Current Research Projects
The Rehabilitation Engineering Laboratory pursues three research lines, ranging from basic research on the mechansims of sensorimotor control and recovery following neurological injury, to improved tools for assessment, therapy and long-term assistance.
This line of research focuses on the investigation of mechanical properties of joints under neural control, as well as the neural and behavioral mechanisms underlying sensorimotor control and recovery following neurological injury. It involves studies with robotic tools and non-invasive neuroimaging in humans and animal models. We work closely with collaborators in movement science, neuroscience and clinical neurorehabilitation.
Dynamic Estimation of Human Knee Joint Impedance during Gait
We are investigating how healthy humans control the mechanical impedance of the knee joint during gait. We developed a powered exoskeleton to apply brief position perturbations at specific points in the gait cycle, while measuring the interaction torque. The gained knowledge can be used for the design and control of robotic prostheses and orthoses.
ETH Pattus – A Platform to Study Motor Learning and Recovery in Rodents
The ETH Pattus is a robotic platform designed for motor learning experiments with rodents. This device allows the implementation of force fields to guide or perturb forelimb motion in a well-controlled and reproducible manner, and provides a quantitative assessment of movement kinematics. The ETH Pattus opens up new research avenues for the characterization of motor learning stages, both in healthy and in stroke animal models.
Accessing the Brain Using Functional Near-Infrared Spectroscopy (fNIRS)
Functional Near-Infrared Spectroscopy (fNIRS) is an emerging technology allowing to assess brain activity by measuring the attenuation of infrared light transmitted through the brain. Brain-Computer-Interfaces (BCIs) using fNIRS can be small, lightweight and inexpensive, offering a spatially localized readout of brain activity without posing any risk to the user. We are developing a wearable and modular fNIRS module for BCI applications based on silicon photomultipliers (SiPMs).
Neural Self-Manipulation of Food Reward by means of Real-Time fMRI Neurofeedback
Certain foods are more rewarding than others. The mesolimbic dopaminergic pathway in the brain induces incentive motivation to acquire rewards. Training humans to down-regulate their brain response to visually presented appetizing food stimuli in a real-time fMRI neurofeedback paradigm could help reduce the relative reinforcing value of the stimuli. This may facilitate development of non-invasive, stimulus-specific interventions for eating disorders.
Functional Imaging and Robotics for Sensorimotor Transformation
There is evidence that the central nervous system processes tactile information similarly to visual information. However, little is known about the brain networks which process tactile information. We are using robotics and resting state functional magnetic resonance imaging to gain a better understanding of the neural mechanisms involved in using touch for dexterous control of the hand.
Clinical assessments have the potential to provide valuable insights on the effect of different therapeutic approaches. However, they are time-consuming to administer and suffer from a number of limitations, such as subjectivity, flooring/ceiling effects and poor sensitivity. The aim of this line of research is to provide clinicians and therapists with objective and sensitive tools to precisely quantify impariments, both motor and somatosensory, in the clinic as well as in the home environment.
Robotic Somatosensory Assessment of Hand Function
Somatosensory deficits following neurological injury such as stroke are poorly investigated and rarely addressed during rehabilitation, despite the importance of somatosensory input in motor learning and control. This can be at least partially explained by the lack of quantiative and sensitive assessment tools. Our aim is to develop quantitative and sensitive assessment tools based on robotics technology, allowing to diagnose and monitor somatosensory deficits, compare different therapies and develop assessment-driven somatosensory therapy for a more effective and complete recovery.
Long-Term Activity Monitoring
Neurological diseases such as stroke, Parkinson’s disease or spinal cord injuries can cause various forms of motor disabilities. Wearable sensors are a promising approach to collecting physical activity data of patients in order to objectively monitor their recovery outside of a clinical environment as well as to investigate the transfer of therapy to activities of daily living.
Assessment of Upper Limb Function with the Virtual Peg Insertion Test
The Virtual Peg Insertion Test is a tool for the objective assessment of grasping and manipulation function. It combines a commercial haptic device, an instrumented handle and a virtual reality environment. Performance and impariment related parameters can be extracted from position and grip force traces to obtain an objective evaluation of upper limb function. Clinical studies are ongoing with patients suffering from different neurological disorders.
The aim of this line of research is to integrate the tools and insights from the Explore and Assess research lines, bringing them to clinical application. This ranges from robot-assisted rehabilitation of hand function, to long term assistance with novel prosthetic/orthotic devices and surgical robotics.
Exoskeleton with Variable Impedance Actuation for Paraplegics
Robotic systems to restore the ability of walking have been emerging in the recent past. However, unlike human legs they are typically not able to adjust their impedance to the terrain and the task, which makes them more susceptible for falls and mechanical damage to the structure. The VariLeg exoskeleton is the first exoskeleton with a variable stiffness actuation designed to support the full weight of a user. It has been developped in the course of a Focus Project and will be competing at the Cybathlon.
Robot-Assisted Rehabilitation and Assessment of Hand Function
Stroke survivors suffering from hand impairment are often severely limited in the execution of activities of daily living. We are developing and clinically evaluating robotic devices to train and assess motor and sensory hand/forearm function. These systems can complement conventional therapy, and allow for unsupervised rehabilitation in the clinic and at home.
Robotic Hand Orthosis for Therapy and Assistance in Activities of Daily Living
A compact and lightweight hand exoskeleton has been developed in collaboration with Kyushu University (Japan). The EMG-controlled device assists stroke patients during grasping tasks in rehabilitation training and during activities of daily living. Its soft actuation mechanism allows for grasping of a variety of objects. Thanks to 3D-rapid prototyping, it can be tailored to the each individual user.
Robotic Platform for Autonomous Surgery
A robotic platform was developed as part of the European project I-SUR (Intelligent Surgical Robot), with the aim of performing simple surgical tasks automatically. The developed I-SUR robot has a modular structure with up to 18 actuated degrees of freedom allowing for the execution of precise needle insertion and suturing.