Cable-driven exosuits assist users in ambulatory activities by transmitting assistive torques from motors to the actuated joints. State-of-the-art exosuits typically use Bowden cable transmissions, albeit their limited efficiencies (40–60 %) and non-linear response in curved paths. This paper evaluates the efficiency and responsiveness of a new cable-pulley transmission compared to a Bowden transmission, using both steel and Dyneema cables. The analysis includes three experiments: a test bench simulating a curved transmission path, followed by a static and dynamic experiment where six unimpaired participants donned an exosuit featuring both transmissions across the hips and knees. Our findings demonstrate that the pulley transmission consistently outperformed the Bowden's efficiency by absolute margins of 18.77$\pm 7.29\%$using a steel cable and by 40.60$\pm 6.76\%$using a Dyneema cable across all experiments. Additionally, the steel cable was on average 19.19$\pm 5.29\%$more efficient than the Dyneema cable in the pulley transmission and 41.02$\pm 6.34\%$in the Bowden tube. These results led to the development of the Stillsuit, a novel lower-limb cable-driven exosuit that uses a pulley transmission and steel cable. The Stillsuit sets a new benchmark for exosuits with 87.56$\pm$3.92 % transmission efficiency, generating similar biological torques to those found in literature (16.4% and 19.0% of the biological knee and hip torques, respectively) while using smaller motors, resulting in a lighter actuation unit (1.92 kg).

Current blood pressure (BP) estimation methods have not achieved an accurate and adaptable approach for ambulatory diagnosis and monitoring applications of populations at risk of cardiovascular disease, generally due to a limited sample size. This paper introduces an algorithm for BP estimation solely reliant on photoplethysmography (PPG) signals and demographic features. It automatically obtains signal features and employs the Markov Blanket (MB) feature selection to discern informative and transmissible features, achieving a robust space adaptable to the population shift. This approach was validated with the Aurora-BP database, compromising ambulatory wearable cuffless BP measurements for over 500 individuals. After evaluating several machine-learning regression methods, Gradient Boosting emerged as the most effective. According to the MB feature selection, temporal, frequency, and demographic features ranked highest in importance, while statistical ones were deemed non-significant. A comparative assessment of a generic model (trained on unclassified BP data) and specialized models (tailored to each distinct BP population), demonstrated a consistent superiority of our proposed MB feature space with a mean absolute error of [Formula: see text] for systolic BP and [Formula: see text] for diastolic BP on the whole dataset. Moreover, we present a first comparison of in-clinic vs. ambulatory models, with performance significantly lower for the latter with a drop of [Formula: see text] in systolic ( ) and [Formula: see text] for diastolic ( ) estimation errors. This work contributes to the resilient understanding of BP estimation algorithms from PPG signals, providing causal features in the signal and quantifying the disparities between ambulatory and in-clinic measurements.

Robots were introduced in the field of upper limb neurorehabilitation to relieve the therapist from physical labor, and to provide high-intensity therapy to the patient. A variety of control methods were developed that incorporate patients' physiological and biomechanical states to adapt the provided assistance automatically. Higher level states, such as selected type of assistance, chosen task characteristics, defined session goals, and given patient impairments, are often neglected or modeled into tight requirements, low-dimensional study designs, and narrow inclusion criteria so that presented solutions cannot be transferred to other tasks, robotic devices or target groups. In this work, we present the design of a modular high-level control framework based on invariant states covering all decision layers in therapy. We verified the functionality of our framework on the assistance and task layer by outlaying the invariant states based on the characteristics of 20 examined state-of-the-art controllers. Then, we integrated four controllers on each layer and designed two algorithms that automatically selected suitable controllers. The framework was deployed on an arm rehabilitation robot and tested on one participant acting as a patient. We observed plausible system reactions to external changes by a second operator representing a therapist. We believe that this work will boost the development of novel controllers and selection algorithms in cooperative decision-making on layers other than assistance, and eases transferability and integration of existing solutions on lower layers into arbitrary robotic systems.

