Content uploaded by Giovanni Morone
Author content
All content in this area was uploaded by Giovanni Morone on Mar 07, 2014
Content may be subject to copyright.
The value of robotic systems
in stroke rehabilitation
Expert Rev. Med. Devices Early online, 1–12 (2014)
Stefano Masiero*
1
,
Patrizia Poli
1
,
Giulio Rosati
2
,
Damiano Zanotto
3
,
Marco Iosa
4
,
Sefano Paolucci
4
and
Giovanni Morone
4,5
1
Department of Neuroscience, Unit of
Rehabilitation, University of Padua,
Padua, Italy
2
Department of Innovation in
Mechanics and Management, University
of Padua, Padua, Italy
3
University of Delaware, Newark, DE,
USA
4
Clinical Laboratory of Experimental
Neurorehabilitation, IRCCS Santa Lucia
Foundation, Rome, Italy
5
School of Doctorate, University of
Padua, Padua, Italy
*Author for correspondence:
Tel.: +39 049 821 1270
Fax: +39 821 796
stef.masiero@unipd.it
In this paper, we discuss robot-mediated neurorehabilitation as a significant emerging field in
clinical medicine. Stroke rehabilitation is advancing toward more integrated processes, using
robotics to facilitate this integration. Rehabilitation approaches have tremendous value in
reducing long-term impairments in stroke patients during hospitalization and after discharge,
of which robotic systems are a new modality that can provide more effective rehabilitation.
The function of robotics in rehabilitative interventions has been examined extensively,
generating positive yet not completely satisfactory clinical results. This article presents state-
of-the-art robotic systems and their prospective function in poststroke rehabilitation of the
upper and lower limbs.
KEYWORDS:activities of daily living • motor learning • neurorehabilitation • plasticity • robot-assisted training • stroke
Stroke that is caused by an ischemic or hemor-
rhagic intracranial vascular event is a leading
cause of disability in the USA and Europe [1].
The WHO estimates that stroke events will
increase by 30% between 2000 and 2025 in
Europe [2]. Hemiparesis and hemiplegia are
the most common outcomes of stroke, leading
to deficits in movement in the limbs that are
contralateral to the side of the brain that is
affected by the stroke. The main clinical char-
acteristics in hemiparetic patients are weakness
of specific muscles, abnormal muscle tone,
abnormal postural adjustments, lack of mobil-
ity, incorrect timing of components within a
pattern, abnormal movement synergies, loss of
interjoint coordination and loss of sensation.
Due to residual arm impairments and inabil-
ity to perform activities of daily living (ADLs),
stroke has a significant social impact: 38% of
severely affected patients experience partial
recovery of dexterity compared with complete
recovery in 11.6% [3]. Regarding the recovery of
mobility, a 2008 study demonstrated that on
discharge from a rehabilitation hospital, approx-
imately half of stroke patients remain in a
wheelchair, <15% can walk inside without aid,
<10% can walk outside and <5% can climb
stairs [4]. Thus, the goal of rehabilitation in post-
stroke subjects is to promote the recovery of lost
function, regain independence and effect early
reintegration into social and domestic life.
The number of people who require rehabili-
tation after stroke is climbing rapidly due to
population aging, [2], and the resulting growth
in related expenses will place increasing pres-
sure on healthcare budgets. For example, in
the USA, nearly 795,000 persons experience a
new or recurrent stroke annually, and the esti-
mated direct and indirect costs of stroke for
2009 were $68.9 billion and $73.7 billion
for 2010, whereas the mean lifetime cost of
ischemic stroke is $140,048 [5]. The poststroke
patient requires continuous medical care
and intensive rehabilitation, often one-on-one
manual interaction with therapists. But, the
demands with regard to organization of time
and budget restrictions render this intensive
rehabilitative program difficult to administer.
Thus, new modalities and technologies are
needed to implement efficacious rehabilitative
treatments. Rehabilitative interventions are
more effective if they ensure early, intensive,
task specific and multisensory stimulation. We
have become aware of the CNS’s capacity to
adapt its structural organization after the
development of a brain lesion, influenced pri-
marily by sensory input, experience and learn-
ing [6,7]. Further, as demonstrated recently in
rats, motor cortex neurons couple sensory
input to multiple-related motor programs dur-
ing learning [8]. Nudo [9] first reported that a
subtotal lesion that is confined to a small por-
tion of the representation of one hand effects
further loss of hand territory in the adjacent,
undamaged cortex in adult squirrel monkeys
and that retraining the skilled hand prevented
www.expert-reviews.com 10.1586/17434440.2014.882766 Ó2014 Informa UK Ltd ISSN 1743-4440 1
Review
Expert Review of Medical Devices Downloaded from informahealthcare.com by 2.192.91.7 on 01/30/14
For personal use only.
the loss of hand territory that was adjacent to the infarct. Thus,
increasing evidence has shown that following stroke, the motor
system is plastic and can be influenced by motor training [9].
The cerebellum has recently been demonstrated to modulate
cortical motor output and function in motor learning [10].
Thus, although it is impossible to repair neural damage by cel-
lular proliferation, partial compensation can be elicited through
these mechanisms and by variations in neural schemes by
unmasking neural pathways and synapses that, although they
are not normally used, can be activated when the dominant
system fails. In this regard, the literature suggests a significant
effect of intensity on motor recovery of rehabilitation to people
affected by stroke. Thus, well-defined training methods, espe-
cially those with intensive multisensory stimulation, might
induce neural adaptations and enhance motor and functional
recovery of paretic upper extremities, based on such plastic
rearrangements.
The idea for using automatic devices was conceived from
this basis to help therapists increase the intensity of therapies,
generate multisensory stimulation and reduce working costs.
This novel concept dates back to the early 1990s, with a new
family of robotic machines, called haptic interfaces. These
mechanical devices are designed to interact with humans, guid-
ing the upper limb into passive and active-assisted mobilization,
aiding in certain movement tasks by biofeedback systems and
measuring changes in movement kinematics and forces. Thus,
robotic therapy might complement standard poststroke multi-
disciplinary programs.
The function of robotics in poststroke rehabilitation
To re-establish a patient’s independence with regard to ADLs,
hand and arm motor function must be restored. Rehabilitation
is fundamental to mitigating residual motor deficits in stroke
patients during the acute/subacute and chronic phases. Treat-
ments that emphasize intense [11], highly repetitive [12] and
task-oriented [13] movements are believed to be extremely valu-
able in this regard. This type of exercise increases strength,
accuracy and functional use in subjects with stroke-induced
paresis [12,13]. Based on animal models, the key component of
the rehabilitative pathway (e.g., early rehabilitation <30 days
from event) entails a substantial increase in the intensity and
dose of treatments, placing an emphasis on recovery from
impairments, alone and in combined approaches [14].
One such therapeutic approach is robotic technology. The
potential of robotic systems in poststroke rehabilitation is broad
and tremendous. Robotic systems are well suited to perform
intensive, task-oriented motor training of a patient’s limbs
under therapist supervision, as part of an integrated set of reha-
bilitation tools that also include simpler nonrobotic approaches
[15]. Robots enhance traditional poststroke treatment by specifi-
cally providing therapy for long periods consistently and pre-
cisely without fatigue. They are programmed to perform
various functional modes and have many automated functions;
they can also measure and record a range of behaviors in paral-
lel with therapeutic applications [16].
Nevertheless, organizationally, the parity of robotics-based
and physical therapy is positive – at the very least, the intro-
duction of robotic systems into clinical practice promotes cost-
effective use of human resources and standardization of rehabil-
itation programs. The labor-intensive aspects of physical reha-
bilitation can be alleviated, allowing the physiotherapist to
concentrate on functional rehabilitation during individual train-
ing and supervising patients simultaneously during robot-
assisted therapy sessions. This approach would exploit the
expertise and time of physiotherapists, improving the efficacy
and efficiency of the rehabilitation program [17].
