Original article 1
Effectiveness of a single session of dual-transcranial direct
current stimulation in combination with upper limb
robotic-assisted rehabilitation in chronic stroke patients:
a randomized, double-blind, cross-over study
Stéphanie Dehema,e, Maxime Gilliauxa,e, Thierry Lejeunea,b,e,
Emmanuelle Delaunoisb, Paul Mbondab, Yves Vandermeerenc,d,e,
Christine Detrembleura,e and Gaëtan Stoquarta,b,e
The impact of transcranial direct current stimulation (tDCS)
is controversial in the neurorehabilitation literature. It has
been suggested that tDCS should be combined with other
therapy to improve their efficacy. To assess the
effectiveness of a single session of upper limb robotic-
assisted therapy (RAT) combined with real or sham-tDCS in
chronic stroke patients. Twenty-one hemiparetic chronic
stroke patients were included in a randomized, controlled,
cross-over double-blind study. Each patient underwent two
sessions 7 days apart in a randomized order: (a) 20 min of
real dual-tDCS associated with RAT (REAL + RAT) and (b)
20 min of sham dual-tDCS associated with RAT
(SHAM + RAT). Patient dexterity (Box and Block and Purdue
Pegboard tests) and upper limb kinematics were evaluated
before and just after each intervention. The assistance
provided by the robot during the intervention was also
recorded. Gross manual dexterity (1.8 ± 0.7 blocks,
P = 0.008) and straightness of movement (0.01 ± 0.03,
P < 0.05) improved slightly after REAL + RAT compared with
before the intervention. There was no improvement after
SHAM + RAT. The post-hoc analyses did not indicate any
difference between interventions: REAL + RAT and
SHAM + RAT (P > 0.05). The assistance provided by the
en esta primera parte habla de la discapacidad residual en MMSS tras ictus, de como mejorar esta discapacidad con
nuevs intervenciones, entre ellas tDCS y com esta se combina con terapia convencional
Introduction activities of daily living (Elsner et al., 2013; Pollock et al.,
Upper limb (UL) poststroke rehabilitation remains a
challenge (Langhorne et al., 2011; Pollock et al., 2014).
Despite intensive interdisciplinary rehabilitation, many
patients retain permanent poststroke disability. To
reduce this disability, researchers are looking for new
effective treatments to improve motor recovery after
stroke. In this context, transcranial direct current stimu-
lation (tDCS) has been shown to reinforce rehabilitative
treatments (Pollock et al., 2014). tDCS involves applica-
tion of a small electrical current to the skull through two
surface electrodes, which can modulate neuronal excit-
ability in stroke patients (Grefkes and Fink, 2012; Kandel
et al., 2012; Page et al., 2015). Recent Cochrane reviews
have reported that compared with sham or conventional
therapy, tDCS reduced UL impairment in stroke
patients, but did not improve their ability to perform
0342-5282 Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.
Copyright r 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
robot was similar during both interventions (P > 0.05). The
results showed a slight improvement in hand dexterity and
arm movement after the REAL + RAT tDCS intervention. The
observed effect after a single session was small and not
clinically relevant. Repetitive sessions could increase the
benefits of this combined approach. International Journal of
Rehabilitation Research 00:000–000 Copyright © 2018
Wolters Kluwer Health, Inc. All rights reserved.
International Journal of Rehabilitation Research 2018, 00:000–000
Keywords: rehabilitation, robotic, stroke, transcranial direct current
stimulation, upper limb
a
Secteur des Sciences de la Santé, Institut de Recherche Expérimentale et
Clinique (IREC), Neuro Musculo Skeletal Lab (NMSK), bCliniques Universitaires
Saint-Luc, Service de Médecine Physique et Réadaptation, cInstitute of
Neuroscience (IoNS), Université Catholique de Louvain, Brussels, dCHU UCL
Namur, Neurology Department, Stroke Unit/NeuroModulation Unit (NeMU), Yvoir
and eUniversité Catholique de Louvain, Louvain Bionics, Louvain-la-Neuve,
Belgium
Correspondence to Stéphanie Dehem, PT, Secteur des Sciences de la Santé,
Institut de Recherche Expérimentale et Clinique (IREC), Neuro Musculo Skeletal
Lab (NMSK), Université Catholique de Louvain (UCL), Avenue Mounier 53,
B-1200 Brussels, Belgium
Tel: + 32 327 645 375; e-mail: stephanie.dehem@uclouvain.be
Received 20 October 2017 Accepted 22 January 2018
2014). Overall, the small effect size reported does not
support generalized use of tDCS alone in routine clinical
practice (Butler et al., 2013; Podubecká et al., 2014;
Pollock et al., 2014; Tedesco Triccas et al., 2015a, 2015b).
The effect of a single dual-tDCS intervention has been
studied by Lefebvre et al. (2014) who explored the effect
of a 20 min single session on motor skill learning in a
double-blind, randomized, cross-over study. The preci-
sion grip and dexterity of stroke patients’ paretic hand
improved after the stimulation (Lefebvre et al., 2014). In
a previous study, they also analysed online motor skill
during a single session of dual-tDCS in combination with
an UL motor skill task. The study showed a qualitative
and quantitative improvement in patients’ motor skill
with long-term retention (1 week) and generalization
(Lefebvre et al., 2013).
DOI: 10.1097/MRR.0000000000000274
 2 International Journal of Rehabilitation Research 2018, Vol 00 No 00
habla de pq combinar tDCS con TAR
To enhance motor recovery after stroke, several authors
