –.006 (.465) –.058 (.223)
Low beta .035 (.258) .014 (.135) –.069 (.124) –.029 (.164) –.097 (.183) –.033 (.082)
High beta .014 (.190) .015 (.109) –.046 (.085) .128 (.593) –.047 (.152) –.030 (.070)
Beta .034 (.164) .053 (.010) –.064 (.094) –.029 (.077) –.050 (.137) –.082 (.112)
634 Cheng et al.
JSEP Vol. 37, No. 6, 2015
site, but this explanation is inconsistent with the lack
of signicant changes on difference scores in the theta,
alpha, low beta and high beta frequency bands. These
results suggest that it is primarily SMR power that
accounts for the facilitating effect of SMR NFT on put-
ting performance rather than other neighboring frequency
bands. Our demonstration of electrode and frequency
specicity strengthens the hypothesis that improved put-
ting performance was the result of reduced sensorimotor
activity before putting execution.
The SMR NFT group improved the putting perfor-
mance through the rened strategy for controlling the
SMR power and reached the training goal as a result
of the training program. First, our data showed that the
SMR group demonstrated a higher successful training
ratio than did the control group. Second, previous studies
proposed that the training effect would emphasize daily
training improvement (Gruzelier et al., 2014). In our
control analysis, we compared the successful training
ratio of the rst and the last trial within eight sessions. A
signicantly higher successful training ratio for the last
trial than for the rst trial was observed, suggesting that
golfers in the SMR NFT group learned the tuning strategy
successfully after the initial trials and that the strategies
were effective in the subsequent trials of the remaining
sessions. This result lends support to the concept of neuro-
feedback trainability and further conrms the possibility
of EEG tuning within a single training session (Kao et
al., 2014; López-Larraz, Escolano, & Minguez, 2012).
Furthermore, we found a signicant threshold increase
after the rst session only in SMR NFT group, suggest-
ing that our training protocol is facilitative to golfers.
This evidence was in line with previous work in which
the SMR amplitude increased above the daily adjusted
threshold (Weber, Köberl, Frank, & Doppelmayr, 2011).
We have several suggestions with regard to future
neurofeedback studies. First, combining these studies
with neuroimaging tools is necessary. Although we have
provided evidence that the regulation of SMR power
can enhance putting performance, this result would be
benet from the experiments conducted with high-spatial-
resolution neuroimaging tools, such as fMRI, to provide
a more precise anatomical description of the NFT effect.
Second, the phenomenological report of neurofeedback
learning and its effects is often overlooked (Gruzelier,
2014b). A sophisticated measurement of subjective
mental state, such as an in-depth questionnaire or scale,
is needed to further elucidate the mental state associ-
ated with NFT (Gruzelier, 2014a). Third, the retention
of learning driven by NFT must be examined. Thus far,
this issue has received little attention, but it is critical
from a practical viewpoint to determine how long the
performance enhancement due to NFT lasts. Fourth, to
explore the effect of SMR NFT on anticipative motor
planning is needed. Future study should investigate the
link between neurophysiological and cognitive processes
by using the priming tests to further understand the neu-
rocognitive architecture of golf performance. Last but
not least, the changes in network dynamics after NFT
should be further examined to ll the knowledge gap
of cortical interaction caused by NFT. For example, the
parietal and sensorimotor cortex networks are thought
to be functionally relevant during motor performance
(Baumeister et al., 2013).
Our ndings should be interpreted with caution due
to the limitations of the study. First, the sample size was
limited. Some of our statistical analyses reached only
marginal signicance, likely due to the small sample size.
Furthermore, given the exploratory nature of the study,
it is reasonable to speculate implications regarding the
the marginally signicant effects. Second, although the
neurophysiological source of the SMR could not be pre-
cisely located due to limited spatial resolution by surface
EEG, the nding of a marginally signicant larger SMR
difference score at the Cz site compared with the Fz and
Oz sites as well as the nding that the largest magnitude
of 12–15 Hz differences occurred at the Cz site rather
than other frequency bands in the SMR group provide
indirect evidence to support the impact of somatosensory
activity on superior putting performance after SMR NFT.
Third, putting is only one of many fundamental motor
skills involved in golf performance. Our results may be
difcult to generalize to other golf motor skills (e.g., the
drive shot and tee shot). Future studies should, therefore,
examine different skills involved in golf performance to
determine the generalizability of the present ndings.
Fourth, the skill levels of the participants may impact
the effect of NFT, and caution should be exercised when
generalizing these ndings to golfers at other skill levels.