Wagner et al.[18] performed an economic analysis of robot-
assisted therapy for long-term upper limb impairments after
stroke compared with intensive traditional therapy and standard
care. At 36 weeks postrandomization, the total costs were com-
parable between the three groups, and changes in quality of life
were modest and did not differ significantly.
However, robotic rehabilitation does not merely increase the
amount and intensity of the therapy. Robotic systems produce
simple and repetitive stereotyped movement patterns, as do
most existing devices, but they also effect more complex, con-
trolled multisensory stimulation of the patient. It appears that
the impact of rehabilitative technology on functional outcomes
could be optimized by offering more opportunities for the ner-
vous system to experience actual activity-related sensorimotor
input during upper limb training [9].
Thus, greater interaction and stimulation can be elicited
with respect to what is usually experienced during hand-over-
hand therapy. Extrinsic feedback can be also used to inform
the patient of his/her results and performance during robotic
training, facilitating achievement of the goal movement and
engaging the subject in the rehabilitation exercise [19].An
important objective of robotic rehabilitation is to provide a
therapy that increases a patient’s function, activity and partici-
pation. Although people who develop robotic tools accept that
task-oriented approaches are beneficial and highlight them as
the trend along which future technologies should advance [20],
most rehabilitation systems support training methods that are
not based on task-oriented exercises, particularly robots for
upper limb functioning recovery, whereas those for lower limbs
are focused primarily on gait with few exceptions (such climb-
ing up and down stairs for the G-EO system, as described
below). Only T-Wrex, ADLER, TheraDrive, ARMin and
AutoCITE provide task-oriented training for the upper
extremity.
Clinical trials have demonstrated that training with robotics
improves short-term and long-term strength and analytical
upper limb movements in stroke patients substantially, and
results are pending with regard to clinical studies on robotics in
task-oriented training. However, there is experimental evidence
that robotic upper limb training fails to transfer improvements
to the activity level [20,21].
Environmentally, contextual training is likely to improve
functional outcomes. ADLER is a robot that permits a subject
to interact with the environment, leaving the hands free to
Review Masiero, Poli, Rosati et al.
doi: 10.1586/17434440.2014.882766 Expert Rev. Med. Devices
Expert Review of Medical Devices Downloaded from informahealthcare.com by 2.192.91.7 on 01/30/14
For personal use only.
manipulate objects [22]. This feature is missing from systems in
which forces are applied to a handgrip such as T-WREX [21].
To provide realistic sensorimotor input and encourage task-
related problem solving, Edmans et al. proposed that robotic
systems use mixed reality systems in which movement-sensitive
objects and machine vision create a virtual reality environment
that is steered by actual manipulation of objects [23].
Unfortunately, most existing robotic devices for neurorehabili-
tation are designed and programmed to generate simple stereo-
typed movement patterns, which are often unrelated to
functional activity; further, virtual reality, typically used instead
of mixed reality and environmentally contextual training, can
hinder the transfer of training results to everyday situations.
Moreover, the time and effort that are spent by patients to learn
how to interact with a robotic/virtual environment can decrease
the efficacy of robot-assisted training with respect to the manipu-
lation of real objects. All of these devices include gaming aspects,
drawing attention to and involving the patient in what is happen-
ing during the training. Encouraging the patient to take an active
role can also attract his attention: technology should offer exer-
cises that approximate those with which the patient prefers to
train [24]. Few applications offer enough variability in the exer-
cises to set goals according to individual needs.
There is robust evidence that specific and difficult goals can
improve patient performance [25]. Most robots focus on the
proximal section of the upper limb (Mirror Image Movement
Enhancer [MIME], T-Wrex and Arm-Guide), whereas Master
II trains the hand and fingers exclusively. Only ADLER trains
the entire arm (shoulder, elbow, forearm, wrist) and hand. In
daily life, we use the entire arm, which might explain the lack
of efficacy at the functional level. Krebs suggests creating a
‘gym’with several types of robotic systems with which the
patient can train his entire body [26].
It is not possible to exercise over a full range of joint motion
and with all necessary degrees of freedom (DOF) with robot
training, which is a limitation of task-oriented training. Con-
versely, most robotic systems, such as MIT-Manus, Haptic Mas-
ter and MIME, deliver a patient-tailored and goal-tailored
training load. Actuators can deliver assistance in executing move-
ments when necessary and resistance whenever possible, rendering
robotic systems valuable for training arms and hands in patients
with lower functional levels. Fine-tuned assistance encourages
patients to use all of their abilities to improve their movement.
A potential advantage of robotic systems is that they can
measure several kinematic and dynamic parameters while the
patient’s limb is in motion, allowing for performance-related
indicators (e.g., range of motion, velocity, smoothness) to be
evaluated online and offline. These values can be used to quan-
tify patient progress more objectively with respect to clinical
evaluation scales. Yet, the engineering parameters that have
been proposed are usually related to the robotic hardware and
the type of exercise that are being implemented remain far
from being accepted as a valid alternative to traditional evalua-
tion scales. The acceptance of robotic technology by patients
and physiotherapists might be an issue although there have
been no major concerns over the devices that have been devel-
oped. Moreover, the cultural gap between technology providers,
rehabilitation professionals and end users is shrinking due to
the recent dissemination of knowledge.
The chief goal for rehabilitation of the lower limb in
patients after stroke is to walk independently. As in upper limb
rehabilitation, there are machines that support gait training.
Treadmill training, with and without body weight support, was
recently introduced for the rehabilitation after stroke. To
restore gait, clinicians prefer a repetitive task-specific approach
[27], and greater intensities in walking practice programs (result-
ing in more repetitions) have effected better outcomes for
patients after stroke [28,29]. However, repetitive practice of com-
plex gait cycles by these patients requires specific devices such
as a treadmill, with and without partial body weight support.
Treadmill training has one disadvantage – therapists must
set the paretic limbs and control weight shifts – which might
limit the therapeutic intensity, particularly in more severely
disabled patients.
To decrease the dependence on therapists, automated electro-
mechanical gait machines have been developed. Gait machines
comprise an electromechanical solution with two driven
foot plates that simulate the phases of gait (i.e., the ‘Gait
Trainer’[30]) or a robot-driven exoskeleton orthotic system such
as ‘Lokomat’[31].‘Lokomat’is a computer-controlled robotic
gait orthotic system that guides the patient in which gait train-
ing is automated, following a preprogrammed gait pattern.
Gait training with the ‘Gait Trainer’is also automated.
Electromechanical devices can be used by nonambulatory
patients to practice intensive, highrepetition, complex gait
cycles, which a single therapist can not do alone. Compared
with treadmill training with partial body weight support, these
robotic devices can reduce the effort by therapists, who would
no longer need to set the paretic limbs or assist with trunk
movements [32]. A recent update Cochrane revision of 17 trials
in 837 participants demonstrated that electromechanically
assisted gait training, combined with physiotherapy, accelerates
the recovery of independent walking in patients after stroke.
Further, this review determined the ideal frequency, duration
and time after stroke of electromechanically assisted gait train-
ing. How long these benefits last is unknown and require addi-
tional study [33].
As highlighted by Dobkin and Duncan, clinicians should
not administer robotic therapy routinely to disabled, vulnerable
persons in place of or in addiction to conventional therapy out-
side of a scientific efficacy trial [34].
Robotic devices
Machine for upper limb rehabilitation
Existing upper limb robotic systems can be classified by [35]:
• the upper limb function on which they focus (unilateral or
bilateral shoulder, elbow, wrist or hand movements);
• mechanical characteristics (exoskeleton or operational
machines);
The value of robotic systems in stroke rehabilitation Review
www.expert-reviews.com doi: 10.1586/17434440.2014.882766
Expert Review of Medical Devices Downloaded from informahealthcare.com by 2.192.91.7 on 01/30/14
For personal use only.
• control strategy: robots can be programmed to deliver various
exercises (passive movement, active nonassisted, active assisted
or resistive mode, bimanual exercise).