Exoskeleton robots found application in neurorehabilitation, telemanipulation, and power augmentation. The human–robot attachment system of an exoskeleton should transmit all the interaction forces while keeping the anatomical and robotic joint axes aligned. Existing attachment concepts were bounding the performance of modern exoskeletons due to insufficient stiffness for high-performance force control, time-consuming adaption processes, and/or bulkiness. Therefore, we developed an augmented attachment system for a recent fully actuated nine-degree-of-freedom upper limb exoskeleton. The proposed system was compared to a conventional solution in a case study with four participants. The proposed attachment system lowered the relative motion between the human and the robot under static loads for all defined landmarks by 45% on average. The occurrence of undesired contacts in the trials was mitigated by 74%, thus improving conditions for closed-loop force control. Furthermore, the proposed system adapted better to the user's anatomy facilitating more accurate alignment and less obstruction. On average, self-attachment took$\mathbf {43(8.3)}$$\mathrm{s}$to don(doff). Thereby, the alignment of anatomic landmarks had typically less than 15 mm offset to a thorough expert alignment, making self-attachment eligible. The augmented attachment system and the insights gained by the case study are expected to enable improvement of the physical human–robot interaction of exoskeletons.

We developed an exoskeleton for neurorehabilitation that covered all relevant degrees of freedom of the human arm while providing enough range of motion, speed, strength, and haptic-rendering function for therapy of severely affected (e.g., mobilization) and mildly affected patients (e.g., strength and speed). The ANYexo 2.0, uniting these capabilities, could be the vanguard for highly versatile therapeutic robotics applicable to a broad target group and an extensive range of exercises. Thereby, the practical adoption of these devices in clinics will be fostered. The unique kinematic structure of the robot and the bio-inspired controlled shoulder coupling allowed training for most activities of daily living. We demonstrated this capability with 15 sample activities, including interaction with real objects and the own body with the robot in transparent mode. The robot's joints can reach$200 \%$,$398 \%$, and$354 \%$of the speed required during activities of daily living at the shoulder, elbow, and wrist, respectively. Further, the robot can provide isometric strength training. We present a detailed analysis of the kinematic properties and propose algorithms for intuitive control implementation.

The intention-detection strategy used to drive an exosuit is fundamental to evaluate the effectiveness and acceptability of the device. Yet, current literature on wearable soft robotics lacks evidence on the comparative performance of different control approaches for online intention-detection. In the present work, we compare two different and complementary controllers on a wearable robotic suit, previously formulated and tested by our group; a model-based myoelectric control (myoprocessor), which estimates the joint torque from the activation of target muscles, and a force control that estimates human torques using an inverse dynamics model (dynamic arm). We test them on a cohort of healthy participants performing tasks replicating functional activities of daily living involving a wide range of dynamic movements. Our results suggest that both controllers are robust and effective in detecting human–motor interaction, and show comparable performance for augmenting muscular activity. In particular, the biceps brachii activity was reduced by up to 74% under the assistance of thedynamic armand up to 47% under themyoprocessor, compared to a no-suit condition. However, themyoprocessoroutperformed thedynamic armin promptness and assistance during movements that involve high dynamics. The exosuit work normalized with respect to the overall work was$68.84 \pm 3.81\%$when it was ran by themyoprocessor, compared to$45.29 \pm 7.71\%$during thedynamic armcondition. The reliability and accuracy of motor intention detection strategies in wearable device is paramount for both the efficacy and acceptability of this technology. In this article, we offer a detailed analysis of the two most widely used control approaches, trying to highlight their intrinsic structural differences and to discuss their different and complementary performance.

Wearable robots are undergoing a disruptive transition, from the rigid machines that populated the science-fiction world in the early 1980s to lightweight robotic apparel, hardly distinguishable from our daily clothes. In less than a decade of development, soft robotic suits have achieved important results in human motor assistance and augmentation. In this article, we start by giving a definition of soft robotic suits and proposing a taxonomy to classify existing systems. We then critically review the modes of actuation, the physical human–robot interface and the intention-detection strategies of state-of-the-art soft robotic suits, highlighting the advantages and limitations of different approaches. Finally, we discuss the impact of this new technology on human movements, for both augmenting human function and supporting motor impairments, and identify areas that are in need of further development.

During robot-assisted therapy of hemiplegic patients, interaction with the patient must be intrinsically safe. Straight-forward collision avoidance solutions can provide this safety requirement with conservative margins. These margins heavily reduce the robot's workspace and make interaction with the patient's unguided body parts impossible. However, interaction with the own body is highly beneficial from a therapeutic point of view. We tackle this problem by combining haptic rendering techniques with classical computer vision methods. Our proposed solution consists of a pipeline that builds collision objects from point clouds in real-time and a controller that renders haptic interaction. The raw sensor data is processed to overcome noise and occlusion problems. Our proposed approach is validated on the 6 DoF exoskeleton ANYexo for direct impacts, sliding scenarios, and dynamic collision surfaces. The results show that this method has the potential to successfully prevent collisions and allow haptic interaction for highly dynamic environments. We believe that this work significantly adds to the usability of current exoskeletons by enabling virtual haptic interaction with the patient's body parts in human-robot therapy.