In conclusion, an eight-session SMR NFT exhib-
ited a putting performance enhancement and increased
SMR power in SMR NFT group compared with control
group, suggesting that SMR NFT is an effective protocol
for enhancing putting performance through ne-tuning
somatosensory interference, as reected by augmented
SMR.
Acknowledgments
The work of Tsung-Min Hung was supported in part by the
National Science Council (Taiwan) under grant NSC 98-2410-
H-003 -124 -MY2.
References
Arns, M., Kleinnijenhuis, M., Fallahpour, K., & Breteler, R.
(2008). Golf Performance Enhancement and Real-Life
Neurofeedback Training Using Personalized Event-
Locked EEG Proles. Journal of Neurotherapy, 11(4),
11–18. doi:10.1080/10874200802149656
Babiloni, C., Del Percio, C., Iacoboni, M., Infarinato, F., Lizio,
R., Marzano, N., . . . Eusebi, F. (2008). Golf putt outcomes
are predicted by sensorimotor cerebral EEG rhythms.
The Journal of Physiology, 586(1), 131–139. PubMed
doi:10.1113/jphysiol.2007.141630
Baumeister, J., Von Detten, S., Van Niekerk, S.M., Schubert,
M., Ageberg, E., & Louw, Q.A. (2013). Brain activity in
predictive sensorimotor control for landings: An EEG pilot
Neurofeedback Training and Golf Putting Performance 635
JSEP Vol. 37, No. 6, 2015
study. International Journal of Sports Medicine, 34(12),
1106–1111. PubMed doi:10.1055/s-0033-1341437
Chatrian, G.E., Lettich, E., & Nelson, P.L. (1985). Ten percent
electrode system for topographic studies of spontaneous
and evoked EEG activity. The American Journal of EEG
Technology, 25, 83–92.
Cheng, M.Y., Hung, C.L., Huang, C.J., Chang, Y.K., Lo, L.C.,
Shen, C., & Hung, T.M. (2015). Expert-novice differ-
ences in SMR activity during dart throwing. Biological
Psychology, 110, 212–218. PubMed doi:10.1016/j.bio-
psycho.2015.08.003
Chuang, L.Y., Huang, C.J., & Hung, T.M. (2013). The differ-
ences in frontal midline theta power between successful
and unsuccessful basketball free throws of elite basketball
players. International Journal of Psychophysiology, 90(3),
321–328. PubMed doi:10.1016/j.ijpsycho.2013.10.002
Cohen, J. (1992). A power primer. Psychological Bul-
letin, 112(1), 155–159. PubMed doi:10.1037/0033-
2909.112.1.155
Davidson, R.J. (1988). EEG measures of cerebral asymmetry:
conceptual and methodological issues. The International
Journal of Neuroscience, 39(1-2), 71–89. PubMed
doi:10.3109/00207458808985694
Del Percio, C., Iacoboni, M., Lizio, R., Marzano, N., Infarinato,
F., Vecchio, F., . . . Babiloni, C. (2011). Functional coupling
of parietal alpha rhythms is enhanced in athletes before
visuomotor performance: a coherence electroencepha-
lographic study. Neuroscience, 175, 198–211. PubMed
doi:10.1016/j.neuroscience.2010.11.031
Doppelmayr, M., & Weber, E. (2011). Effects of SMR and theta/
beta neurofeedback on reaction times, spatial abilities, and
creativity. Journal of Neurotherapy, 15(2), 115–129. doi:
10.1080/10874208.2011.570689
Egner, T., & Gruzelier, J. (2001). Learned self-regulation of
EEG frequency components affects attention and event-
related brain potentials in humans. Neuroreport, 12(18),
4155–4159. PubMed doi:10.1097/00001756-200112210-
00058
Egner, T., Strawson, E., & Gruzelier, J. (2002). EEG signa-
ture and phenomenology of alpha/theta neurofeedback
training versus mock feedback. Applied Psycho-
physiology and Biofeedback, 27(4), 261–270. PubMed
doi:10.1023/A:1021063416558
Egner, T., Zech, T.F., & Gruzelier, J. (2004). The effects of
neurofeedback training on the spectral topography of the
electroencephalogram. Clinical Neurophysiology, 115(11),
2452–2460. PubMed doi:10.1016/j.clinph.2004.05.033
Enriquez-Geppert, S., Huster, R.J., Scharfenort, R., Mokom,
Z.N., Zimmermann, J., & Herrmann, C.S. (2014). Modula-
tion of frontal-midline theta by neurofeedback. Biological
Psychology, 95(1), 59–69. PubMed doi:10.1016/j.biopsy-
cho.2013.02.019
Fitts, P.M., & Posner, M.I. (1967). Human Performance. Basic
Concepts in Psychology. Retrieved from http://lccn.loc.
gov/67011662
Grosse-Wentrup, M., Schölkopf, B., & Hill, J. (2011).