have recommended using combinations of UL inter-
ventions, such as robotic-assisted therapy (RAT) (Hesse
et al., 2011; Mehrholz et al., 2012; Ochi et al., 2013) with
tDCS stimulation (Harvey and Stinear 2010; Butler et al.,
2013). RAT follows the current guidelines for neuroreh-
abilitation of the UL as it intensifies the therapy, pro-
vides assistance as needed, and provides feedback and
quantified measurement of patient movements (Krebs
et al., 2003; Pignolo 2009; Langhorne et al., 2011). Recent
systematic reviews recommended RAT in stroke patients
to reduce UL impairment and improve activities of daily
living performance (Norouzi-Gheidari et al., 2012; Pollock
et al., 2014; Mehrholz et al., 2015). Moreover, the assis-
tance provided by the robot is a good reflection of motor
control during rehabilitation.
RAT alone has been shown to improve UL kinematics
(Daly et al., 2005; Finley et al., 2005; Kahn et al., 2006),
whereas tDCS alone improved fine and gross manual
dexterity (Lefebvre et al., 2013; Lefebvre et al., 2014;
Huyun-Kyu et al., 2014). RAT and tDCS could then be
complementary, and their combination could lead to
additional improvement in motor control. A recent review
analysed the effects of tDCS coupled with RAT in
poststroke UL rehabilitation (Simonetti et al., 2017). This
review showed that the combination led to the same
clinical improvement as RAT alone. The authors high-
lighted the difficulty in comparison of the studies
because of their heterogeneity and small number of col-
lected studies. The study of Giacobbe et al. (2013) was
the only study that analysed the combination of RAT and
tDCS in a single session. This study showed an
improvement in motor kinematics after a single session of
anodal tDCS provided before RAT in chronic stroke
patients (Giacobbe et al., 2013). To our knowledge, no
study has quantified the assistance provided to the
patient during combined RAT tDCS.
The aim of the present study was to assess the effec-
tiveness of a single session of RAT combined with dual-
tDCS in chronic stroke patients using a double-blind,
cross-over randomized control trial. We hypothesized that
the association of real dual-tDCS and RAT would
improve manual ability and kinematics in stroke patients
compared with sham-tDCS and RAT.
Methods
The ethics board of our Faculty of Medicine of the
Université Catholique de Louvain approved the study.
All patients provided written informed consent. The
study was carried out in accordance with the Declaration
of Helsinki. The trial was registered at ClinicalTrials.gov
(NCT02512289).
Patients
Fifty patients were screened and 21 fulfilled the inclu-
sion criteria and agreed to participate. All patients were
Copyright r 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
recruited at the Cliniques Universitaires Saint-Luc
(Brussels). Inclusion criteria were as follows: an ischae-
mic or haemorrhagic stroke, a 6-month delay since the
stroke, a minimum age of 18 years and the ability to
understand instructions. Exclusion criteria were as fol-
lows: presence of intracranial metal epilepsy, alcoholism,
pregnancy, UL injection of Botulinum toxin in the pre-
vious 3 months, another significant orthopaedic or neu-
rological disease that could alter active or passive mobility
of the ULs and a mini-mental state examination score
less than 18 indicating significant cognitive disorder
(Tombaugh and McIntyre, 1992; Lefebvre et al., 2014).
Neurological impairment was characterized using the
Stroke Impairment Assessment Set (SIAS) (Chino et al.,
1996). The SIAS score ranges from 0 to 76 (higher scores
correspond to less neurological impairment). The mod-
ified Rankin Scale was used to categorize the patient’s
functional independence, ranging from 0 to 5 (higher
scores corresponding to greater disability) (van Swieten
et al., 1988).
Each patient performed two treatment sessions in a
random order, 7 days apart: (a) real dual-tDCS associated
with RAT (REAL + RAT) and (b) sham dual-tDCS
associated with RAT (SHAM + RAT). Patients who
started with REAL + RAT and then underwent
SHAM + RAT were included in group A; patients allo-
cated to the reverse order were included in group B. A
consort flow diagram (Moher et al., 2010) of the protocol
is shown in Fig. 1.
Interventions
During each session of 20 min, the patient performed a
RAT with the tDCS electrodes positioned on the head.
The RAT interventions were identical in both sessions.
The dual-tDCS interventions used either real or sham
stimulation.
Dual transcranial direct-current stimulation
Dual-tDCS was delivered by an Eldith DC-Stimulator
(NeuroConn, Ilmenau, Germany) following the method
used by Lefebvre et al. (2014). The electrodes (35 cm2)
were soaked in 0.9% NaCl. The anode was positioned
using the EEG-10-20 system over the ipsilesional pri-
mary motor cortex (M1) and the cathode was placed over
the contralesional M1. During the real dual-tDCS
(REAL + RAT), the stimulator delivered DC at 1 mA
for 20 min (fade-in/fade-out 8 s). During the sham dual-
tDCS (SHAM + RAT), a short up-ramp (8 s fade-in) was
followed by 40 s of DC and 8 s of fade-out, after which
ineffective current pulses (110 μA over 15 ms, peak cur-
rent 3 ms) were delivered every 550 ms. This sham
intervention gave the patients the sensation that they
were receiving a stimulation without actually being the
case. One investigator established a list with the Eldith
codes (real/sham) for each patient session. Another

