Stroke: robot-assisted gait training (FS-07)

Focused symposium
Chair/speaker
Speakers
MAXIMIZE THE EFFECTIVENESS OF ROBOT-ASSISTED GAIT TRAINING AFTER STROKE

Swinnen E1Mehrholz J2Lefeber N1Buurke J3Tamburella F4 
1Vrije Universiteit Brussel, Rehabilitation Research - Neurological Rehabilitation, Brussels, Belgium, 2Wissenschaftliches Institut, Klinik Bavaria in Kreischa GmbH, Kreischa, Germany, 3University of Twente, Rehabilitation Technology Research, Roessingh Research & Development (RRD), Enschede, Netherlands, 4Santa Lucia Foundation, SPInal REhabilitation Lab (SPIRE) and Laboratory of Robotics Applied to Neurological Rehabilitation (NeuroRobot Lab), Neurological and Spinal Cord Injury Rehabilitation Department 1, Rome, Italy

Learning objective 1: to extend the participants knowledge about the Cochrane evidence of robotic devices for gait rehabilitation after stoke
Learning objective 2: to increase the participants insight in how to maximize the effectiveness of robotic devices based on results of cardiorespiratory and biomechanical parameters
Learning objective 3: to extend the participants insight in robotic device selection (e.g. wearable exoskeletons, treadmill based exoskeletons, end-effectors) and customization procedures (e.g. level of assistance) according to residual functional abilities of each individual.
Description: At the end of last century, growing data on brain plasticity forced critical changes in the neurological rehabilitation approaches to people after stroke. Traditionally, rehabilitation possibilities after brain damage were considered minimal and thus all efforts were directed to provide ways to substitute for a lost function. Nowadays, brain plasticity is a fact. The introduction of many different technology-based devices into clinical rehabilitation represents one of the main novelties in the field(1). 
Current evidence - recent high-quality Cochrane reviews with meta-analysis, network-meta-analysis and meta-regression of randomized controlled - show promising results in favor of robotic gait training(2). For instance, people after stroke who receive electromechanical-assisted gait training in combination with physiotherapy are more likely to achieve independent walking than people who receive gait training without these devices. Further, clinical examples include evidence for improving activities of daily living, community ambulation and specific gait parameters, such as walking speed and walking capacity. However, one size does not fit all: physiotherapy should be adjusted to the individual needs and characteristics of patients. So the question "How to increase the effectiveness of robot-assisted gait training after stroke" arises. 
One aspect is the optimal training intensity during gait rehabilitation with robotic devices(3). Aerobic exercise is recommended in several guidelines, but its implementation often meets barriers - including individuals´ physical impairment and fatigue, and lack of therapeutic staff. Robotic devices tackle these barriers by reducing the number of required therapists and enabling longer periods of practice for the patient, while addressing large muscle groups. Current evidence - which is limited - suggests that at this point the exercise intensity of robot-assisted gait rehabilitation is below aerobic training recommendations(4). However, adjusting the training parameters (such as duration and robotic guidance) and using the right type of robotic system (treadmill-based versus wearable systems) according to the patients´ disabilities, may play an important role in achieving sufficient exercise intensity.
A second aspect is that many robotic gait-trainers restrict the shift of the pelvis in the coronal plane towards the weight-bearing leg during the double-stance phase - one of the most important movements during walking. The lower extremity powered exoskeleton (LOPES) is a robotic gait trainer that can support mediolateral pelvic movements (5), and can therefore allow or even enforce weight shift in people after stroke. Data from people after stroke and healthy participants, who walked in the LOPES with three different configurations (allowed weight shift, enforced weight shift, and no weight shift), showed that allowing and/or supporting more degrees of freedom of the pelvis in robotic gait trainers, resembles a more normal gait pattern than fixating the pelvis.
A third aspect is the use of control algorithms in robotic devices. In particular, control algorithms that are more flexible or adjustable to the patients´ needs appear to provide better results. Most approaches tend to develop control algorithms to minimize support and guidance according to the concept of "assistance as needed". In spite of the plausibility of the approach, it is common experience that rehabilitation protocols cannot be only based on predetermined support and guidance. Patients´ attention and mental states, differences in motor learning paradigms, resistance vs assistance, control of abnormal postures and synergies are among the factors to be considered. By customizing the information provided to patient and therapist, it is possible to improve treatment efficacy. The possibility to use an exoskeleton for gait rehabilitation or as substitutional device introducing the concept of "assistance as needed", has been tested on people after stroke. Now, the challenge is to merge technology based and traditional rehabilitation approaches to develop clinical friendly technologies easy to implement in multidisciplinary personalized rehabilitation programs.
Implications / Conclusions: Clinicians and researchers should be aware of the many possibilities, but also the limitations of the actual robotic systems. Robotic gait rehabilitation may not be a stand-alone therapy, but should be implemented in the global rehabilitation plan. Beside this, therapists must also have knowledge about the differences in biomechanics, muscle activity and cardiorespiratory load of training with robotic-systems compared to conventional training. This gives them the opportunity to adequately respond on this by implementing or changing aspects in their therapy plan. 
Key-words: 1. Robotics 2. Gait 3. Stroke
Funding acknowledgements: Nina Lefeber is a SB (Strategic Basic Research) PhD fellow funded by the Research Foundation Flanders (FWO), Belgium.
Relevance to physical therapy globally: Gait rehabilitation robots are increasingly being deployed and incorporated in rehabilitation centres as part of the therapy. Physiotherapists should be aware of the evidence, many possibilities, but also the limitations of the robotic systems. They must also have knowledge about the immediate effects of training with robot-systems compared to conventional gait training. This information gives therapists the opportunity to respond on this by implementing or changing aspects in their therapy plan.
Target audience: Mainly researchers, clinicians, engineers and educators interested in robot-assisted gait training for neurorehabilitation, and to a lesser extent policy makers and managers.