Causal inuence of gamma oscillations on the senso-
rimotor rhythm. NeuroImage, 56(2), 837–842. PubMed
doi:10.1016/j.neuroimage.2010.04.265
Gruzelier, J. (2014a). Differential effects on mood of 12–15
(SMR) and 15–18 (beta1) Hz neurofeedback. International
Journal of Psychophysiology, 93(1), 112–115. PubMed
doi:10.1016/j.ijpsycho.2012.11.007
Gruzelier, J. (2014b). EEG-neurofeedback for optimising
performance. III: A review of methodological and theo-
retical considerations. Neuroscience and Biobehavioral
Reviews, 44, 159–182. PubMed doi:10.1016/j.neubio-
rev.2014.03.015
Gruzelier, J., Foks, M., Steffert, T., Chen, M.J.L., & Ros, T.
(2014). Benecial outcome from EEG-neurofeedback
on creative music performance, attention and well-being
in school children. Biological Psychology, 95(1), 86–95.
PubMed doi:10.1016/j.biopsycho.2013.04.005
Gruzelier, J., Inoue, A., Smart, R., Steed, A., & Steffert, T.
(2010). Acting performance and flow state enhanced
with sensory-motor rhythm neurofeedback comparing
ecologically valid immersive VR and training screen sce-
narios. Neuroscience Letters, 480(2), 112–116. PubMed
doi:10.1016/j.neulet.2010.06.019
Hoedlmoser, K., Pecherstorfer, T., Gruber, G., Anderer, P.,
Doppelmayr, M., Klimesch, W., & Schabus, M. (2008).
Instrumental conditioning of human sensorimotor rhythm
(12-15 Hz) and its impact on sleep as well as declarative
learning. Sleep, 31(10), 1401–1408. PubMed
Kao, S.C., Huang, C.J., & Hung, T.M. (2013). Frontal midline
theta is a specic indicator of optimal attentional engage-
ment during skilled putting performance. Journal of
Sport & Exercise Psychology, 35(5), 470–478 Retrieved
from http://www.ncbi.nlm.nih.gov/pubmed/24197715.
PubMed
Kao, S.C., Huang, C.J., & Hung, T.M. (2014). Neurofeedback
training reduces frontal midline theta and improves put-
ting performance in expert golfers. Journal of Applied
Sport Psychology, 26(3), 271–286. doi:10.1080/104132
00.2013.855682
Klimesch, W., Doppelmayr, M., Pachinger, T., & Ripper, B.
(1997). Brain oscillations and human memory: EEG cor-
relates in the upper alpha and theta band. Neuroscience
Letters, 238(1-2), 9–12. PubMed doi:10.1016/S0304-
3940(97)00771-4
Kober, S.E., Witte, M., Stangl, M., Väljamäe, A., Neuper, C., &
Wood, G. (2014). Shutting down sensorimotor interference
unblocks the networks for stimulus processing: An SMR
neurofeedback training study. Clinical Neurophysiology.
PubMed
Krane, V., & Williams, J.M. (2006). Psychological characteris-
tics of peak performance. In J.M. Williams (Ed.), Applied
sport psychology: Personal growth to peak performance.
New York: McGraw-Hill.
Landers, D.M., Petruzzello, S.J., Salazar, W., Crews, D.J.,
Kubitz, K.A., Gannon, T.L., & Han, M. (1991). The inu-
ence of electrocortical biofeedback on performance in
pre-elite archers. Medicine and Science in Sports and Exer-
cise, 23(1), 123–129. PubMed doi:10.1249/00005768-
199101000-00018
Logan, G.D., Hockley, W.E., & Lewandowsky, S. (1991). Auto-
maticity and memory. In W.E. Hockley & S. Lewandowsky
(Eds.), Relating Theory and Data: Essays on Human
636 Cheng et al.
JSEP Vol. 37, No. 6, 2015
Memory in Honor of Bennet B. Murdock (pp. 347–366).
East Sussex, UK: Psychology Press.
López-Larraz, E., Escolano, C., & Minguez, J. (2012). Upper
alpha neurofeedback training over the motor cortex
increases SMR desynchronization in motor tasks. Pro-
ceedings of the Annual International Conference of the
IEEE Engineering in Medicine and Biology Society,
EMBS, 2012, 4635–4638. http://doi.org/doi:10.1109/
EMBC.2012.6347000
Mann, C.A., Sterman, M.B., & Kaiser, D.A. (1996). Suppres-
sion of EEG rhythmic frequencies during somato-motor
and visuo-motor behavior. International Journal of Psy-
chophysiology, 23(1-2), 1–7. PubMed doi:10.1016/0167-
8760(96)00036-0
Nicholls, A.R. (2007). A longitudinal phenomenological analy-
sis of coping effectiveness among Scottish international
adolescent golfers. European Journal of Sport Science,
7(3), 169–178. doi:10.1080/17461390701643034
Niemarkt, H.J., Jennekens, W., Pasman, J.W., Katgert, T., Van
Pul, C., Gavilanes, A.W.D., . . . Andriessen, P. (2011).