Fig. 1
Consort flow diagram of the stroke patients through each stage of the
study. tDCS, transcranial direct current stimulation.
investigator used these codes in a double-blinded
manner.
Robot-assisted therapy
RAT was performed using the REAplan robot (Axinesis,
Wavre, Belgique) (Fig. 2a) (Gilliaux et al., 2014a, 2014b).
REAplan is an end-effector robot than can move the
patient’s UL in a horizontal plane by a handle that the
patient grasps. The workspace was kept as large as pos-
sible depending on his/her arm length. A large, immer-
sive screen was installed in front of the patient to provide
visual feedback (Fig. 2b). The tasks consisted of moving
the paretic UL along reference trajectories while passing
through checkpoints. Indeed, a robotic ball collector had
to catch golf balls by following a specific path (Fig. 2b).
During the 20 min of RAT, the patients had to perform as
many movements as possible. During the game, the robot
guided the patients with an assistance-as-needed control
strategy. Indeed, the robot is has force and position
sensors to record the force on the handle and the position
of the handle (acquisition frequency = 125 Hz). The
sensor measurements are sent to a control unit that
determines the assistive forces on the basis of the dif-
ference between the patient’s hand position and the
reference trajectory. Two types of force can assist the
patient as needed: a lateral interaction force (Flat) and a
Copyright r 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Robotic and tDCS in stroke rehabilitation Dehem et al. 3
longitudinal interaction force (Flong) (Fig. 3) (Galinski,
2014). Flat limits deviation of the end-effector along the
reference trajectory. The control unit changes the stiff-
ness coefficient, Klat, during each movement on the basis
of the movement straightness (higher Klat corresponding
to more lateral assistance) (Sapin et al., 2017). Flong helps
the patient move along the reference trajectory at a
reference speed. The control unit changes the damping
coefficient, Clong, during each movement on the basis of
the smoothness of movement (high Clong corresponding
to a stronger longitudinal interaction force) (Sapin et al.,
2017).
Assessments
All patients were assessed by a functional and a robotic
evaluation before and after each intervention. The
assistance provided by the robot during the session was
also quantified. All patients were familiarized with the
protocol 7 days before the first intervention session. One
investigator performed all the assessments in a double-
blinded manner.
The primary outcome was paretic UL kinematics mea-
sured with the REAplan. It was chosen as the primary
outcome as it provides a quantitative and objective
assessment of active movements of the UL (Gilliaux
et al., 2014a, 2014b). Two tasks (free amplitude and tar-
get) were assessed in this study. For the free amplitude
task, the participants had to reach straight out in front of
them as far as they could and bring the arm back to the
starting position. For the target task, the participants had
to move the handle in the most precise and direct manner
towards a specific target placed at a distance of 14 cm in
front of them. These tasks were performed without any
assistance at spontaneous speeds. Each task was per-
formed 10 times consecutively, during which the end-
effector position was recorded. The rest period between
tasks was 1 min. For the free Amplitude task, the com-
puted kinematic included the speed, the straightness
(ratio between the amplitude and path length covered by
the participant; ratios closer to 1 indicate more rectilinear
paths) and the smoothness (ratio between the mean and
peak speed; ratios closer to 0 indicate less smooth
movements) (Rohrer et al., 2002). For the target task, the
accuracy index was also recorded (distance between
the target position and the end position achieved by the
participant, higher scores indicating more inaccurate
movements). Each index in this protocol was computed
from the 10 cycles of movement and averaged.
The secondary outcome was the gross and fine manual
dexterity assessed using the Box and Block Test (BBT)
and the Purdue Pegboard Test (PPT), respectively
(Mathiowetz et al., 1985; Brown and Jahanshahi, 1998).
During the RAT, motor control was quantified by the
movement velocity, the movement initiation time (Tinit) and
the mean assistance provided by the robot. The assistance
 4 International Journal of Rehabilitation Research 2018, Vol 00 No 00
Fig. 2
(a)
Illustrations of (a) the set-up of RAT combined with tDCS stimulations and (b) example of the patient’s visual interface during the RAT. 1 = planar-end
effector; 2 = patient’s visual interface; 3 = therapist’s visual interface; 4 = tDCS set up; 5 = electrode, 6 = Cursor to move; 7 = reference trajectory;
8 = target to reach; 9 and 13= shows to the patient how many checkpoints he/she need to cross. RAT, robotic-assisted therapy; tDCS, transcranial
direct current stimulation.
Fig. 3
Illustration of the assistance-as-needed used in the game. The force
computation is based on the definition of a reference trajectory and
combines two components: a lateral interaction force (Flat) proportional
to the stiffness coefficient (Klat) and a longitudinal interaction force
(Flong) proportional to the damping coefficient (Clat) (Galinski, 2014).
provided was represented by the stiffness coefficient (Klat)
and the damping coefficient (Clong).
Statistical analyses
Statistical analyses were carried out using the SigmaStat
3.5 software (WPCubed GmbH, Munich, Germany). For
each analysis, the normality and equality of variances
were checked with a significance level fixed at 0.05.
The homogeneity between group A and B was analysed
using a t-test. The equivalence of both baselines scores
was analysed using a paired t-test. To evaluate the
treatment effect (REAL + RAT or SHAM + RAT), a
one-way repeated measure analysis of variance
(RMANOVA) was used. If this test showed an interaction,
it was further analysed with a post-hoc test
(Holm–Sidak). However, if the data were not normally
distributed or variances were not equal, the nonpara-
metric Friedman RMANOVA and the post-hoc Turkey
Copyright r 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
(b)
test were used. Finally, the order effect (group A or B)
was analysed using a two-factor RMANOVA [(group A or
group B) and (REAL + RAT or SHAM + RAT)].
Results
Twenty patients completed both sessions with full col-
laboration and no adverse event. One patient dropped
out; his data were not taken into account for the analyses,
which were therefore performed on 20 participants.
Patient characteristics and results are shown in Tables 1
and 2, respectively.
Group comparison
At baseline, both groups (A and B) were similar for age,
time after stroke, Mini-Mental State Examination and
UL impairments assessed by SIAS, modified Rankin
Scale, BBT, PPT and kinematics (P > 0.05) (Table 1).
Baseline comparisons
There was no significant difference between both base-
lines obtained before each assessment (P > 0.05) (data
not shown); thus, both baselines’ data were averaged.
Combination of robotic-assisted therapy and real-
transcranial direct current stimulation or sham-
transcranial direct current stimulation
Table 2 shows the results of the combination of RAT
with REAL-tDCS or SHAM-tDCS. RMANOVA identi-
fied a significant difference in only two assessments.
Indeed, analyses showed that the BBT (P = 0.008) and
the straightness during the free amplitude task
(P = 0.019) were significantly better after REAL + RAT
than before the intervention. Patients moved on average
1.8 (± 0.7) more blocks. The movement straightness was
Copyright r 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