Existing interactive systems with 1 DOF are useful for
stroke patients with lower functional levels; those with multiple
DOF might be more valuable to patients with lower and
higher functional levels. Several upper limb robotic devices are
available, some of which have been discussed in a recent
Cochrane [36]:
InMotion robot (Massachusetts Institute of Technology,
Mit-Manus): wrist robot with three active DOF, mounted at
the tip of a companion planar robot (MIT-MANUS), allowing
five active DOFs at the shoulder, elbow and wrist. MIT-
MANUS (developed by Krebs et al.) was one of the first
robotic rehabilitation systems for upper limb training after
stroke [37]. It allows 2 DOFs and trains wrist, elbow and shoul-
der movements. The exercise can be performed in passive,
active and interactive training modes. Patients with all levels of
muscle strength can use the system. Visual, tactile and auditory
feedback are provided during the movements [38,39]. This sys-
tem improved motor function of the hemiparetic upper
extremities in acute, subacute and chronic stroke patients [40,41].
MIME: robot manipulator with 6 DOFs; performs unilateral
and bilateral upper limb (shoulder and elbow) training. MIME
has been validated in chronic stroke patients [42].
BI-MANU-TRACK: a 1-DOF system that performs forearm
pronation/supination and wrist flexion/extension. BI-MANU-
TRACK provides bilateral training in passive or active mode
but does not give feedback to the patient. It has been validated
in subacute and chronic stroke patients [43,44].
Robot-mediated therapy system GENTLE/S: 3-DOF robot
manipulator (HapticMaster, FCS Robotics, the Netherlands)
with an extra 3-DOF passive gimbal mechanism (allows for pro-
nation/supination of the elbow and flexion and extension of the
wrist), an exercise table, computer screen, overhead frame and
chair. [45]. Arm function has improved in subacute and chronic
stroke patients using the Haptic Master [46] in clinical trials [47].
Arm robot ARMin: semi-exoskeleton for movement of the
shoulder (3 DOFs), the elbow (1 DOF), forearm (1 DOF)
and wrist (1 DOF), matched with an audiovisual display to
illustrate the movement to the patient. It has been tested in
chronic stroke patients [48].
Assisted Rehabilitation and Measurement Guide (Arm-
Guide): 4-DOF robotic device provides arm reaching therapy
for patients with chronic hemiparesis, giving real-time visual
feedback of the location of the arm to the patient. It has been
tested in the chronic phase [49].
REHAROB Robotic Rehabilitation System for upper limb
motion therapy for the disabled: It comprises two unmodified
industrial robots. It is a 7-DOF system that supplies passive
shoulder and elbow physiotherapy. It has been applied in spastic
hemiparesis [50].
NeReBot (FIGURE 1): a 3-DOF robot based on direct drive wire
actuation. Easily transportable NeReBot provides visual and
auditory feedback to the patient and has been clinically tested
in acute–subacute stroke [51,52].
Many other devices have been tested in chronic and acute–
subacute stroke patients. For example, the Active Joint Brace
[53] is a lightweight exoskeleton robotic brace that is controlled
by surface EMG from the affected elbow flexor and extensor
muscles; it has 1 DOF and does not provide feedback on exer-
cise performance. Another example is T-WREX [20] which
offers training on several activities (e.g., eating, shopping, wash-
ing the arm, making lemonade). Unfortunately, limitations in
the movement of the shoulder (especially rotations) and fore-
arm (no pronation or supination) create a discrepancy between
the functional relevance of the exercise and the actual move-
ment that is performed. In telerehabilitation, UniTherapy [54],
a computer-assisted neurorehabilitation tool for tele-assessment
and telerehabilitation, has been validated in chronic stroke
patients [55], providing visual and auditory cues in response to
success/failure.
There are many robotic arms that have been designed for
rehabilitation but have not been examined clinically in large
samples including the Reha-Slide (FIGURE 2), Reha-Slide Duo [56],
H-O-H, Bimanual Lifting Rehabilitator, ARCMIME, Braccio
di Ferro, Bimanual Handlebar, Driver’s SEAT, Bimanual
Coordinate Training System, Hand Robotic Rehabilitation
Device, APBT (The Rocker) and Able-X [57].
Machines for lower limb rehabilitation
Automated electromechanical gait machines were developed in
the 1990s. They can consist of:
Figure 1. Robotic device for upper limb rehabilitation
(NeReBot).
Review Masiero, Poli, Rosati et al.
doi: 10.1586/17434440.2014.882766 Expert Rev. Med. Devices
Expert Review of Medical Devices Downloaded from informahealthcare.com by 2.192.91.7 on 01/30/14
For personal use only.
• a robot-driven exoskeleton orthosis [58]; for example, Loko-
mat (FIGURE 3).
• an electromechanical solution with two driven foot plates
(End-Effector system), simulating the phases of gait [30];for
example, Gait Trainer (FIGURE 4).
Lokomat [58] is a robotic gait orthosis that is combined with a
harness-supported body weight apparatus and a treadmill. The
principal difference with its use from treadmill training is that
the patient’s legs are guided by the robotic device according to a
preprogrammed gait pattern. A computer-controlled robotic gait
orthosis guides the patient – the gait training is automated.
Gait Trainer is based on a double crank and rocker gear sys-
tem. In contrast to a treadmill, it consists of two foot plates that
are positioned on two bars, rockers and cranks each, which pro-
vide the propulsion. The foot plates symmetrically simulate the
stance and the swing phases of walking while the patient is on
the device [30]. In contrast to treadmill training, the gait training
is automated and supported by an electromechanical solution.
Other similar electromechanical devices include the Haptic-
Walker [30,32], Anklebot [59], ALex [60] and lower extremity pow-
ered exoskeleton [61].
The G-EO System (Reha Technology AG; Olten, Switzer-
land; eo comes from Latin for ‘I walk’) is a gait robot that is
based on the end-effector principle and was designed to mini-
mize the therapeutic effort that is needed to relearn walking
and stair climbing after stroke. The trajectories of the foot
plates and the vertical and horizontal movements of the center
of mass are fully programmable. This device trains wheelchair-
bound subjects through repetitive practice of simulated floor
walking and climbing up and down stairs. The device follows
the HapticWalker, a research prototype with limited clinical
applicability due to its size [62].
WalkTrainer (Swortec SA, Monthey, Switzerland) allows for
overground training, using a motor to follow patient move-
ments and has a parallel robotic structure to control pelvic
motion in 6 DOF [63].
The KineAssist (Kinea Design, Evanston, IL, USA) has a
mobile base and provides partial body weight support and
assistance for movements of the pelvis and torso, leaving the
patient’s legs unobstructed to allow therapists to assist [64].
The hybrid assistive leg is a wearable system that comprises
an exoskeleton that is driven by electric motors [65].
The ambulation-assisting robotic tool for human rehabilita-
tion is an end-effector system in which leg movements are con-
trolled by moving coil forcers [66].
The pelvic assist manipulator allows for movement of the
pelvis in 5 DOF [67].
The lower extremity powered exoskeleton is an exoskeleton
that permits more DOF versus other similar devices, especially
into the frontal plane [61]. The ankle can be controlled depend-
ing on the version. Although this system has been tested in
patients with stroke, there have been no randomized controlled
trials with an experimental group of patients who have been
trained with this device.
Figure 2. Patients performing bilateral mechanical device
with video game feedback (ReaSlide).
Figure 3. Exoskeleton robotic device with treadmill and
weight support: Lokomat, Hocoma, Switzerland.
The value of robotic systems in stroke rehabilitation Review
www.expert-reviews.com doi: 10.1586/17434440.2014.882766
Expert Review of Medical Devices Downloaded from informahealthcare.com by 2.192.91.7 on 01/30/14
For personal use only.
Clinical efficacy of robotic approaches
In 1995, the first robotic system for rehabilitation (the MIT-
Manus) was introduced. Since then, the use of robotics in
poststroke rehabilitation has been examined extensively, gener-
ating positive, yet not fully satisfactory, clinical results.
Despite the potential benefits of robot-mediated movement
training after stroke, the clinical efficacy of this approach is
unknown. Although motor benefits have been noted, there is
no evidence that robotic training brings improvements to the
functional level versus traditional therapy [33,36].