Maturational changes in automated EEG spectral power
analysis in preterm infants. Pediatric Research, 70(5),
529–534. PubMed doi:10.1203/PDR.0b013e31822d748b
Nierhaus, T., Forschack, N., Piper, S.K., Holtze, S., Krause,
T., Taskin, B., . . . Villringer, A. (2015). Imperceptible
somatosensory stimulation alters sensorimotor background
rhythm and connectivity. The Journal of Neuroscience,
35(15), 5917–5925. PubMed doi:10.1523/JNEURO-
SCI.3806-14.2015
Pelz, D., & Frank, J.A. (2000). Dave Pelz’s putting bible: The
complete guide to mastering the green. New York, NY:
Doubleday.
Raymond, J., Sajid, I., Parkinson, L.A., & Gruzelier, J. (2005).
Biofeedback and dance performance: A preliminary
investigation. Applied Psychophysiology and Biofeedback,
30(1), 65–73. PubMed doi:10.1007/s10484-005-2175-x
Ring, C., Cooke, A., Kavussanu, M., McIntyre, D., & Masters,
R. (2015). Investigating the efcacy of neurofeedback
training for expediting expertise and excellence in
sport. Psychology of Sport and Exercise, 16, 118–127.
doi:10.1016/j.psychsport.2014.08.005
Ros, T., Moseley, M.J., Bloom, P.A., Benjamin, L., Parkinson,
L.A., & Gruzelier, J. (2009). Optimizing microsurgical
skills with EEG neurofeedback. BMC Neuroscience, 10(1),
87. PubMed doi:10.1186/1471-2202-10-87
Sauseng, P., Hoppe, J., Klimesch, W., Gerloff, C., & Hummel,
F.C. (2007). Dissociation of sustained attention from cen-
tral executive functions: Local activity and interregional
connectivity in the theta range. The European Journal
of Neuroscience, 25(2), 587–593. PubMed doi:10.1111/
j.1460-9568.2006.05286.x
Schneider, W., & Shiffrin, R.M. (1977). Controlled and
automatic human information processing: I. Detection,
search, and attention. Psychological Review, 84(1), 1–66.
doi:10.1037/0033-295X.84.1.1
Semlitsch, H.V., Anderer, P., Schuster, P., & Presslich, O.
(1986). A solution for reliable and valid reduction of ocular
artifacts, applied to the P300 ERP. Psychophysiology,
23(6), 695–703. PubMed doi:10.1111/j.1469-8986.1986.
tb00696.x
Vernon, D., Egner, T., Cooper, N., Compton, T., Neilands,
C., Sheri, A., & Gruzelier, J. (2003). The effect of
training distinct neurofeedback protocols on aspects of
cognitive performance. International Journal of Psycho-
physiology, 47(1), 75–85. PubMed doi:10.1016/S0167-
8760(02)00091-0
Weber, E., Köberl, A., Frank, S., & Doppelmayr, M. (2011).
Predicting successful learning of SMR neurofeedback
in healthy participants: Methodological considerations.
Applied Psychophysiology and Biofeedback, 36(1), 37–45.
PubMed doi:10.1007/s10484-010-9142-x
Wu, T., Chan, P., & Hallett, M. (2008). Modications of the
interactions in the motor networks when a movement
becomes automatic. The Journal of Physiology, 586(Pt 17),
4295–4304. PubMed doi:10.1113/jphysiol.2008.153445
Manuscript submitted: June 30, 2015
Revision accepted: October 08, 2015
... It should be noted however that the characteristics of the so-called ''SMR'', especially temporal domain characteristics (i.e., frequency bands), vary quite extensively between studies. Broadly speaking, SMR have been defined as EEG rhythms in the central region, in particular two kinds of EEG band activities: mu (7-11 Hz) and beta (12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30) [43]. Mu rhythms were described by Gastaut [34]. ...
... Nonetheless, the neurofeedback procedure was not associated with any clear EEG pattern changes. Following this study, several other neurofeedback experiments were conducted in the aim of enhancing athletes' performance in different sports, e.g., in golf [17,82], swimming [28], dance [36,81] or athleticism [63]. Mirifar et al. [65] proposed a systematic review of these neurofeedback studies. ...