Table 1 Global characteristics of the stroke patients

Age Time since Type Localization of Dominant Impaired MMSE SIAS mRS BB Order of
Patients nos (years) Sex stroke (months) stroke main lesions upper limb upper limb (/30) (/76) (/5) (blocks/min) interventions

1 63 Male 10 I SC R L 26 56 3 14 R-S
2 59 Male 18 I C R R 28 72 1 28 R-S
3 58 Female 15 I C R R 27 70 2 33 R-S
4 63 Male 6 H SC R R 28 54 4 0 R-S
5 53 Male 11 I C L L 30 41 3 0 S-R
6 60 Male 48 H SC R L 29 57 3 6 R-S
7 65 Male 6 I C R R 29 69 3 32 S-R
8 61 Male 240 H SC R R 28 71 1 26 S-R
9 75 Male 36 H SC R R 26 53 4 0 S-R
10 55 Male 6 I C R L 28 52 2 23 S-R
11 56 Male 24 I SC L R 29 49 4 1 S-R
12 68 Male 43 I C R L 29 64 1 27 S-R
13 50 Female 22 I SC R R 29 36 3 0 S-R
14 50 Female 14 I C R L 27 46 3 20 R-S
15 65 Male 8 I C R L 28 52 4 4 R-S
16a 58 Female 49 I C R L 30 62 3 20 R-S
17 62 Male 144 I C R L 23 69 2 27 S-R
18 36 Female 53 I C R L 29 44 3 0 S-R
19 62 Male 6 I C R R 28 64 1 32 R-S