Robotic approaches have disparate responses depending on
rehabilitation problem and patient type (FIGURE 5). Thus, tailor-
ing the robotic approach to user needs (patients, therapist,
clinicians) might improve outcomes.
Lower & upper limb
A recent Cochrane review [36] concluded that electromechani-
cal and robot-assisted arm training improve performance in
ADLs and arm function but not arm muscle strength. More-
over, the improvement in ADLs varies between acute and
subacute stroke patients and those in the chronic phase, sug-
gestingthatearlyroboticinterventionismoreeffectivewith
regard to functional outcomes. However, Lo et al.[68].
reported long-term benefits of intensive rehabilitation (robot-
aided or intensive traditional therapy) in patients with mod-
erate-to-severe impairments – even years after stroke.
Due to the various systems, exercises and treatment protocols
that have been used in studies, the results of reviews and meta-
analyses’ must be interpreted carefully. There is not a single,
fixed treatment protocol – robot-assisted therapy can be added
to or replaced by traditional rehabilitation.
Several groups are studying the function of treatment inten-
sity in upper limb rehabilitation, the preliminary results (pilot
study) of which have shown that higher intensity robotic treat-
ment (assisted forearm and wrist movements) affects greater
improvements in motor ability and functional performance in
stroke patients [69]. Higher intensity of robotic rehabilitation in
acute phase appears to induce more substantial improvements
than lower intensity treatments at discharge, but this difference
does not remain after 6 months [70].
With regard to lower limb rehabilitation, a recent review on
gait-training robotic devices [71] concluded that patients who
received robot-assisted gait training in combination with phys-
iotherapy attained independent walking more easily than
patients who were trained without these devices. However,
clinical trials suggest that manual therapy is more effective
than robotic gait training in the subacute and chronic
phases [72,73]. The implication of findings is that voluntary pos-
tural control is reduced during robot-assisted gait training, due
to the restraint of the pelvis and trunk, coupled with the pas-
sive swing assistance that is provided by the robotic system
employed [72].
Proximal & distal approach
There is a positive trend toward robot-assisted therapy of the
proximal upper limb compared with conventional treatment
modalities with regard to motor recovery, as measured by the
Fugl-Meyer assessment scale, but the performance on ADLs
Bilateral
Unilateral
Controller
of end-
point
Trajectories
Upper
limp
Lower
limp
Exoskeleton
Figure 5. Various robotic therapy approaches for arm and
leg as discussed in the review.
Figure 4. End-effector electromechanical devices for
walking-like training: GTII, Rehastim, Berlin.
Review Masiero, Poli, Rosati et al.
doi: 10.1586/17434440.2014.882766 Expert Rev. Med. Devices
Expert Review of Medical Devices Downloaded from informahealthcare.com by 2.192.91.7 on 01/30/14
For personal use only.
does not improved significantly, as measured by the Functional
Independence Measure scale [36]. The ADL scales do not reflect
recovery of the paretic upper limb properly [36].
In their recent review, Oujamaa et al. claimed that proximal
robot-aided therapy decreases the time that a physiotherapist
spends with the patient and hospitalization time. However,
during rehabilitation of chronic stroke, proximal robot-aided
therapy does not effect any clinically relevant motor improve-
ment in severely impaired subjects – regardless of whether the
movements are performed in unimanual or bimanual, active
aided or counter-resistance mode (InMotion2, MIME,
BATRAC) [74].
Bilateral/monolateral approach
Bilateral movement training, validated in a meta-analysis by
Cauraugh et al.[75], is not clearly superior to or as efficient as
the unilateral mode [76], differences that might be attributed to
poststroke delay, degree of motor impairment, type of bilateral
movement training (proximal or distal, functional or sensori-
motor) and amount of movement repetitions.
In particular, a greater understanding of bihemispheric plas-
ticity after stroke will guide the development of new (robot-
assisted) therapies to enhance learning-dependent neuroplastic-
ity [76,77]. Although a small pilot study by Lum et al. suggested
that unilateral training is superior to bilateral training [42],
larger studies should examine the efficacy of bilateral robotic-
assisted training, with or without auditory or visual cueing, and
feedback that promotes interhemispheric activation of the
limbs. One group recently demonstrated that symmetrical and
bilateral robotic practice, combined with functional task train-
ing, significantly improved motor function, arm activity and
self-perceived bilateral arm ability in 20 patients late after
stroke [78]. There is other positive evidence of unilateral and
bilateral robotic treatment: the former improves upper limb
motor impairments, muscle power and strength at the distal
joints, and the latter is optimal for enhancing proximal muscle
power [79]. Further research is needed to determine the value of
bilateral exercise.
New developments in robotic innovations and capabilities
are expected and will most likely expand their applicability to
other and more specific motor functions.
End-effector/exoskeleton
Based on the end-effector principle, a patient’s extremities
(hands or feet) are placed on a support (such as foot plates)
that imposes specific trajectories. In walking robots, such sup-
ports move the feet, simulating the stance and swing phases
during gait training.
Exoskeletons are outfitted with programmable drives or pas-
sive elements that move the limb joints during training. For
example, in robotic walking training, the ankles, knees and
hips can be controlled by robots during the phases of gait.
In both robot types, the weight support is the conditio sine
qua non, especially for walking training in nonambulatory
patients, for practicing the exercises in intensively and safely.
Weight supports allow more repetitions of standardized move-
ments to be performed in a training session than conventional
therapy, in which body or limb weight is usually supported
manually by the physiotherapist or an aid [80]. Both types of
robotic devices have their advantages and drawback, requiring
one to consider the rationale and relative benefits and disadvan-
tages between devices.
In particular, end-effector walking devices might allow the
patient to extend his knee with more freedom. Also, maintain-
ing his balance might be more demanding, depending on how
the harness is set up and whether the patient holds the hand
rails. With exoskeleton devices, gait cycles might be controlled
more easily. Except for a single case report, no study has com-
pared electromechanical devices for gait rehabilitation directly
in patients with cerebral damage [81].
Regarding arm robot/electromechanical devices, the end-
effector approach is considered when a manipulandum guides
the movement, whereas the shoulder and elbow are not con-
trolled in executing kinematic movements and are merely
weight supported. Arm exoskeleton robots control single joint
movements (e.g., shoulder, elbow, wrist) – not only those at
the end of the arm.
Notably, these approaches train patients disparately, differing
with regard to the constriction and freedom of the patient’s
ability and for this reason, different approach described might
represent best option for a specific kind of patients’
impairment.
Who might benefit more?
As confirmed recently by Cochrane [33], the patient characteris-
tics must be defined to determine who will benefit more from
robotic therapy. Thus, are robotic devices effective for all types
of patients with stroke?
The widespread and correct use of new technologies must
rely on the information regarding the types of patients and the
phase of rehabilitation that will benefit from specific technolo-
gies. Based on this principle, Morone and colleagues reported
that patients with more severe impairments in the motor leg
benefited more from robotic-assisted therapy, in combination
with conventional therapy [82,83], likely because robotic devices
allow increases in intensity of the therapy versus conventional
methods, especially for more impaired patients.
Conversely, patients with greater voluntary motor function
in the affected limb can perform intensive training in conven-
tional therapy. Neurorehabilitators might prefer less constrained
and more variable exercises in these patients. For example, in
ambulant patients, overground walking training is more effec-
tive in improving balance and preventing falls [84]. Patients
might benefit from machines that provide external support
until they recover the ability to walk over ground unsupported.
Robots favor such recovery, allowing for progressive decreases
in external support to match the patient’s level of gait
dependency [83].
The psychological profiles of patients are important in
attaining superior motor outcomes with robot training versus
The value of robotic systems in stroke rehabilitation Review
www.expert-reviews.com doi: 10.1586/17434440.2014.882766
Expert Review of Medical Devices Downloaded from informahealthcare.com by 2.192.91.7 on 01/30/14
For personal use only.
conventional therapy. A recent study implicated anxiety as a
negative prognostic factor for robotic therapy, whereas internal
recovery of a locus of control (e.g., patients who believed them-
selves to be chief causal factor in managing their recovery) was
a positive prognostic factor of better outcomes [85].