... The first case study investigating the use of SMR neurofeedback (reward production of low beta activity (15)(16)(17)(18)(19)(20)(21) and down-regulation of production of theta activity (4-8 Hz) in a patient with a stroke was published in 1995, and suggested a possible beneficial effect of neurofeedback on motor recovery [89]. However, the protocol was not specifically designed in line with the literature on MI in post-stroke. ...
Article
Many Brain Computer Interface (BCI) and neurofeedback studies have investigated the impact of sensorimotor rhythm (SMR) self-regulation training procedures on motor skills enhancement in healthy subjects and patients with motor disabilities. This critical review aims first to introduce the different definitions of SMR EEG target in BCI/Neurofeedback studies and to summarize the background from neurophysiological and neuroplasticity studies that led to SMR being considered as reliable and valid EEG targets to improve motor skills through BCI/neurofeedback procedures. The second objective of this review is to introduce the main findings regarding SMR BCI/neurofeedback * Corresponding author. Service d'explorations fonctionnelles du système nerveux, clinique du sommeil, CHU de Bordeaux, place Amélie Raba-Léon, 126 C. Jeunet et al. in healthy subjects. Third, the main findings regarding BCI/neurofeedback efficiency in patients with hypokinetic activities (in particular, motor deficit following stroke) as well as in patients with hyperkinetic activities (in particular, Attention Deficit Hyperactivity Disorder, ADHD) will be introduced. Due to a range of limitations, a clear association between SMR BCI/neurofeedback training and enhanced motor skills has yet to be established. However, SMR BCI/neurofeedback appears promising, and highlights many important challenges for clinical neurophysiology with regards to therapeutic approaches using BCI/neurofeedback.
... Many studies have used NFT to improve the sports performance of athletes (Raymond et al., 2005;Faridnia et al., 2012;Strizhkova et al., 2012;Cheng et al., 2015a;Mikicin et al., 2015;Ring et al., 2015). For instance, Raymond et al. (2005) improved the dance performance of the college dance sports team by increasing the alpha/theta ratio of the Pz electrode. ...
... Mikicin et al. (2015) reduced student-athletes' attention-reaction by increasing beta1 and SMR and decrease theta and beta2 of C3 and C4 electrodes. Cheng et al. (2015a) improved golfers' putting performance by increasing SMR of the Cz electrode. ...
... Collura (2013) suggest this could be a kind of influence of the ''entrainment'' effect: in addition to the frequency bands and channels involved in feedback, the EEG power of other nearby frequency bands and the cerebral cortex also showed a certain degree of change. Other neurofeedback also has reported similar phenomena (Cheng et al., 2015a). ...
Article
Full-text available
Previous literature on shooting performance neurofeedback training (SP-NFT) to enhance performance usually focused on changes in behavioral indicators, but research on the physiological features of SP-NFT is lacking. To explore the effects of SP-NFT on trainability and neuroplasticity, we conducted a study in which 45 healthy participants were randomly divided into three groups: based on sensory-motor rhythm of C3, Cz and C4 (SMR group), based on alpha rhythm of T3 and T4 (Alpha group), and no NFT (control group). The training was performed for six sessions for 3 weeks. Before and after the SP-NFT, we evaluated changes in shooting performance and resting electroencephalography (EEG) frequency power, participant’s subjective task appraisal, neurofeedback trainability score, and EEG feature. Statistical analysis showed that the shooting performance of the participants in the SMR group improved significantly, the participants in the Alpha group decreased, and that of participants in the control group have no change. Meanwhile, the resting EEG power features of the two NFT groups changed specifically after training. The training process data showed that the training difficulty was significantly lower in the SMR group than in the Alpha group. Both NFT groups could improve the neurofeedback trainability scores and change the feedback features by means of their mind strategy. These results may provide evidence of trainability and neuroplasticity for SP-NFT, suggesting that the SP-NFT is effective in brain regulation and thus provide a potential method to improve shooting performance.
... Nonetheless, the NF procedure was not associated with any clear modifications of neurophysiological patterns. Following this study, several other NF experiments have been led in the aim of enhancing athletes' performance in different sports including golf 11,12 , swimming 13 , dance 14,15 or athletics 16 . Mirifar et al. 3 provided a systematic review of these NF studies. ...
... Covert attention has been shown to be underlain by specific neurophysiological patterns, and notably by a lateralised modulation of α-power (8)(9)(10)(11)(12)(13)(14) in the visual cortex 21,22 . More specifically, EEG studies have suggested that CVSA would be underlain by an increased power in the α frequency band over occipital areas (i.e., supposedly in the region of the visual cortex) ipsilateral to the hemi-field containing the object the person is focusing on. ...