Robotic and tDCS in stroke rehabilitation Dehem et al. 5

20 81 Male 6 H C R R 29 65 2 40 R-S
21 71 Female 12 H C R R 29 63 2 9 R-S

Mean (SD) Male/female (n) Mean (SD) I/H (n) SC/C (n) R/L (n) R/L (n) Median (Q1–Q3) Median (Q1–Q3) Median (Q1–Q3) Mean (SD) R-S/S-R (n)

Total 60.5 (9.5) 15/6 38.6 (57.0) 15/6 7/14 19/2 11/10 28 (28–29) 57 (52–65) 3 (2–3) 16.3 (13.7) 11/10
Total R-S group 62.73 (8.0) 7/4 17.47 (15.8) 7/4 8/3 11/0 6/5 28 (27–29) 62 (54–65) 3 (2–3) 18.73 (13.3) –
Total S-R group 58.1 (10.8) 8/2 58.5 (75.5) 8/2 6/4 8/2 5/5 29 (28–29) 52.5 (44–69) 3 (2–3) 13.6 (14.3) –

C, cortical; H, haemorrhage; I, ischaemia; L, left; MMSE, mini mental state examination; mRS, modified Rankin Scale; Q1, first quartile; Q3, third quartile; R, right; R-S, real dual-tDCS then sham dual-tDCS; SC, subcortical;
SIAS, Stroke Impairment Assessment Set; S-R, sham dual-tDCS, then real dual-tDCS; tDCS, transcranial direct current stimulation.
a
Drop out patient.
 6 International Journal of Rehabilitation Research 2018, Vol 00 No 00

Table 2 Outcomes of robotic-assisted therapy combined with REAL-transcranial direct current stimulation or SHAM-transcranial direct
current stimulation

RMANOVA between REAL + RAT

Baseline (N = 20) Post-RAT + REAL intervention Post-RAT + SHAM intervention and SHAM + RAT intervention
[mean (SD)] (N = 20) [mean (SD)] (N = 20) [mean (SD)] (P value)

Kinematic indices
Free amplitude
Amplitude (cm) 39.90 (29.22–45.16)a 42.48 (31.42–45.20)a 44.01 (29.22–45.15)a 0.705
Straightness 0.97 (0.91–0.99)a 0.97 (0.94–0.99)a 0.97 (0.95–0.99)a 0.019
Velocity (cm/s) 12.27 (6.12) 13.53 (6.90) 14.05 (6.14) 0.069
Smoothness 0.38 (0.14) 0.40 (0.15) 0.41 (0.15) 0.080
Target
Accuracy (cm) 1.40 (1.50) 1.32 (1.29) 1.96 (2.52) 0.085
Straightness 0.91 (0.09) 0.95 (0.05) 0.92 (0.09) 0.193
Velocity (cm/s) 9.47 (3.10) 10.20 (3.29) 10.36 (3.51) 0.116
Smoothness 0.48 (0.07) 0.49 (0.06) 0.50 (0.07) 0.360
Gross manual dexterity
BB (blocks/min) 16.60 (14.32) 18.35 (15.14) 17.90 (16.02) 0.021
Fine manual dexterity
a a a
PPT (Pegs/30 s) 0.97 (0.00–4.07) 1.5 (0.00–4.83) 0.15 (0.00–5.30) 0.115
Robotic assistance
Klat (N/m) – 226.77 (121.14–804.90) a
443.46 (113.39–1707.08) a
0.491
Clong (N s/m) – 14.96 (0.00–141.54)a 18.06 (10.06–107.24)a 0.796
Klat/velocity (N s/m2) – 23.59 (11.76–142.08)a 63.45 (15.80–120.78)a 0.074
Tinit (s) – 0.62 (0.36) 0.62 (0.34) 0.923
Velocity (cm/s) – 11.21 (5.91) 11.29 (5.94) 0.928

Post-hoc analyses Baseline vs. post-RAT + REAL (P value) Baseline vs. post-RAT + SHAM (P value) RAT + REAL vs. RAT + SHAM (P value)

Straightness (free amplitude) 0.020 0.099 0.802

Gross manual dexterity (BB) 0.023 0.084 0.473

BB, Box and Block; Clong, damping coefficient for the longitudinal interaction force; Klat, stiffness coefficient for the lateral interaction force; PPT, Purdue Pegboard Test;
RAT, robotic-assisted therapy; RMANOVA, repeated measure analysis of variance; Tinit, time to initiate the movement.
a
Median (Q1–Q3) of nonparametric test.
P values in bold are significant.