Thus, instead of asking ourselves whether robotic devices are
effective in rehabilitation, we should determine who will benefit
more from robotic rehabilitation.
Top-down/bottom-up approach
The ‘bottom-up’approach of certain conventional therapies
and many electromechanical devices and robots has been
suggested to require modification, based on the hypothesis
that moving the limbs passively will enhance the recovery of
limb functions through indirect changes in the central neural
system. It has also been suggested that robots that are
based on a top-down approach be redesigned or used to
increase active participation by the patients during robotic
training [86].
Certain robots that provide walking-like training include a
system that facilitates cognitive involvement by biofeedback. [86].
Following this approach, brain–computer interface (BCI) sys-
tems allow specific measurable neurophysiological signals to be
recorded, decoded and translated into effector actions [87].
Thus, BCIs are a potentially valuable tool that can be com-
bined with robotic devices as part of a top-down approach, as
recently supported by Wang and colleagues, who measured
patient engagement during Lokomat walking training by elec-
troencephalography, providing biofeedback to the patient per
BCI principles [88]. However, this approach is not feasible for
moderate and severely affected patients because it requires sub-
stantial active participation by the patients.
In other top-down approaches, noninvasive brain stimulators
can be combined with robots to enhance rehabilitative out-
comes. There are limited data, however, on the combined use
of transcranial direct current stimulation and robots [89,90], but
the stimulation parameters and the candidates must be
matched [91]. We expect that a top-down approach will be par-
ticularly effective in patients with central damage, such as
stroke, but the type of impairment and side of lesion can influ-
ence rehabilitative outcomes.
Expert commentary
In the light of the rapid development of rehabilitation technol-
ogy, we must understand the value of robotics-based rehabilita-
tion in the optimal rehabilitation program. Our review
confirms the commentary of Johnson [92] – that technology for
upper and lower limb training after stroke needs to align with
the evolution of rehabilitation toward functionally oriented
approaches that influence function, activity and participation
levels. We do not understand how various rehabilitation
approaches contribute to restoration of the CNS after stroke.
Although several rehabilitation technology approaches have
shown promising results in small-scale studies, findings from
large-scale clinical trials are needed.
Robot-mediated neurorehabilitation is a rapidly advancing
field that seeks to use advances in robotics, virtual reality and
haptic interfaces, coupled with neuroscience and rehabilitation
theories, to develop new methods of treating neurological inju-
ries such as stroke, spinal cord injury and traumatic brain
injury. Machine-mediated neurorehabilitation faces many chal-
lenges both in engineering and clinical practice. In the former,
more integrated solutions are needed that administer therapy in
a safe environment for the subject and therapist with high
patient acceptance. New machines for rehabilitation must be
tested clinically, the results of which should be published.
Future research must determine whether ADL tasks, through
technical design or new treatment exercises and protocols, can
actually be enhanced by robotic training. The need to find a
solution that permits easy interaction between the therapist and
patients appears to be mandatory.
Ideally, machine-mediated neurorehabilitation should be
available within several days of the initial brain injury and
thereafter throughout the rehabilitation [93–95]. As suggested in
the Cochrane review by Mehrholz and colleagues [36], robotic
treatment of upper limb impairments is likely to be more effi-
cacious in the acute and subacute phase with regard to func-
tional outcomes in the recovery of ADLs; thus solutions for
early and intensive robotic intervention during patient hospital-
ization should be the focus of future research.
The clinical effects differ when robotic-assisted gait training
is used in the early stages to treat lower limb impairments post-
stroke in mild to moderately impaired patients [72,73,84,96]. How-
ever, early robotic-assisted gait training can facilitate the return
of walking ability in stroke patients with moderate-to-severe
gait impairments [82,83]. In the acute phase of stroke, a patient
will occupy a hospital bed; thus, the initial device must be
operable in such an environment. The equipment should be
designed into a device that is available to a potentially unre-
sponsive patient, whereas later, in the chronic phase, the patient
can visit an inpatient or outpatient rehabilitation gymnasium
to receive more specific treatments for recovery of the limbs.
The option to continue rehabilitation at home is attractive
to a patient, who is keen on returning to familiar surroundings,
and economically sensible for the hospital, which would like
to increase the turnover. The high costs of hospitalization
should encourage shorter hospital stays and outpatient treat-
ment. This model demands a new treatment modality – for
example, machine-mediated therapies with techniques in
telerehabilitation.
Five-year view
The robotic option is growing rapidly but is accompanied by
many unknowns, despite the myriad results from the past cen-
tury [97]. This technique has considerable benefits with regard
to cost and intensification of rehabilitation.
The efficacy of such rehabilitative techniques has not been
demonstrated, in part, due to the difficulty in objectively quan-
tifying inputs and outcomes. Robots can objectively quantify
the amount and nature of multisensory stimuli and measure
Review Masiero, Poli, Rosati et al.
doi: 10.1586/17434440.2014.882766 Expert Rev. Med. Devices
Expert Review of Medical Devices Downloaded from informahealthcare.com by 2.192.91.7 on 01/30/14
For personal use only.
patient outputs, such as times, movements, coordination and
strength improvement, which suits the aims of evidence-based
medicine. Future research should identify the features and
modalities that are essential to the efficacy of robotic manipula-
tion and determine the trajectories, strengths and multisensorial
inputs that robots must provide to improve the quality and
efficiency of rehabilitation.
Successful clinical implementation of this fascinating new
area requires that fears that robots will replace the human work
force and ‘dehumanize’patient rehabilitation be addressed. The
chief objective of these machines is to be adjunctive tools that
increase the intensity of therapies as per the modern principles
of motor rehabilitation. A robot can never replace the multi-
level interactions between a patient and therapist, which take
into account the exchange of glances and the manual ability of
experienced operators.
In poststroke patients with upper arm impairment, these
devices can be applied in the acute, subacute and chronic
phase. Most robotic therapy-based treatment protocols focus on
persons with chronic impairments. Nevertheless, applying this
approach to acute or subacute stroke patients might effect bet-
ter clinical outcomes, primarily because the brain has additional
capacity for plasticity earlier after stroke.
The effectiveness of this approach is supported by a recent
Cochrane review by Mehrholz et al.[36], who concluded that
robotic-assisted training in the acute and subacute phases (i.e.,
within 3 months after stroke), in association with traditional
treatment protocols, has a greater impact on the ADLs of par-
ticipants compared with robotic therapy in the chronic phase
(>3 months after stroke). Other clinical experiences with this
concept in the development of a rehabilitation robot have been
promising [51].
In contrast to other engineering disciplines, the effective-
ness of a robot is often examined after its development and
commercial launch. At this stage, the clinical researcher
determines whether the robot is effective. Thus, determining
for whom this technology is effective instead of whether a
robotic technology is effective, as suggested by Morone
et al.[83,98], might improve robotic efficacy and its use by
rehabilitation teams.
In recent years, disparate robotic systems and approaches
have been used in the rehabilitation of impaired upper and
lower limbs in poststroke patients. Such robots interact with
the patient in real time and can manipulate a powerless limb,
like any hand-over-hand therapy. In robotic therapy, several
principles of motor learning must be considered in choosing
between these systems:
• The modality in which the subjects perform. Brain stimuli
and motor gains are greater in intense, actively assisted
repetitive movements than in unassisted and passive
movements [99].
• The extent and type of feedback (visual, auditory, haptic
feedback) in relation to the degree of active subject move-
ments or to the degree of attention by the patient or active
participation. A virtual reality interface and the use of real
objects in a natural or purposeful context might maximize
attention to the task and enhance motor performance in
individuals with hemiparesis. However, there are little data
on the true relationship between sensory information and
patient engagement and effort, which should be examined to
guide the design of novel robotic systems for rehabilitation.
• The multiplanarity of the exercises that induce greater motor
cortex excitation [35,100,101].
