Motion analysis: an useful tool to evaluate the effects of rehabilitation robotics in patients with stroke
Over the last years, the introduction of robotic technologies in rehabilitation of stroke patients has garnered an increasingly strong interest. Robots could be more effective than conventional treatment at improving lower and upper limb motor function, since they raise the degree of motivation and participation among patients and provide a more intensive treatment (in terms of the amount of repetitions required). In fact, robotic devices allow a more symmetric and assisted gait training and a standardized-assisted upper limb training.
Several studies investigated the effects of robot-assisted training in restoring upper limb function and gait ability, compared to conventional treatments, whose results are summarized in two meta-analyses (Mehrholz J et al. 2015 and 2017, respectively).
In these papers, clinical scales are usually employed; however, it is worth noting that movement analysis can provide additional and quantitative information about the changes obtained after the treatment, also in terms of movement patterns.
For this reason, we investigated the effects of robotic interventions both for lower and upper limb, comparing them to those obtained using a traditional approach.
With respect to the robotic gait training, we investigated the effects of the use of end-effect robots both in subacute and chronic stroke patients, by using clinical outcomes and movement analysis.
Our results show that in chronic stroke robotic gait training, combined with conventional physiotherapy, improves functional and motor outcomes to a greater extent than conventional gait rehabilitation alone.
Although neither rehabilitation treatment changes the compensatory strategies in chronic patients, the robotic gait training provides a more intensive and controlled gait training, thereby increasing gait endurance and reducing lower limb spasticity.
In fact, training based on the robotic device offers the patient a more intensive, repetitive and automatic form of exercise that more closely reflects the characteristics of ambulation.
In brief, robot-assisted training improves ambulation, reduces muscle tone and muscle stiffness and increases muscle recruitment. Instead, in subacute stroke patients both rehabilitation treatments produce promising effects on functional and motor outcomes, but in robotic groups more physiological and symmetric gait strategies were observed after treatment.
We suppose that, in subacute stroke patients, robotic therapy, by offering a more intensive and controlled gait training, is able to promote a positive functional and structural-anatomical cortical reorganization, due to the higher neuroplasticity of the brain in this phase after stroke.
With respect to the upper limb, movement analysis can provide a great support to quantitatively measure the effect of a robotic treatment, and to tailor the treatment itself to the patient’s needs and residual ability.
In fact, even if the indices provided by the robotic devices are demonstrated to be sensitive, reliable and well correlated with the clinical scales, they often are limited to the end-effector (i.e., the hand); in addition, the evaluation tasks are similar to those used during the therapy, with a consequent lack of information of the acquired ability to generalize the motor improvement to different tasks.
Instead, advanced rehabilitation strategies of the upper limb in stroke patients should focus on the recovery of the most important daily activities and, therefore, it is crucial to evaluate patient’s improvement during such functional tasks.
In a previous work we had analyzed quantitatively the motor strategies employed by stroke patients when reaching and drinking from a glass, by using an ad-hoc kinematic protocol.
This task is paradigmatic since it includes two relevant phases: a first open-chain movement (reaching for the glass) and a second movement (bringing the glass to the mouth) where a strict coordination between upper limb, trunk, and head movements is required, thus challenging the motor control system even in the presence of UL minimal disability.
Comparing stroke patients and healthy subjects, we highlighted the compensatory strategies adopted by patients to perform the investigated task, in terms of reduced arm elongation and trunk axial rotation, and increased trunk forward inclination and head movements; moreover, an increment of the time required to perform the task and a reduction of the smoothness were also observed in stroke patients, when compared to healthy subjects.
After that, we used the same protocol to investigate the effect of a robotic treatment, using a robotic device mainly focused on shoulder movements, in a group of stroke patients.
After the treatment, we found that patients increased the general quality of the movement (as measured by the time required to perform the task and its smoothness), while their motor strategies did not change significantly, even if a trend toward an improvement was detected.
This led us to suppose that, given the complexity of the upper limb, a single robotic device was not able to treat it in a comprehensive way.
To support our hypothesis, we used the same kinematic protocol to evaluate the effects of a robotic treatment, using a set of four robotic and sensor-based devices, each acting on a different plane and/or function, comparing them with those of a conventional treatment.
The results of this study are expected for the end of August.