Article
Full-text available
Advances in sports sciences and neurosciences offer new opportunities to design efficient and motivating sport training tools. For instance, using NeuroFeedback (NF), athletes can learn to self-regulate specific brain rhythms and consequently improve their performances. Here, we focused on soccer goalkeepers’ Covert Visual Spatial Attention (CVSA) abilities, which are essential for these athletes to reach high performances. We looked for Electroencephalography (EEG) markers of CVSA usable for virtual reality-based NF training procedures, i.e., markers that comply with the following criteria: (1) specific to CVSA, (2) detectable in real-time and (3) related to goalkeepers’ performance/expertise. Our results revealed that the best-known EEG marker of CVSA—increased α-power ipsilateral to the attended hemi-field— was not usable since it did not comply with criteria 2 and 3. Nonetheless, we highlighted a significant positive correlation between athletes’ improvement in CVSA abilities and the increase of their α-power at rest. While the specificity of this marker remains to be demonstrated, it complied with both criteria 2 and 3. This result suggests that it may be possible to design innovative ecological training procedures for goalkeepers, for instance using a combination of NF and cognitive tasks performed in virtual reality.
... As shown in Table 5, during a prolonged driving task, the driver's quality of attention decreases. Fast-tempo music will further deteriorate the quality of attention (Cheng et al., 2015). Fast-tempo and high-intensity music (e.g., rock and roll) directly affect driving safety and operational efficiency (Venter, 2011). ...
... Table for Attention Level Evaluation Based on EEG(Cheng et al., 2015).Note. EEG ¼ electroencephalogram; SMR ¼ sensorimotor rhythm. ...
... Various biofeedback modalities (notably heart rate variability and surface electromyography) have been used to enhance performance psychology in athletes (Rijken et al., 2016;Jiménez Morgan and Molina Mora, 2017;Rusciano et al., 2017). EEG-based biofeedback (i.e., neurofeedback) has also been shown to improve athletic performance (Vernon, 2005;Cheng et al., 2015;Mirifar et al., 2017;Liu et al., 2018;Xiang et al., 2018). EEG studies on expertise and skilled performance have mostly reported on changes or differences in alpha and beta power, which are often interpreted as reflecting cortical activation and sensorimotor rhythm, respectively. ...
Article
Full-text available
The elite sports environment provides a unique setting for studying human performance, where both cognitive and physical demands are high. Successful performance in sport is contingent upon key cognitive skills such as attention, perception, working memory and decision-making. The demands of competitive sport also increase loading on the central nervous system (CNS). Neuroimaging methods such as functional magnetic resonance imaging (fMRI), functional near infrared spectroscopy (fNIRS) and electroencephalography (EEG) offer the potential to investigate the cognitive demands of sport, neuroplasticity of athletes, and biofeedback training. However, practical and technical limitations of these methods have generally limited their use to laboratory-based studies of athletes during simulated sporting tasks. This review article, provides a brief overview of research that has applied neuroimaging technology to study various aspects of cognitive function during sports performance in athletes, alternative methods for measuring CNS loading [e.g., direct current (DC) potential], possible solutions and avenues of focus for future neuroergonomics research in sport.
Chapter
The brain is the main locus of control for our behaviour and psychological states. Superior sport performance occurs when both the physical and mental dimension of an athlete converge in an adaptive manner to meet the challenges of the task. This chapter provides an overview of how the brain works for athletes at different levels of expertise and discusses how brain activity can be controlled to achieve superior sports performance. The multi-action plan model provides an alternative perspective for understanding the relationship between performance effectiveness and the utilization of resources in sports performance. The chapter discusses several limitations arising from the current literature and makes suggestions for future research in the hope of establishing a consensus on protocols to be followed in future studies. It provides an overview of neurofeedback training (NFT) studies used to investigate sports performance and proposes various criteria that might be used to assess the effects of NFT in sports domains.
Article
Full-text available
The present study extends the sport neuroscience literature by comparing elite and amateur golfers during golf putting and examining the essential cognitive-motor processes that may contribute to understanding the superior cognitive- motor performance of skilled performers. Twenty elite and 18 amateur golfers were recruited to perform 60 putts while individual EEGs were recorded. Compared with the amateur golfers, the elite golfers were characterized by (1) lower alpha 2 power at Pz and T8 two seconds before putt release; (2) lower alpha 2 power at Fz and T8 and lower mu 2 power one second before putting; and (3) lower alpha 2 coherence at Fz–T7 and Fz–T8. This suggests that the elite golfers had higher levels of attention to response motor programming and visuospatial attention and less cognitive-motor interference before putting. These findings not only point to the importance of refining brain processes but also specify essential cognitive-motor processes for superior performance in athletes.