Financial & competing interests disclosure
The authors have no relevant affiliation or financial involvement with
any organization or entity with a financial interest in or financial conflict
with the subject matter or materials discussed in the manuscript. This
includes employment, consultancies, honoraria, stock ownership or options,
expert testimony, grants or patents received or pending, and royalties.
Professional copyediting services courtesy of Blue Pencil Science.
Key issues
• Stroke (cerebrovascular disease) is the leading cause of long-term disability in developed countries.
• There is increasing evidence that the motor system is plastic following stroke and can be influenced by motor training.
• Due to population aging, the number of people who require rehabilitation training after stroke will rise in the coming decades.
• There is a need to reduce the social cost of stroke with regard to residual motor impairments in stroke survivors and direct (inpatient
and treatment) and indirect costs.
• Robotic systems have the potential to improve clinical outcomes of rehabilitation treatments and decrease treatment costs.
• Compared with physical therapy, existing robotic therapy devices confer few benefits with regard to functional outcome.
• Future research should determine whether daily living-related tasks can be enhanced by robotic training through technical design and
new treatment exercises and protocols.
• The use of existing and developing robotic technology in poststroke rehabilitation will continue to grow, and the first at-home
rehabilitation devices are likely to be developed and brought to market.
• Tailoring the robotic approach to user needs (patients, therapist and clinicians) might improve outcomes.
• The robotic technology is used in various domains of rehabilitation, as the large family of neurodegeneration.
The value of robotic systems in stroke rehabilitation Review
www.expert-reviews.com doi: 10.1586/17434440.2014.882766
Expert Review of Medical Devices Downloaded from informahealthcare.com by 2.192.91.7 on 01/30/14
For personal use only.
References
1. Rosamond W, Flegal K, Furie K, et al.
Heart disease and stroke statistics-2007
update: a report from the American heart
association statistics committee and stroke
statistics subcommittee. Circulation 2007;
15(5):69-17
2. WHO. World health statistics. WHO Press;
Geneva, Switzerland: 2008
3. Kwakkel G, Kollen BJ, van der Grond J,
Prevo AJ. Probability of regaining dexterity
in the flaccid upper limb: impact of severity
of paresis and time since onset in acute
stroke. Stroke 2003;34(9):2181-6
4. Paolucci S, Bragoni M, Coiro P, et al.
Quantification of the probability of reaching
mobility independence at discharge from a
rehabilitation hospital in nonwalking early
ischemic stroke patients: a multivariate
study. Cerebrovasc Dis 2008;26(1):16-22
5. Lloyd-Jones D, Adams RJ, Brown TM,
et al. Writing Group Members; American
Heart Association Statistics Committee and
Stroke Statistics Subcommittee. Heart
disease and stroke statistics – 2010 update:
a report from the American Heart
Association. Circulation 2010;121(7):46-215
6. Masiero S, Carraro E. Upper limb
movements and cerebral plasticity in
post-stroke rehabilitation. Aging Clin Exp
Res 2008;20(1):103-8
7. Wolpert DM, Diedrichsen J, Flanagan JR.
Principles of sensorimotor learning. Nature
Rev Neurosci 2011;12(12):739-51
8. Huber D, Gutnisky DA, Peron S, et al.
Multiple dynamic representations in the
motor cortex during sensorimotor learning.
Nature 2012;484(7395):473-8
9. Nudo RJ. Postinfarct cortical plasticity and
behavioral recovery. Stroke 2007;38(2):
845-0
10. Oulad Ben Taib N, Manto M, Laute MA,
Brotchi J. The cerebellum modulates rodent
cortical motor output after ripetitive
somatosensory stimulation. Neurosurgery
2005;56(4):811-20
11. Nelles G. Cortical reorganization – effects
of intensive therapy. Restor Neurol
Neurosci 2004;22(3-5):239-44
12. Butefisch C, Hummelsheim H, Denzler P,
Mauritz KH. Repetitive training of isolated
movements improves the outcome of motor
rehabilitation of the centrally paretic hand. J
Neurol Sci 1995;130(1):59-68
13. Bayona NA, Bitensky J, Salter K, Teasell R.
The role of task-specific training in
rehabilitation therapies. Top Stroke Rehabil
2005;12(3):58-65
14. Krakauer JW, Carmichael ST, Corbett D,
Wittenberg GF. Getting neurorehabilitation
right: what can be learned from animal
models? Neurorehabil Neural Repair 2012;
26(8):923-31
15. Johnson MJ, Feng X, Johnson LM,
Winters JM. Potential of a suite of robot/
computer-assisted motivating systems for
personalized, home-based, stroke
rehabilitation. J Neuroeng Rehabil 2007;4:6
16. Dobkin BH. Strategies for stroke
rehabilitation. Lancet Neurol 2004;3(9):
528-36
17. Kahn LE, Lum PS, Rymer WZ,
Reinkensmeyer DJ. Robot-assisted
movement training for the stroke-impaired
arm: does it matter what the robot does?
J Rehabil Res Dev 2006;43(5):619-30
18. Wagner TH, Lo AC, Peduzzi P, et al. An
economic analysis of robot-assisted therapy
for long-term upper-limb impairment after
stroke. Stroke 2011;42(9):2630-2
19. Van Vliet PM, Wulf G. Extrinsic feedback
for motor learning after stroke: what is the
evidence? Disabil Rehabil 2006;28(13-14):
831-40
20. Prange GB, Jannink MJ,
Groothuis-Oudshoorn CG, et al. Systematic
review of the effect of robot-aided therapy
on recovery of the hemiparetic arm after
stroke. J Rehabil Res Dev 2006;43(2):
171-84
21. Sanchez RJ, Liu J, Rao S, et al. Automating
arm movement training following severe
stroke: functional exercises with quantitative
feedback in a gravity-reduced environment.
IEEE Trans Neural Syst Rehabil Eng 2006;
14(3):378-89
22. Wisneski KJ, Johnson MJ. Quantifying
kinematics of purposeful movements to real,
imagined, or absent functional objects:
implications for modelling trajectories for
robot-assisted ADL tasks. J Neuroeng
Rehabil 2007;4:7
23. Edmans JA, Gladman JR, Walker M, et al.
Mixed reality environments in stroke
rehabilitation: development as rehabilitation
tools. 5th International Conference of
Disability, Virtual Reality & Assoc Tech
(ICDVRAT); Oxford, UK; 2004
24. Timmermans AA, Seelen HA,
Willmann RD, et al. Arm and hand skills:
training preferences after stroke. Disabil
Rehabil 2009;31(16):1344-52
25. Levack WM, Taylor K, Siegert RJ, et al. Is
goal planning in rehabilitation effective?
A systematic review. Clin Rehabil 2006;
20(9):739-55
26. Krebs HI. Robot-mediated movement
therapy: a tool for training and evaluation.
Eur Symp Tech Aids Rehabil TAR
2007;2007
27. French B, Thomas L, Leathley M, et al.
Does repetitive task training improve
functional activity after stroke? A Cochrane
systematic review and meta-analysis.
J Rehabil Med 2010;42(1):14.9
28. Wevers L, van de Port I, Vermue M, et al.
Effects of task-oriented circuit class training
on walking competency after stroke:
a systematic review. Stroke 2009;40(7):
2450-9
29. Van de Port IG, Wood-Dauphinee S,
Lindeman E, Kwakkel G. Effects of exercise
training programs on walking competency
after stroke: a systematic review. Am J Phys
Med Rehabil 2007;86(11):935-51
30. Schmidt H, Werner C, Bernhardt R, et al.
Gait rehabilitation machines based on
programmable footplates. J Neuroeng
Rehabil 2007;4:2
31. Riener R, Lu
¨nenburger L, Jezernik S, et al.
Patient-cooperative strategies for robot-aided
treadmill training: first experimental results.