improved by 0.01 (± 0.03) on average (+ 1.0 ± 3.1%).
Other kinematics and fine manual dexterity were not
significantly improved after REAL + RAT (P > 0.05).
The combination of SHAM + RAT did not significantly
alter any parameter (P > 0.05). The analyses did not
objectify any difference between the assessments per-
formed after REAL + RAT and SHAM + RAT (P > 0.05).
The assistance provided by the robot during the rehabi-
litation was equivalent during both interventions
(P > 0.05). There was no difference between the results
of patients from group A and group B (P > 0.05) (data not
shown).
Discussion
Combination of robotic-assisted therapy and real-
transcranial direct current stimulation or sham-
transcranial direct current stimulation
There is currently a lack of evidence to support the use of
tDCS alone in routine clinical practice. Some reviews
have reported a small reduction in UL impairment after a
REAL-tDCS intervention compared with SHAM-tDCS
(Elsner et al., 2013; Pollock et al., 2014). However, the
observed effect is small and transient if not combined
with a rehabilitation intervention (Butler et al., 2013;
Podubecká et al., 2014; Tedesco Triccas et al., 2015a,
2015b). The present study evaluated the effectiveness of
a single session of RAT combined with a REAL-tDCS or
a SHAM-tDCS intervention. Kinematic parameters and
manual dexterity were similar under both conditions. None
of these parameters improved after the SHAM + RAT
intervention. However, after the REAL + RAT interven-
tion, the straightness of movement and the gross manual
dexterity (BBT) were significantly improved. The assistance
provided during RAT was similar during REAL + RAT and
SHAM + RAT interventions.
A study of Giacobbe et al. (2013) analysed the kinematic
effect of administering tDCS before, during and after a
single session of UL RAT. In their study, movement
smoothness improved by 15% when tDCS was adminis-
tered before RAT (Giacobbe et al., 2013). In the present
study, we only observed a small improvement in
straightness after REAL + RAT. The effect was not
observed during the target task and there was no inter-
action between processing time and the intervention.
This lack of (or small) improvement could be because of
the fact that tDCS was administered during RAT,
whereas Giacobbe et al. (2013) recommended adminis-
tration of tDCS before RAT.
In terms of manual dexterity, the study of Lefebvre et al.
(2014) observed an improvement in fine manual dexter-
ity assessed by PPT after a single session of tDCS. The
maximal improvement was observed 20 min after the end
of the stimulation (+ 2.1 Pegs, + 38% after REAL-tDCS

