The incidence of stroke and other neurological conditions is set to increase over the coming years as the demographics of the European population change. While robots are already used to provide certain types of therapy, researchers in the European Network on Robotics for NeuroRehabilitation aim to improve them further, as Dr Thierry Keller explains
Intelligent robotics for neurological conditions The aging nature
of the European population is likely to have a significant impact on healthcare services. With demand for effective rehabilitation therapies set to increase, the work of the European Network on Robotics for NeuroRehabilitation takes on real importance. “The aim in this project is to coordinate and stimulate European research in rehabilitation therapy. We also aim to link this neurorehabilitation research with computational modelling, motor learning and neuroimaging,” outlines Dr Thierry Keller, the Action’s Chair. Robots are already used to provide certain types of therapy, now Dr Keller and his colleagues are working to develop new, robot-assisted therapies that are tailored to the needs of individual patients. “There are systems available that train people in for example locomotion, in grasping and reaching. These systems are primarily used to provide therapy to patients who have been quite severely affected by stroke,” he says.
Intelligent robotics Evidence suggests that the key factor in determining whether a robot-assisted therapy is effective or not is the frequency with which it is applied. Therapies which help patients repeat certain actions have already proved effective; now researchers are aiming to develop robots that are able to learn from experience and adapt their approach accordingly. “The working hypothesis for the future development of robotics is that when the robot is more intelligent, it can also guide the therapy in the right direction for each individual patient. So it can not only repeat actions, but also adapt to what the patient is doing, and then correct itself in the support levels it provides in a broad range, from assistance to resistance,” explains Dr Keller. This is a key element of the Action’s agenda, which includes both fundamental and applied research. “The Action brings together researchers from several different disciplines. We have members from neuroscience and specialists in motor control, as well as medical therapists and engineers,” says Dr Keller.
16
The majority of robots work either with fixed moment patterns or follow predemonstrated pattern in a so-called teach and repeat mode, where the therapist demonstrates the movement, that is subsequently recorded and repeated. Now the Action aims to improve robotics technologies further; Dr Keller and his colleagues at Tecnalia have developed an arm robot, which conveys instructions to patients. “A screen gives patients the instruction on the type of movements they should do through a visual, hi-definition pack. We call this the exercise gaming approach, where the patient has to put
Personalised treatment A key question in research is the degree to which therapy can be personalised through the application of robot-assisted therapies. This depends to a degree on accurate assessment of the patients’ responses at an early stage; Dr Keller says this is an important element in the exercise gaming approach. “There are assessment games in the beginning and at the end of the exercise. We can compare how the patient does during these exercises; for example their range of motion, or the force that they can apply, and then we can also quantify their progress from one day to the next,” he outlines.
The working hypothesis for the future development of robotics is that when the robot is more intelligent, therapy can be guided in the right direction for each individual patient together a puzzle, or follow a specific trajectory,” he explains. The robot interacts with the patient during both these tasks, and can be connected with therapists over a network. “The therapists can then remotely observe the performance of the patient, but also analyse the available information and then modify the level of difficulty of the exercise accordingly and put together a tailored rehabilitation programme,” continues Dr Keller.
Image taken at the Military Rehabilitation Center Aardenburg in the Netherlands.
Therapists play a central role in assessing patients, yet robots can also provide key information. “We aim to find out which types of assessments the clinicians will need to provide and which robots can provide. Which of those assessments will robots be able to provide in future? And what will robots need in order to make these assessments?” asks Dr Keller. The backdrop of this research is the aging nature of European society and its likely impact on healthcare services. The incidence of neurological diseases is predicted to increase over the coming years, which will place a heavy burden on healthcare services; robotic therapies take on real importance in this context. “Many private hospitals now apply rehabilitation robotics by having one therapist that then supervises four patients at the time. They have several robots, and can then basically go from station to station,” says Dr Keller. This approach combines the professional expertise of the therapist with the repeatability of the robot. “The therapist can give the human input – which is of course always extremely important – but they don’t have to do the strenuous manual therapy work,” points out Dr Keller. “For
EU Research