IEEE Trans Neural Syst Rehabil Eng 2005;
13(3):380-94
32. Hesse S, Mehrholz J, Werner C.
Robot-assisted upper and lower limb
rehabilitation after stroke: walking and
arm/hand function. Dtsch Arztebl Int 2008;
105(18):330-6
33. J, Mehrholz J, Werner C, KuglerPohl M.
Electromechanical-assisted training for
walking after stroke. Cochrane Database
Syst Rev 2007;4:CD006185
34. Dobkin BH, Duncan PW. Should body
weight-supported treadmill training and
robotic-assistive steppers for locomotor
training trot back to the starting gate?
Neurorehabil Neural Repair 2012;26(4):
308-17
35. Masiero S, Carraro E, Ferraro C, et al.
Upper limb rehabilitation robotics after
stroke: a perspective from the University of
Padua, Italy. J Rehabil Med 2009;41(12):
981-5
36. Mehrholz J, Ha
¨drich A, Platz T, et al.
Electromechanical and robot-assisted arm
training for improving generic activities of
daily living, arm function, and arm muscle
strength after stroke. Cochrane Database
Syst Rev 2012;6:CD006876
37. Krebs HI, Hogan N, Aisen ML, Volpe BT.
Robot-aided neurorehabilitation. IEEE
Trans Rehabil Eng 1998;6(1):75-87
38. Reinkensmeyer D, Mahoney R, Rymer WZ,
Burgar C. Robotic devices for movement
Review Masiero, Poli, Rosati et al.
doi: 10.1586/17434440.2014.882766 Expert Rev. Med. Devices
Expert Review of Medical Devices Downloaded from informahealthcare.com by 2.192.91.7 on 01/30/14
For personal use only.
therapy after stroke: current status and
challenges to clinical acceptance. Top Stroke
Rehabil 2002;8(4):40-53
39. Dipietro L, Ferraro M, Palazzolo JJ, et al.
Customized interactive robotic treatment for
stroke: EMG-triggered therapy. IEEE Trans
Neural Syst Rehabil Eng 2005;13(3):325-34
40. Stein J, Krebs HI, Frontera WR, et al.
Comparison of two techniques of
robot-aided upper limb exercise training
after stroke. Am J Phys Med Rehabil 2004;
83(9):720-8
41. Volpe BT, Krebs HI, Hogan N, et al.
A novel approach to stroke rehabilitation:
robot-aided sensorimotor stimulation.
Neurology 2000;54(10):1938-44
42. Lum PS, Burgar CG, Van der Loos M,
et al. MIME robotic device for upper-limb
neurorehabilitation in subacute stroke
subjects: a follow-up study. J Rehabil Res
Dev 2006;43(5):631-42
43. Hesse S, Schulte-Tigges G, Konrad M,
et al. Robot-assisted arm trainer for the
passive and active practice of bilateral
forearm and wrist movements in
hemiparetic subjects. Arch Phys Med
Rehabil 2003;84(6):915-20
44. Hesse S, Werner C, Pohl M, et al.
Computerized arm training improves the
motor control of the severely affected arm
after stroke: a single-blinded randomized
trial in two centers. Stroke 2005;36(9):
1960-6
45. Coote S, Murphy B, Harwin W, Stokes E.
The effect of the GENTLE/s robot-
mediated therapy system on arm function
after stroke. Clin Rehabil 2008;22(5):
395-405
46. Loureiro R, Amirabdollahian F, Topping M,
et al. Upper limb robot mediated stroke
therapy –GENTLE/s approach. Auton
Robots 2003;15(1):35-51
47. Amirabdollahian F, Loureiro R, Gradwell E,
et al. Multivariate analysis of the
Fugl-Meyer outcome measures assessing the
effectiveness of GENTLE/S robotmediated
stroke therapy. J Neuroeng Rehabil
2007;4:4
48. Guidali M, Duschau-Wicke A, Broggi S,
et al. A robotic system to train activities of
daily living in a virtual environment. Med
Biol Eng Comput 2011;49(10):1213-23
49. Kahn LE, Lum PS, Rymer WZ,
Reinkensmeyer DJ. Robot-assisted
movement training for the stroke-impaired
arm: does it matter what the robot does?
J Rehabil Res Dev 2006;43(5):619-30
50. Fazekas G, Horvath M, Toth A. A novel
robot training system designed to
supplement upper limb physiotherapy of
patients with spastic hemiparesis. Int J
Rehabil Res 2006;29(3):251-4
51. Masiero S, Celia A, Rosati G, Armani M.
Robotic-assisted rehabilitation of the upper
limb after acute stroke. Arch Phys Med
Rehabil 2007;88(2):142-9
52. Masiero S, Armani M, Rosati G.
Upper-limb robot-assisted therapy in
rehabilitation of acute stroke patients:
focused review and results of new
randomized controlled trial. J Rehabil Res
Dev 2011;48(4):355-66
53. Stein J, Narendran K, McBean J, et al.
Electromyography-controlled exoskeletal
upper-limb-powered orthosis for exercise
training after stroke. Am J Phys Med
Rehabil 2007;86(4):255-61
54. Johnson MJ, Feng X, Johnson LM,
Winters JM. Potential of a suite of robot/
computer-assisted motivating systems for
personalized, home-based, stroke
rehabilitation. J Neuroeng Rehabil 2007;4:6
55. Johnson MJ, Ramachandran B,
Paranjape RP, Kosasih JB. Feasibility study
of TheraDrive: a low-cost game-based
environment for the delivery of upper arm
stroke therapy. Conf Proc IEEE Eng Med
Biol Soc 2006;1:695-8
56. Buschfort R, Brocke J, Heß A, et al. Arm
studio to intensify upper limb rehabilitation
after stroke: concept, acceptance, utilization
and preliminary clinical results. J Rehabili
Med 2010;42(4):310-14
57. van Delden AL, Peper CL, Kwakkel G,
Beek PJ. A systematic review of bilateral
upper limb training devices for poststroke
rehabilitation. Stroke Res Treat
2012;2012:972069
58. Colombo G, Joerg M, Schreier R, Dietz V.
Treadmill training of paraplegic patients
using a robotic orthosis. J Rehabil Res Dev
2000;37(6):693-700
59. Roy A, Forrester LW, Macko RF.
Short-term ankle motor performance with
ankle robotics training in chronic
hemiparetic stroke. J Rehabil Res Dev 2011;
48(4):417-29
60. Banala S, Agrawal S, Scholz J. Active Leg
Exoskeleton (ALEX) for gait rehabilitation
of motor-impaired patients. In 10th IEEE
International Conference on Rehabilitation
Robotics (ICORR); Noordwijk, The
Netherlands; 2007. p. 401-7
61. Koopman B, van Asseldonk EH,
van der Kooij H. Selective control of gait
subtasks in robotic gait training: foot
clearance support in stroke survivors with a
powered exoskeleton. J Neuroeng Rehabil
2013;10:3
62. Hesse S, Waldner A, Tomelleri C. Research
Innovative gait robot for the repetitive
practice of floor walking and stair climbing
up and down in stroke patients. J Neuroeng
Rehabil 2010;7:30
63. Bouri M, Stauffer Y, Schmitt C, et al. The
WalkTrainer. a robotic system for walking
rehabilitation. In Proceedings of the
2006 IEEE international conference on
robotics and biomimetics, Kunming,
China;2006
64. Peshkin M, Brown D, Santos-Munne
´J,
et al. KineAssist: a robotic overground gait
and balance training device. In 9th IEEE
International Conference on Rehabilitation
Robotics (ICORR); Chicago, USA; 2005.
p. 241-6
65. Kawamoto H, Sankai Y. Power assist system
HAL-3 for gait disorder person. Proceedings
of the 8th International Conference on
Computers Helping People with Special
Needs (ICCHP); Linz, Austria; 2002
66. Reinkensmeyer D, Wynne J, Harkema S.
A robotic tool for studying locomotor
adaptation and rehabilitation. In 24th
Annual Conference and the Annual Fall
Meeting of the Biomedical Engineering
Society (EMBS/BMES); Houston, U.S;
2002. p. 2353-4
67. Aoyagi D, Ichinose W, Harkema S, et al.
An assistive robotic device that can
synchronize to the pelvic motion during
human gait training. In 9th IEEE
International Conference on Rehabilitation
Robotics (ICORR); Chicago, U.S; 2005.
p. 565-8
68. Lo AC, Guarino PD, Richards LG, et al.
Robot-assisted therapy for long-term upper-
limb impairment after stroke. N Engl J
Med 2010;362(19):1772-83
69. Hsieh YW, Wu CY, Liao WW, et al.
Effects of treatment intensity in upper limb
robot-assisted therapy for chronic stroke:
a pilot randomized controlled trial.
