Upper
limb motor impairment is one of the most frequent causes of long term
disability following stroke and it is particularly problematic given its
negative impact on Activities of Daily Living (ADL) [1].
Physical therapy and exercise promote the motor recovery after stroke
with consequent regain of function and changes in cortical
reorganization according to residual neuroplasticity [2].
It has been demonstrated that the amount and intensity of practice and
the degree of participation, as well as the task-oriented training, play
a crucial role in positively affecting the neuroplastic changes [3].
Apart the intrinsic ability of providing a high number of specific
practice movements, robot-mediated therapy can be successfully coupled
with virtual reality (VR) technology allowing patients to train in a
more ecological and enriched environment which could give an opportunity
to practice functional movements and everyday activities that are not
or cannot be practiced within the hospital environment [4].
However,
although scientific literature provides supporting evidence of the
efficacy of upper limb robotic treatments after stroke compared to
manual therapy [5, 6],
it is still arguable the achievement of an effective improvement in
terms of regained upper limb function and consequent transfer of
abilities to ADL. One recent, large pragmatic randomized controlled
trial performed with the MIT-Manus robotic gym system [7]
concluded that robot-assisted training did not lead to improvement in
upper limb function in ADLs compared with usual care, measured by ARAT
test. To overcome this potential limit of some robotic rehabilitation
programs, it has been so far hypothesized in literature that robotic
training with exoskeletons, based on three-dimensional spatial,
task-oriented and more naturalistic movements [8], is likely to provide higher benefits in terms of recovery in ADLs and improvement of upper limb function.
In
the scientific literature there are still however not only a limited
number of randomized controlled trials (RCT) concerning robotic therapy
with three-dimensional spatial robotic exoskeletons to support this
hypothesis, but also contrasting evidences. The asymmetry of studies
conducted with End Effector (EE) devices vs Exoskeletons (Exo) is for
example evident from data published in one recent meta-review [5],
where only 3 exoskeletons RCT are reported, of which one based on
passive exoskeleton device only,vs. 11 trials employing EE devices.
One
of the first clinical studies addressing this issue was the randomized
trial conducted with Pneu-WREX in a group 26 patients [9],
wherein three dimensional movement against gravity in the context of
simulated functional tasks that required use of the hand conducted with
assist-as-needed controller robotic training was found to be more
effective than conventional table-top training. According to authors’
the observed results benefits may also have arisen in part due to the
fact that the robot allowed 3D movements that incorporate hand grip and
release, rather than just planar or single-joint movements.
In a large controlled study (77 patients) [10],
the robotic treatment conducted with the ARMin exoskeleton was compared
with the manual physical and occupational therapy, showing that robotic
training enhanced arm motor function more effectively than manual
therapy, as measured by the upper extremity portion of the Fugl-Meyer
scale (FMA-UE).
Also we observed in our previous study [11]
in chronic stroke through instrumental study of the reaching
performance that exoskeleton training produced positive effects in
movement execution, in terms of decreased execution time, improved
movement smoothness and increased active joint ranges of motion.
On
the other side, another recent randomized controlled trial compared End
Effector (EE) and Exoskeleton (Exo) robot therapy in patients with
stroke [12]
after 4 weeks of intervention, suggesting that the EE robot
intervention is better than the Exo robot intervention among chronic
stroke patients with moderate-to-severe impairment of upper extremity
function.
Also within the cross-over study conducted with BONES exoskeleton [13],
patients were assigned with different random order to both single joint
and multiple joint robotic training. The results of the study showed
how multi-joint functional robotic training was not superior to single
joint robotic training for Box and Block Test score (primary outcome)
and for other secondary outcome measurements (FMA, Wolf Motor Function
Test WMFT, Motor Activity Log MAL scales).
So to what extent the
3D nature of therapy robotic assistance provided with exoskeletons can
be a determinant factor for motor recovery?
To provide further
clinical and experimental evidences to answer this question, we have
compared within a randomized controlled clinical trial the effects of a
robotic exoskeleton training in three-dimensional task-oriented
exercises versus an equally intensive program of manual therapy
intervention (1) to assess if the observed motor improvements are
reflected into higher functional outcomes—and so improved transfer of
abilities into ADL—than conventional manual therapy and (2) to
understand how the eventual observed changes can be interpreted in terms
of kinematic measurements automatically extracted by the exoskeleton.
As a second aspect, several clinical studies, including animal ones [14],
support with growing consensus that individualized approach to stroke
rehabilitation, for instance based on stratification of patients into
groups with different probabilities of upper limb recovery, could
enhance the recovery of lost motor function. In this context, the use of
biomarkers plays an important role [15, 16].
Beside neurophysiological and neuroimaging biomarkers, robotic
biomarkers may be a valuable clinical instrument for determining the
effect of a rehabilitation therapy [17].
These robotic biomarkers have the great advantage to be entirely
objective in capturing the quality of movement which can be immediately
provided as an index of the recovery progress [18].
The extraction and the analysis of robotic biomarkers can be used for
both monitoring the ongoing recovery process during treatment and for
investigating the relationship with primary clinical outcome. It is
reasonable to think that as next step robotic biomarkers can be used to
optimize the design of rehabilitation therapies tailored to the need of
individual patients.
In [19]
first, it was demonstrated the high potential of prediction of the
outcome of a therapeutic treatment in stroke, performing an objective
and analytical assessment of motor recovery through the acquisition of
kinesiological and kinetic parameters and finding a prediction,
supported by a statistically significant correlation with clinical
scales, while in [20] it was confirmed the capability of predicting Fugl-Meyer assessment scale through robotic and clinical biomarkers.
Predictive
clinical models of post-stroke motor recovery allow specific early
interventions which is the phase in which the largest treatment effect
can be obtained. Patient-specific prognostic models for monitoring
post-stroke recovery have been developed and validated to assess their
clinical effectiveness [21].
The most reliable predictors of the functional outcome are age and
motor function assessed on clinical scales immediately after the acute
event [22].
The
use of metrics based on biomechanical parameters to estimate movement
capabilities can raise the knowledge about motor recovery mechanisms.
However, because of insufficient validation, the clinical integration of
those methods is still limited.
So based on the characteristics
of spatial movement involved in exoskeleton rehabilitation, a secondary
goal of the study is to investigate whether in the robotic group the
measured robotic performance biomarkers, based on patient’s performance
automatically extracted at the enrollment of treatment, could predict
the clinical and functional outcome of the robotic rehabilitation
treatment.
More at link.