A high-quality free-motion rendering is one of the most vital traits to achieve an immersive human-robot interaction. Rendering free-motion is notably challenging for rehabilitation exoskeletons due to their relatively high weight and powerful actuators required for strength training and support. In the presence of dynamic human movements, accurate feedback linearization of the robot's dynamics is necessary to allow for a linear synthesis of interaction wrench controllers. Hence, we introduce a virtual model controller that uses two 6-DoF force sensors to control the interaction wrenches of a multi-DoF torque-controlled exoskeleton over the joint accelerations and inverse dynamics. Furthermore, we propose a disturbance observer for controlling the joint acceleration to diminish the influence of modeling errors on the inverse dynamics. To provide a high-bandwidth, low-bias estimation of the system's acceleration, we introduce a bias-observer which fuses the information from joint encoders and seven low priced IMUs. We have validated the performance of our proposed control structure on the shoulder and arm exoskeleton ANYexo. The experimental comparison of the controllers shows a reduction of the felt inertia and maximum reflected joint torque by a factor of more than three compared to state of the art. The controllers' robustness w.r.t. a model mismatch is validated. The experiments show that the closed-loop acceleration control improves the tracking, particularly at joints with low inertia. The proposed controllers' performance sets a new benchmark in haptic transparency for comparable devices and should be transferable to other applications.

Multiplayer environments are thought to increase and prolongate active participation in robot-aided rehabilitation. We expect that environments linking patients with their spouses will particularly foster active participation. Thus, we developed two multiplayer games to link the game experience of two players: an Air Hockey game and a Haptic Kitchen game. In the competitive Air Hockey game, differences in skill levels between players were balanced by individualizing haptic guidance or damping forces. In the Haptic Kitchen game, a healthy player could support the patient's movements using a virtual force field. The two players could control the haptic interaction since both the force field and the point of application were visualized. We tested the haptic performance balancing algorithm of the Air Hockey game and the spouse-controlled haptic support of the Kitchen game with patients post-stroke who trained both single- (i.e., alone) and multiplayer training (i.e., with spouse) in eight therapy sessions lasting 45 min each. Mean total rating in Intrinsic Motivation Inventory was 46.9 points (out of 63 points) for multiplayer modes, and 42.7 points for single player modes, respectively. The spouses applied the haptic support in the Haptic Kitchen game during 42 % of the total game duration. We are currently testing more patient-spouse couples to better understand the effects of using these haptic approaches on the behavior and recovery of patients. We foresee this approach can improve the motivation during training and positively influence the at-home behavior of patients, an important goal of rehabilitation training efforts.

Error feedback is critical for supporting motor adaptation in rehabilitation, sports, piloting, and skilled manual tasks. Error augmentation interventions, in which participants' errors are amplified with either visual or haptic feedback during training has shown success over repetitive practice. Here we show that the statistical tendencies arising from free movement exploration can improve error augmentation with customized training forces that vary across the trajectory. We hypothesized that with customized error augmentation participants will adapt faster to learning a visual-motor distortion and have greater improvement than participants receiving standard error augmentation and participants repetitively practicing the task. We tested twenty-one participants using a robotic exoskeleton device restricted to two degrees of freedom. We found that participants receiving customized forces adapted faster and consequently changed with smaller forces. Further, change in error was greatest for participants receiving customized forces. These promising results support the need for customization to target subject specific errors.

This paper uses physiological measurements to estimate human workload and effort in physical human–robot interaction. Ten subjects performed 19 consecutive task periods using the ARMin robot while difficulty was varied along two scales. Three physiological modalities were measured: electroencephalography, autonomic nervous system (ANS) responses (electrocardiography, skin conductance, respiration, skin temperature) and eye tracking. After each task period, reference workload and effort values were collected using the NASA Task Load Index. Machine learning was used to estimate workload and effort from physiological data. All three physiological modalities performed significantly better than random, particularly using nonlinear estimation algorithms. The most important ANS responses were respiration and skin conductance, while the most important electroencephalographic information was obtained from frontal and central sites. However, all three physiological modalities were outperformed by task performance and movement data. This suggests that future studies should try to demonstrate advantages of physiological measurements over other information sources.