Neurorehabil Neural Repair 2011;25(6):
503-11
70. Burgar CG, Lum PS, Scremin AM, et al.
Robot-assisted upper-limb therapy in acute
rehabilitation setting following stroke:
department of veterans affairs multisite
clinical trial. J Rehabil Res Dev 2011;48(4):
445-58
71. Mehrholz J, Pohl M.
Electromechanical-assisted gait training after
stroke: a systematic review comparing
end-effector and exoskeleton devices.
J Rehabil Med 2012;44(3):193-9
The value of robotic systems in stroke rehabilitation Review
www.expert-reviews.com doi: 10.1586/17434440.2014.882766
Expert Review of Medical Devices Downloaded from informahealthcare.com by 2.192.91.7 on 01/30/14
For personal use only.
72. Hornby TG, Campbell DD, Kahn JH,
et al. Enhanced gait-related improvements
after therapist- versus robotic-assisted
locomotor training in subjects with chronic
stroke: a randomized controlled study.
Stroke 2008;39(6):1786-92
73. Hidler J, Nichols D, Pelliccio M, et al.
Multicenter randomized clinical trial
evaluating the effectiveness of the Lokomat
in subacute stroke. Neurorehabil Neural
Repair 2009;23(1):5-13
74. Oujamaa L, Relave I, Froger J, et al.
Rehabilitation of arm function after stroke.
Literature review. Ann Phys Rehabil Med
2009;52(3):269-93
75. Cauraugh JH, Lodha N, Naik SK,
Summers JJ. Bilateral movement training
and stroke motor recovery progress:
a structured review and meta-analysis.
Hum Mov Sci 2010;29(5):853-70
76. Tijs E, Matayas TA. Bilateral training does
not facilitate performance of copying tasks
in post-stroke hemiplegia. Neurorehabil
Neural Repair 2006;20(4):473-83
77. Foley NC, Bhogal SK, Teasell RW, et al.
Estimates of quality and reliability with the
physiotherapy evidence-based database scale
to assess the methodology of randomized
controlled trials of pharmacological and
nonpharmacological interventions. Phys
Ther 2006;86(6):817-24
78. Liao WW, Wu CY, Hsieh YW, et al.
Effects of robot-assisted upper limb
rehabilitation on daily function and
real-world arm activity in patients with
chronic stroke: a randomized controlled
trial. Clin Rehabil 2012;26(2):111-20
79. Yang CL, Lin KC, Chen HC, et al. Pilot
comparative study of unilateral and bilateral
robot-assisted training on upper-extremity
performance in patients with stroke. Am J
Occup Ther 2012;66(2):198-206
80. Iosa M, Morone G, Bragoni M, et al.
Driving electromechanically assisted Gait
Trainer for people with stroke. J Rehabil
Res Dev 2011;48(2):135-46
81. Regnaux JP, Saremi K, Marehbian J, et al.
An accelerometry-based comparison of
2 robotic assistive devices for treadmill
training of gait. Neurorehabil Neural Repair
2008;22(4):348-54
82. Morone G, Bragoni M, Iosa M, et al. Who
may benefit from robotic-assisted gait
training? A randomized clinical trial in
patients with subacute stroke. Neurorehabil
Neural Repair 2011;25(7):636-44
83. Morone G, Iosa M, Bragoni M, et al. Who
may have durable benefit from robotic gait
training?: a 2-year follow-up randomized
controlled trial in patients with subacute
stroke. Stroke 2012;43(4):1140-2
84. Duncan PW, Sullivan KJ, Behrman AL,
et al. LEAPS investigative team.
Body-weight-supported treadmill
rehabilitation after stroke. N Engl J Med
2011;364(21):2026-36
85. Bragoni M, Broccoli M, Iosa M, et al.
Influence of psychologic features on
rehabilitation outcomes in patients with
subacute stroke trained with robotic-aided
walking therapy. Am J Phys Med Rehabil
2013;92(10 Suppl 1):e16-25
86. Belda-Lois JM, Mena-del Horno S,
Bermejo-Bosch I, et al. Rehabilitation of
gait after stroke: a review towards a
top-down approach. J Neuroeng Rehabil
2011;8:66
87. Buch E, Weber C, Cohen LG, et al. Think
to move: a neuromagnetic brain-computer
interface (BCI) system for chronic stroke.
Stroke 2008;39(3):910-17
88. Wagner J, Solis-Escalante T, Grieshofer P,
et al. Level of participation in
robotic-assisted treadmill walking modulates
midline sensorimotor EEG rhythms in
able-bodied subjects. Neuroimage 2012;
63(3):1203-11
89. Hesse S, Werner C, Schonhardt EM, et al.
Combined transcranial direct current
stimulation and robot-assisted arm training
in subacute stroke patients: a pilot study.
Restor Neurol Neurosci 2007;25(1):9-15
90. Edwards DJ, Krebs HI, Rykman A, et al.
Raised corticomotor excitability of
M1 forearm area following anodal tDCS is
sustained during robotic wrist therapy in
chronic stroke. Restor Neurol Neurosci
2009;27(3):199-207
91. Ochi M, Saeki S, Oda T, et al. Effects of
anodal and cathodal transcranial direct
current stimulation combined with robotic
therapy on severely affected arms in chronic
stroke patients. J Rehabil Med 2013;45(2):
137-40
92. Johnson MJ. Recent trends in robot-assisted
therapy environments to improve real-life
functional performance after stroke. J
Neuroeng Rehabil 2006;18:3-29
93. Krakauer JW, Carmichael ST, Corbett D,
Wittenberg GF. Getting neurorehabilitation
right: what can be learned from animal
models? Neurorehabil Neural Repair 2012;
26(8):923-31
94. Nielsen R, Soerensen K, Simonsen D,
Jensen W. Effect of early and late
rehabilitation onset in a chronic rat model
of ischemic stroke-assessment of motor
cortex signaling and gait functionality over
time. IEEE Trans Neural Syst Rehabil Eng
2013;21(6):1006-15
95. Mackay-Lyons M, McDonald A,
Matheson J, et al. Dual effects of
body-weight supported treadmill training on
cardiovascular fitness and walking ability
early after stroke: a randomized controlled
trial. Neurorehabil Neural Repair 2013;
27(7):644-53
96. Husemann B, Mu
¨ller F, Krewer C, et al.
Effects of locomotion training with
assistance of a robot-driven gait orthosis in
hemiparetic patients after stroke:
a randomized controlled pilot study. Stroke
2007;38(2):54.349
97. Aisen ML, Krebs HI, Hogan N, et al. The
effect of robot-assisted therapy and
rehabilitative training on motor recovery
following stroke. Arch Neurol 1997;54(4):6.
443
98. Iosa M, Morone G, Fusco A, et al. Seven
capital devices for the future of stroke
rehabilitation. Stroke Res Treat
2012;2012:187965
99. Takahashi CD, Der-Yeghiaian L, Le V,
et al. Robot-based hand motor therapy after
stroke. Brain 2008;131(2):425-37
100. Masiero S, Celia A, Armani M, et al. Robot-
aided intensive training in post-stroke recovery.
Aging Clin Exp Res 2006;18(3):261-5
101. Masiero S, Celia A, Armani M, Rosati G.
A novel robot device in rehabilitation of
post-stroke hemiplegic upper limbs. Aging
Clin Exp Res 2006;18:531-5
Review Masiero, Poli, Rosati et al.
doi: 10.1586/17434440.2014.882766 Expert Rev. Med. Devices
Expert Review of Medical Devices Downloaded from informahealthcare.com by 2.192.91.7 on 01/30/14
For personal use only.