Copyright r 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

vs. 0.4 Pegs, + 5% after SHAM-tDCS) (Lefebvre et al.,
2014). This improvement was not observed in the pre-
sent study. Because of the low number of studies on a
single session of tDCS, the gross manual dexterity results
from the present study were compared with studies
including multiple sessions of tDCS. In the study of
Huyun-Kyu et al. (2014) the experimental group received
20 min of tDCS combined with basic training for func-
tional improvement 5 days per week for 4 weeks. They
observed a greater improvement in BBT in the experi-
mental group (REAL-tDCS) (mean improvement = 26.3
blocks) than in the control group (SHAM-tDCS) (mean
improvement = 9.1 blocks) (P < 0.05) (Huyun-Kyu et al.,
2014). An improvement in gross manual dexterity was
also observed in our study: patients moved significantly
more blocks after the REAL + RAT. However, the mean
improvement was 1.8 blocks, which is less than the
minimal detectable change (six blocks/min) (Chen et al.,
2009). Furthermore, the improvement in the BBT after
the REAL + RAT intervention was not significantly
different from that observed after SHAM + RAT.
Therefore, the improvement in BBT after REAL + RAT
in the present study should be interpreted with caution.
The assistance provided by the robot is a good reflection
of the patients’ motor skill control. It was expected that
the patient would improve his/her online motor skill
learning and would need less assistance during
REAL + RAT as observed in the study by Lefebvre et al.
(2013). However, the robotic assistance provided was
similar during REAL and SHAM interventions, sug-
gesting that motor control was not significantly better
during REAL + RAT. Future studies should administer
tDCS stimulations during more complicated tasks or
tasks involving cognition. Indeed, one could hypothesize
that tasks involving more motor, cognitive and attentional
resources would reinforce learning to a greater extent.
Study limitations
The present study only included 20 participants. One
patient dropped out of the study because of a headache
after the first session (REAL + RAT), which was already
reported in a previous study (Tedesco Triccas et al.,
2015a, 2015b). This is a minor side effect of tDCS.
Conclusion
This study showed a slight motor improvement for the
paretic UL after a single session of REAL + RAT tDCS
treatment. The observed improvement in gross manual
dexterity and movement straightness is not large enough
to support the use of RAT combined with tDCS in
routine clinical practice. A future study should analyse
the combination of RAT + tDCS when tDCS is admi-
nistered before the RAT as suggested by Giacobbe et al.
(2013). Future studies should also analyse RAT + tDCS
with more complicated tasks or tasks involving cognition.
Copyright r 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Robotic and tDCS in stroke rehabilitation Dehem et al. 7
Acknowledgements
This work was supported by the Région Wallonne, the
Fondation Saint-Luc and the Fondation Motrice. The
work of Y.V. was supported by the following grants:
Fonds de la Recherche Scientifique (F.R.S.-FNRS) Post-
Doctorate Clinical Master Specialist (SPD) 1.R.506.16
2015–2017, Fonds Spécial de Recherche (FSR) grant
from the Université Catholique de Louvain 2016,
Fondation Mont-Godinne 2014–2016.
The authors thank Axinesis (Wavre, Belgique) for the
development of the rehabilitation robot REAplan and
Fishing Cactus (Mons, Belgique) for the development of
the serious game. They would also thank all the patients
for their participation in this study.
Conflicts of interest
There are no conflicts of interest.
References
Brown RG, Jahanshahi M (1998). An unusual enhancement of motor performance
during bimanual movement in Parkinson’s disease. J Neurol Neurosurg
Psychiatry 64:813–816.
Butler AJ, Shuster M, O’Hara E, Hurley K, Middlebrooks D, Guilkey K, et al. (2013).
A meta-analysis of the efficacy of anodal transcranial direct current stimulation
for upper limb motor recovery in stroke survivors. J Hand Ther 26:162–171.
Chen HM, Chen CC, Hsueh IP, Huang SL, Hsieh CL (2009). Test-retest repro-
ducibility and smallest real difference of 5 hand function tests in patients
with stroke. Neurorehabil Neural Repair 23:435–440.
Chino N, Sonoda S, Domen K, Saitoh E, Kimura A (1996). Stroke impairment
assessment set (SIAS). In: Chino N, Melvin JL, editors. Functional evaluation
of stroke patients. Springer, Tokyo.
Daly JJ, Hogan N, Perepezko EM, Krebs HI, Rogers JM, Goyal KS, et al. (2005).
Response to upper-limb robotics and functional neuromuscular stimulation
following stroke. J Rehabil Res Dev 42:723–736.
Elsner B, Kugler J, Pohl M, Mehrholz J (2013). Transcranial direct current stimu-
lation (tDCS) for improving function and activities of daily living in patients
after stroke. Cochrane Database Syst Rev 11:CD009645.
Finley M, Fasoli SE, Dipietro L, Ohlhoff J, Macclellan L, Meister C, et al. (2005).
Short-duration robotic therapy in stroke patients with severe upper-limb motor
impairment. J Rehabil Res Dev 42:683–692.
Galinski D (2014). Conception et optimisation d’un robot de rééducation neuro-
motrice du membre supérieur avec compensation active de la gravité. PhD
thesis in engineering sciences, Université catholique de Louvain, Louvain-La-
Neuve (UCL-IMMC) [Design and optimisation of a neuromotor rehabilitation
robot of the upper limb with an active gravity compensation. PhD thesis in
engineering sciences, UniversitÉ catholique de Louvain, Louvain-La-Neuve
(UCL-IMMC)].
Giacobbe V, Krebs HI, Volpe BT, Pascual-Leone A, Rykman A, Zeiarati G (2013).
Transcranial direct current stimulation (tDCS) and robotic practice in chronic