Rehabilitation robots have the potential to greatly improve motor rehabilitation. However, the patient must be properly motivated to actively participate in therapy. Several strategies have been suggested to improve patient motivation, but one element has not yet been explored: playing with other people. We designed a two-player rehabilitation game played by two people using two ARMin III robots. We tested three game modes: single-player (competing against a computer), competitive (competing against a human), and cooperative (cooperating with a human against a computer). All modes were played by 24 healthy subjects who filled out questionnaires about their personality and in-game motivation. Almost all subjects preferred playing the two-player game modes to the single-player one, as they enjoyed being able to talk and interact with another person. However, there were two distinct player groups. One group liked the competitive mode but not the cooperative mode while the other liked the cooperative but not the competitive mode. Subjects who liked the competitive mode also put more effort into it. Finally, subjects' personalities partially predicted what mode they would like. This emphasizes that two-player rehabilitation games have advantages over single-player ones, but that the right game needs to be chosen for each subject. An extended patient study is planned for the near future.

Body weight support systems are frequently used as part of robotic gait training to provide unloading in order to help subjects perform walking, but can also induce stabilizing forces and render the task of maintaining balance less challenging. In this paper, a two-dimensional body weight support system extension is presented which reduces lateral forces induced on the subject by means of linearly translating the cable pulley according to lateral movements of the subject. It is demonstrated that the system accurately tracks lateral movements of the pelvis at different levels of vertical support load and thereby lowers the induced lateral forces. The system will be used in advanced robotic body weight supported treadmill walking incorporating a balance training element.

Functional near-infrared spectroscopy (fNIRS) is a noninvasive optical method that measures cortical activity based on hemodynamics in the brain. Physiological signals (biosignals), such as blood pressure and respiration, are known to appear in cortical fNIRS recordings. Some biosignal components occupy the same frequency band as the cortical response, and respond to the subjects activity. To process an fNIRS signal in a brain-computer interface, it is desirable to know which components of the signal come from cortical response, and which come from biosignal interference. Numerous filtering methods have been proposed to this end with mixed success, possibly because they assume that the cortical and physiological signals combine linearly, or that biosignals do not correlate with subject behavior. Here, we propose an adaptive filter with a cost function based on mutual information to selectively remove information that correlates with blood pressure from the fNIRS signal. The filter was tested with real and simulated data. The real signals were measured on seven healthy subjects performing an isometric pinching task. Cross-correlation and mutual information were employed as performance measures. The filter successfully removed correlations between blood pressure and the fNIRS signal, by an equal or greater amount compared to a traditional recursive least squares adaptive filter. Blood pressure was found to be the most informative signal to classify rest and active periods using linear discriminant analysis. Any task information in the fNIRS signal was redundant to that expressed by blood pressure.

The Cybathlon is a championship for racing pilots with disabilities (i.e., parathletes) who are using advanced assistive devices including robotic technologies. The competitions are comprised by different disciplines that apply the most modern powered devices such as prostheses, wearable exoskeletons, wheelchairs, functional electrical stimulation as well as novel brain-computer interfaces. The main goal of the Cybathlon is to provide a platform for the development of novel assistive technologies that are useful for daily life. Through the organization of the Cybathlon we want to help removing barriers between the public, people with disabilities and science. The first Cybathlon will take place on October 8, 2016. Thereafter, the Cybathlon will be held periodically every 2 to 4 years.

In an affective feedback loop, the computer maps various measurements to affective variables such as enjoyment, then adapts its behavior based on the recognized affects. The affect recognition is never perfect, and its accuracy (percentage of times the correct affective state is recognized) depends on many factors. However, it is unclear how this accuracy relates to the overall user experience. As recognition accuracy is difficult to control in a real affective feedback loop, we describe a method of simulating recognition accuracy in a game where difficulty is increased or decreased after each round. The game was played by 261 participants at different simulated recognition accuracies. Participants reported their satisfaction with the recognition algorithm as well as their overall game experience. We observed that in such a game, the affective feedback loop must adapt game difficulty with an accuracy of at least 80 percent to be accepted by users. Furthermore, users who do not enjoy the game are likely to stop playing it rather than continue playing and report low enjoyment. However, the acceptable recognition accuracy may not generalize to other contexts, and studies of affect recognition accuracies in other applications are needed.