Chapter
When it is applied in sports, biofeedback training (BFT) is a technique that can enable athletes to modify their psychophysiological behavior by regulating their biological signals (referred to as modalities) in response to real-time feedback, which may result in desirable psychological processes and/or behavioral outcomes, such as improved accuracy in shooting performance. The most common modalities include heart rate variability (HRV), electrodermal activity (EDA), muscle activity (EMG), respiration rate, blood pressure, and neural activity (neurofeedback training, NFT). Electroencephalography (EEG) has been the most commonly applied NFT method for sport performance enhancement. In this chapter, we review the current evidence, provide commentary on the level of evidence, and offer directions for future research.
Article
Previous studies have revealed that several cortical signatures are associated with superior motor performance in sports, particularly precision sports. This review examined the strength of the evidence from these studies so that a clear conclusion could be drawn and a concrete direction for future efforts revealed. A total of 26 articles assessing the relationship between cortical activity and precision motor performance were extracted from databases. This review concluded that among the electroencephalographic components examined, only sensorimotor rhythm demonstrated a consistent and causal relationship with superior precision motor performance, whereas findings related to the left temporal alpha and frontal theta and alpha rhythms were not consistent and lacked the evidence needed to draw a causal inference for a role in superior precision motor performance. Future studies would benefit from methodological improvements including larger sample sizes, narrower skill-gap comparisons, evidenced-based and refined neurofeedback-training protocols, and consideration of ecological validity.
Article
The purpose of this study was to investigate the effectiveness of a sensorimotor rhythm (SMR) neurofeedback training (NFT) and biofeedback training (BFT) intervention on ice hockey shooting performance. Specifically, the purpose was to examine (a) whether an NFT/BFT program could improve ice hockey shooting performance, (b) whether the implementation of an SMR-NFT intervention leads to neurological adaptations during performance, and (c) whether such neurological changes account for improvement in shooting performance. Using a longitudinal stratified random control design, results demonstrated that while both SMR-NFT/BFT and control groups improved performance, the rate of improvement for the SMR-NFT/BFT group was significantly higher than the control. Participants in the SMR-NFT/BFT group demonstrated the ability to significantly increase SMR power from pre- to postintervention in the lab. However, no significant changes in SMR power were found during shooting performance. This result may be suggestive of differing cortical activity present during motor-skill preparation.
Article
Full-text available
Most sensory input to our body is not consciously perceived. Nevertheless, it may reach the cortex and influence our behavior. In this study, we investigated noninvasive neural signatures of unconscious cortical stimulus processing to understand mechanisms, which (1) prevent low-intensity somatosensory stimuli from getting access to conscious experience and which (2) can explain the associated impediment of conscious perception for additional stimuli. Stimulation of digit 2 in humans far below the detection threshold elicited a cortical evoked potential (P1) at 60 ms, but no further somatosensory evoked potential components. No event-related desynchronization was detected; rather, there was a transient synchronization in the alpha frequency range. Using the same stimulation during fMRI, a reduced centrality of contralateral primary somatosensory cortex (SI) was found, which appeared to be mainly driven by reduced functional connectivity to frontoparietal areas. We conclude that after subthreshold stimulation the (excitatory) feedforward sweep of bottom-up processing terminates in SI preventing access to conscious experience. We speculate that this interruption is due to a predominance of inhibitory processing in SI. The increase in alpha activity and the disconnection of SI from frontoparietal areas are likely correlates of an elevated perception threshold and may thus serve as a gating mechanism for the access to conscious experience. Copyright © 2015 the authors 0270-6474/15/355917-09$15.00/0.
Article
Full-text available
A single-subject design was used to examine the influence of one session of neurofeedback training (NFT) on reducing frontal midline theta (Fm theta) amplitude and enhancing golf putting performance. Posttraining, three highly skilled golfers improved in putting score or score stability. Although the Fm theta amplitude during the pre-putt period inconsistently decreased across participants, all golfers exhibited lower Fm theta amplitude during the resting condition following NFT, suggesting that the tonic reduction of Fm theta may play a role in subsequent performance improvement. Overall, these results indicate that a short session of NFT may be an effective method of performance enhancement in some sports.
Article
Full-text available
Background: Meta analyses on neonatal mechanical ventilation suggested no difference between conventional and patient triggered ventilation (PTV) in preterm neonates. Newer forms of respiratory support with perceived advantages continue to entice Neonatologists.