stroke: the dimension of timing. NeuroRehabilitation 33:49–56.
Gilliaux M, Renders A, Dispa D, Holvoet D, Sapin J, Dehez B, et al. (2014a). Upper
limb robot-assisted therapy in cerebral palsy: a single-blind randomized
controlled trial. Neurorehabil Neural Repair 29:183–192.
Gilliaux M, Lejeune TM, Detrembleur C, Sapin J, Dehez B, Selves C, Stoquart G
(2014b). Using the robotic device reaplan as a valid, reliable, and sensitive
tool to quantify upper limb impairments in stroke patients. J Rehabil Med
46:117–125.
Grefkes C, Fink GR (2012). Disruption of motor network connectivity post-stroke
and its noninvasive neuromodulation. Curr Opin Neurol 25:1.
Harvey RL, Stinear JW (2010). Cortical stimulation as an adjuvant to upper limb
rehabilitation after stroke. PM R 2:269–278.
Hesse S, Waldner A, Mehrholz J, Tomelleri C, Pohl M, Werner C, et al. (2011).
Combined transcranial direct current stimulation and robot-assisted arm
training in subacute stroke patients: an exploratory, randomized
multicenter trial. Neurorehabil Neural Repair 25:838–846.
Huyun-Kyu C, Sang-Goo J, Myoung-Kwon K, Jong-Sung C (2014). Effect of
transcranial direct current stimulation of function in patients with stroke.
J Phys Ther Sci 26:363–365.
 8 International Journal of Rehabilitation Research 2018, Vol 00 No 00
Kahn LE, Zygman ML, Rymer WZ, Reinkensmeyer DJ (2006). Robot-assisted
reaching exercise promotes arm movement recovery in chronic hemiparetic
stroke: a randomized controlled pilot study. J Neuroeng Rehabil 3:1–13.
Kandel M, Beis JM, Le Chapelain L, Guesdon H, Paysant J (2012). Non-invasive
cerebral stimulation for the upper limb rehabilitation after stroke: a review. Ann
Phys Rehabil Med 55:657–680.
Krebs HI, Palazzolo JJ, Dipietro L, Ferraro M, Krol J, Rannekleiv K, et al. (2003).
Rehabilitation robotics: performance-based progressive robot-assisted ther-
apy. Auton Robots 15:7–20.
Langhorne P, Bernhardt J, Kwakkel G (2011). Stroke care 2 : stroke rehabilitation.
Lancet 377:1693–1702.
Lefebvre S, Laloux P, Peeters A, Desfontaines P, Jamart J, Vandermeeren Y, et al.
(2013). Dual-tDCS enhances online motor skill learning and long-term
retention in chronic stroke patients. Front Hum Neurosci 6:343.
Lefebvre S, Thonnard JL, Laloux P, Peeters A, Jamart J, Vandermeeren Y, et al.
(2014). Single session of dual-tDCS transiently improves precision grip and
dexterity of the paretic hand after stroke. Neurorehabil Neural Repair 28:1–11.
Mathiowetz V, Volland G, Kashman N, Weber K (1985). Adult norms for the Box
and Block Test of manual dexterity. Am J Occup Ther 39:386–391.
Mehrholz J, Pohl M, Platz T, Kugler J, Elsner B (2012). Electromechanical and
robot-assisted arm training for improving generic activities of daily living, arm
function, and arm muscle strength after stroke (review). Cochrane Database
Syst Rev 6:1–65.
Mehrholz J, Pohl M, Platz T, Kugler J, Elsner B (2015). Electromechanical and
robot-assisted arm training for improving activities of daily living , arm function ,
and arm muscle strength after stroke (review). Cochrane Database Syst Rev
11:1–126.
Moher D, Hopewell S, Schulz KF, Montori V, Gøtzsche PC, Devereaux PJ, et al.
(2010). Consort 2010 explanation and elaboration: updated guidelines for
reporting parallel group randomised trials. J Clin Epidemiol 63:1–37.
Norouzi-Gheidari N, Archambault PS, Fung J (2012). Effects of robot-assisted
therapy on stroke rehabilitation in upper limbs: systematic review and meta-
analysis of the literature. J Rehabil Res Dev 49:479–496.
Ochi M, Saeki S, Oda T, Matsushima Y, Hachisuka K (2013). Effects of anodal
and cathodal transcranial direct current stimulation combined with robotic
Copyright r 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
therapy on severely affected arms in chronic stroke patients. J Rehabil Med
45:137–140.
Page SJ, Cunningham DA, Plow E, Blazak B (2015). It takes two: noninvasive
brain stimulation combined with neurorehabilitation. Arch Phys Med Rehabil
96:S89–S93.
Pignolo L (2009). Robitics in neuro-rehabilitation. J Rehabil Med 41:955–960.
Podubecká J, Bösl K, Rothhardt S, Verheyden G, Nowak DA (2014). Transcranial
direct current stimulation for motor recovery of upper limb function
after stroke. Neurosci Biobehav Rev 47:245–259.
Pollock A, Farmer S, Brady MC, Langhorne P, Mead GE, Mehrholz J, et al. (2014).
Interventions for improving upper limb function after stroke. Cochrane
Database Syst Rev 11:1–161.
Rohrer B, Fasoli S, Krebs HI, Hughes R, Volpe B, Frontera WR, et al. (2002).
Movement smoothness changes during stroke recovery. J Neurosci
22:8297–8304.
Sapin J, Dehez B, Gilliaux M (2017). Rehabilitation system and method. Patent
No. PCT/EP2016/072676.
Simonetti D, Zollo L, Milighetti S, Miccinilli S, Bravi M, Ranieri F, et al. (2017).
Literature review on the effects of tDCS coupled with robotic therapy in post
stroke upper limb rehabilitation. Front Hum Neurosci 11:268.
Triccas LT, Burridge JH, Hughes AM, Pickering RM, Desikan M, Rothwell JC,
Verheyden G (2015a). Multiple sessions of transcranial direct current stimu-
lation and upper extremity rehabilitation in stroke: a review and meta-analysis.
Clin Neurophysiol 127:946–955.
Triccas LT, Burridge JH, Hughes A, Verheyden G, Desikan M, Rothwell J (2015b).
A double-blinded randomised controlled trial exploring the effect of anodal
transcranial direct current stimulation and uni-lateral robot therapy for the
impaired upper limb in sub-acute and chronic stroke. NeuroRehabilitation
37:181–191.
Tombaugh T, McIntyre N (1992). The mini-mental state examination: a
comprehensive review. J Am Geriatr Soc 40:922–935.
Van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ, van Gijn J (1988).
Interobserver agreement for the assessment of handicap in stroke patients.
Stroke 19:604–607.