Stimulating Rehabilitation - A new hybrid rehabilitation device combining robot and functional electrical stimulation for early stroke rehabilitation
Cindy Rikhof is a PhD student in the Department of Biomechatronics and Rehabilitation Technology. (Co)Promotors are prof.dr. J.S. Rietman, dr. E.C. Prinsen and dr. G.B. Prange from the Faculty of Engineering Technology and prof.dr. J.H. Buurke from the Faculty of Electrical Engineering, Mathematics and Computer Science.
Stroke is one of the leading causes of disability and frequently results in hemiparesis of the lower extremity. Robotic technology has the potential to deliver intensive rehabilitation with minimal burden for the physical therapist. However, robotic technology is often limited in the extent of active involvement of the participant. To stimulate active contribution by the participant during robot-supported training it can be combined with Functional Electrical Stimulation (FES). FES can directly stimulate the muscle to provide a muscle contraction. The combination of robot and FES, also called hybrid technology, utilizes the strength of each modality, and counteracts the drawbacks. At present, knowledge about the effectiveness and clinical applicability of hybrid devices for early rehabilitation of the lower extremity is limited. Therefore, the main aim of this thesis is to assess the feasibility, efficacy and clinical applicability of a new prototype of a robot rehabilitation device combined with functional electrical stimulation for early rehabilitation of the lower extremities after a stroke. In order to do this, six studies were set up using ROBERT® an existing robotic device that was expanded with a FES module.
In a systematic review (Chapter 2) it was investigated what hybrid devices are available for the lower extremity rehabilitation and the added value of electrical stimulation to robot therapy based on existing literature. All included studies showed improvement on lower extremity function. Furthermore, 64% of the included studies showed superiority of combined robotic and ES training over robot training alone or conventional therapy.
To optimize the potential effect of combining robot and FES support, emphasizing active contribution by the patients themselves is essential. One way to do this, is by actively initiating FES. Actively-initiated FES can be based on voluntary muscle activity. To initiate FES, a threshold is needed as cut-off point above which FES should be provided. This threshold can be estimated based on the muscle activity of the participant. Chapter 3 investigated four different threshold estimation methods based on muscle activity, to initiate FES, in stroke patients. Evaluation in three sub-acute stroke patients revealed higher success rate (86.8% vs. 75%, 35.6% or 40% for the other methods) and better user experience for the estimation method based on rest EMG plus two times the standard deviation. In subsequent studies, an updated prototype was used, in which an Assist-As-Needed (AAN) approach was implemented combining robotic and FES support, in a way that it provided patient-specific assistance and maximized active involvement of the participant. Furthermore, this AAN algorithm could adapt the assistance on a repetition-by-repetition basis. It was a state machine, which selected no assistance, FES alone, or combined FES and robotic (mechanical) assistance, depending on the capabilities of the participant. The feasibility and efficacy of Assist-As-Needed robot and FES support were tested during knee extension and ankle dorsiflexion movements among healthy volunteers (Chapter 4, n=10) and sub-acute stroke patients (Chapter 5, n=9), in a lab-based setting. Based on the results we concluded that FES and robot support combined via AAN control is efficacious in assisting leg movements in stroke patients.
Training at body function level, for example with ROBERT-SAS®, is needed to increase force, range of motion or coordination as prerequisites for functional ability and functional training. For functional activities, such as walking and standing, some level of muscle force is necessary to execute te task. In order to potentially increase force, an important factor is to gain a better understanding of the relation between body function level and activity level. Therefore, in Chapter 6 the relation between voluntarily generated isometric force and functional ability, by means of sit-to-stand transfer and gait velocity was evaluated. The results showed a strong correlation between ankle plantarflexion force and gait velocity and moderate correlations of ankle dorsiflexion force and knee flexion force with gait velocity. These results suggest that there is a relation between voluntarily generated force and functional ability, especially in terms of gait velocity. Whether this translates to improved functional ability with improved strength after training with ROBERT-SAS®, cannot be discerned within the scope of this thesis.
Before time-consuming and expensive randomized clinical trials are conducted, knowledge about potential clinical applicability can be gathered to explore feasibility, usability and potential added value in clinic practice. Therefore, in Chapter 7 the clinical applicability of ROBERT-SAS® training in a clinical setting for early rehabilitation after stroke was evaluated. During these training sessions patients were enthusiastic, considering positive responses on a questionnaire about user experience. From a physiotherapist’s view the training has no added value to the current rehabilitation program. The main disadvantage was the time needed to set-up the device. These results suggest that robot- and FES-supported FES training is feasible to implement in early stroke rehabilitation. However, future research should focus on optimal implementing in other workflows and optimizing set-up time. A hybrid device, such as ROBERT-SAS®, has the potential to increase active movements for other target groups, by providing AAN-controlled lower extremity training at bedside.