Article
Previous evidence suggests that augmented sensorimotor rhythm (SMR) activity is related to the superior regulation of processing cognitive-motor information in motor performance. However, no published studies have examined the relationship between SMR and performance in precision sports; thus, this study examined the relationship between SMR activity and the level of skilled performance in tasks requiring high levels of attention (e.g., dart throwing). We hypothesized that skilled performance would be associated with higher SMR activity. Fourteen dart-throwing experts and eleven novices were recruited. Participants were requested to perform 60 dart throws while EEG was recorded. The 2 (Group: Expert, Novice) x 2 (Time window: –2000 ms to –1000 ms, –1000 ms to 0 ms) ANOVA showed that the dart-throwing experts maintained a relatively higher SMR power than the novices before dart release. These results suggest that SMR might reflect the adaptive regulation of cognitive-motor processing during the preparatory period.
Article
An 81-electrode system is described which is designed for topographic studies of spontaneous and evoked EEG activities. This method combines the standard leads of the International 10-20 System with supplementary electrodes applied midway between leads of the 10-20 system or electrodes in turn situated between 10-20 leads. Auxiliary electrode designations refer to the underlying brain areas and to adjacent leads of the 10-20 method. The utilization of this '10% system' is suggested to promote standardization.
Article
Objectives This experiment examined whether electroencephalographic (EEG)-based neurofeedback could be used to train recreational golfers to regulate their brain activity, expedite skill acquisition, and promote robust performance under pressure. Design We adopted a mixed-multifactorial design, with group (neurofeedback, control) as a between-subjects factor, and pressure (low, high), session (pre-test, acquisition 1, acquisition 2, acquisition 3, post-test), block (putts within each training session), and epoch (cortical activity in the seconds around movement initiation) as within-subject factors. Methods Recreational golfers received three hours of either true (to reduce frontal EEG high-alpha power, N = 12) or false (control, N = 12) neurofeedback training sandwiched between pre-test and post-test sessions during which we collected measures of cortical activity (EEG) and putting performance under both low and high pressure conditions. Results Individuals in the neurofeedback group learned to reduce their frontal high-alpha power before striking putts. Despite causing this more “expert-like” pattern of cortical activity, neurofeedback training failed to selectively enhance performance, as both groups improved their putting performance similarly from the pre-test to the post-test. Finally, both groups performed robustly under pressure. Conclusions Performers can learn to regulate their brain activity using neurofeedback training. However, research identifying the cortical correlates of expertise is required to refine neurofeedback interventions if this training method is to expedite learning. Suggestions for future neurofeedback interventions are discussed.
Article
In the present study, we investigated how the electrical activity in the sensorimotor cortex contributes to improved cognitive processing capabilities and how SMR (sensorimotor rhythm, 12-15Hz) neurofeedback training modulates it. Previous evidence indicates that higher levels of SMR activity reduce sensorimotor interference and thereby promote cognitive processing. Participants were randomly assigned to two groups, one experimental (N=10) group receiving SMR neurofeedback training, in which they learned to voluntarily increase SMR, and one control group (N=10) receiving sham feedback. Multiple cognitive functions and electrophysiological correlates of cognitive processing were assessed before and after 10 neurofeedback training sessions. The experimental group but not the control group showed linear increases in SMR power over training runs, which was associated with behavioural improvements in memory and attentional performance. Additionally, increasing SMR led to a more salient stimulus processing as indicated by increased N1 and P3 event-related potential amplitudes after the training as compared to the pre-test. Finally, functional brain connectivity between motor areas and visual processing areas was reduced after SMR training indicating reduced sensorimotor interference. These results indicate that SMR neurofeedback improves stimulus processing capabilities and consequently leads to improvements in cognitive performance. The present findings contribute to a better understanding of the mechanisms underlying SMR neurofeedback training and cognitive processing and implicate that SMR neurofeedback might be an effective cognitive training tool.
Article
The aim of this study was to explore coping effectiveness among international golfers. Five Scottish international adolescent golfers (mean age 16.6 years, s = 0.6) maintained daily coping effectiveness diaries over a 28-day period during their competitive season. Data were thematically analysed using interpretive phenomenological analysis (Smith & Osborn, 200329. Smith , J.A. and Osborn , M. 2003. “Interpretative phenomenological analysis”. In Qualitative psychology: A practical guide to research methods, Edited by: Smith , J.A. 51–80. London: Sage. View all references). The participants reported 56 effective coping strategies and 23 ineffective coping strategies. The unique finding from this study is that the same coping strategies were often rated as being both effective and ineffective, even when they were employed to manage the same stressor. This suggests that the goodness-of-fit approach and the choice of coping strategy theories may not be adequate explanations of coping effectiveness. Applied practitioners working with golfers are encouraged to teach their clients a variety of coping strategies, which should be deployed in combination.