Prolonged bed rest in severely paralyzed or intensive care patients is associated with adverse secondary effects on cardiopulmonary function. To counteract these effects of immobility in bed-ridden patients, we aim at controlling and stabilizing the cardiovascular system via multiple mechanical input variables. A challenge in this control problem is to provide an accurate model of the plant to be controlled. As humans are time variant systems and show individual physiological reactions to external stimuli the identification of such a model appears to be challenging. The current work presents a model-free predictive controller which takes into account these challenges. In this paper we present data concerning the control of heart rate, systolic and diastolic blood pressures, and mean arterial blood pressure via body tilting and leg mobilization. The controller was validated in a simulation study and feasibility was tested on two healthy subjects. The experimental results with healthy subjects show that the mean value differed in average less than 1 beat per minute (bpm) from the desired heart rate values and less than 1 mm Hg from the desired blood pressure values. The long term goal of this project is to control also breathing via body tilting, stepping and electrical muscle stimulation.

Intensive rehabilitation training of the arm can improve motor recovery in patients with neurological impairment. Actuated robots are becoming more and more common in this field as they serve to actively assist, enhance and assess neurorehabilitation. However, there is currently no actuated robot available specifically designed for the rehabilitation of children with upper extremity motor impairments. In this paper, we describe a completely new designed exoskeleton-based arm robot, ChARMin, with four degrees of freedom to guide and assist shoulder and elbow movements for young patients with motor impairments. The serial mechanical structure includes parallel kinematics for remote center of rotation actuation. This allows to keep a safe distance between parts of the robot and the patient and it reduces friction, while being highly adaptable to cover the anthropometrics for patients aged 5 to 18 years. Additionally, a novel passive weight support mechanism and 6 degrees of freedom force sensors are installed for a safe and transparent operation of the device.

In recent years, gait robots have become increasingly common for gait rehabilitation in non-ambulatory stroke patients. Cardiovascular treadmill training, which has been shown to provide great benefit to stroke survivors, cannot be performed with non-ambulatory patients. We therefore integrated cardiovascular training in robot-assisted gait therapy to combine the benefits of both training modi. We developed a model of human heart rate as a function of exercise parameters during robot-assisted gait training and applied it for automatic control purposes. This structural model of the physiological processes describes the change in heart rate caused by treadmill speed and the power exchanged between robot and subject. We performed physiological parameter estimation for each tested individual and designed a model-based feedback controller to guide heart rate to a desired time profile. Five healthy subjects and eight stroke patients were recorded for model parameter identification, which was successfully used for heart rate control of three healthy subjects. We showed that a model-based control approach can take into account patient-specific limitations of treadmill speed as well as individual power expenditure.

Nowadays, “user-cooperative” control strategies are commonly used in robot-assisted motor (re-)learning. User-cooperative strategies enable compliant haptic interactions between robot and user: the robot only intervenes as needed, instead of forcing the user to follow a fixed predefined movement. However, the effectiveness of user-cooperative control is contro-versially discussed. Recent studies indicate that the effectiveness of user-cooperative control strategies will be enhanced when every user is individually provided with an optimal amount of assistance or challenge. In conventional motor (re-)learning, such an optimal amount of assistance or challenge is successfully applied by physiotherapists and trainers.

The number of ropes for a fully constrained, tendon-based robot has to be larger than the actuated degrees of freedom since ropes only impose unidirectional constraints. This actuation redundancy implicates that more position information is available than would be required for the the determination of the end-effector pose. This leads to an optimization problem for the forward kinematics of the robot which has to be solved in real-time. Furthermore, the kinematics of tendon-based robots are often kept simple in existing systems by guiding the ropes through holes into the workspace. This facilitates the description of the rope vectors. However, this solution is not applicable for high-load applications, as friction would cause excessive non-linearities and wear. To solve the forward kinematics of tenon-based robots, we introduce a physics-based interpretation of the mentioned optimization problem. The robotic system is described as a damped oscillator whose resting position is equal to the optimal solution. As a major advantage over the known algorithms, this physics-based approach is quantifiable in terms of accuracy of the solution and number of iterations. Furthermore, the design and mathematical description of a deflection unit's geometry is presented. This deflection unit guides the rope smoothly into the workspace and its relevant influence on the kinematic equations can be compensated. The physics-based approach is experimentally evaluated on a tendon-based haptic interface, the r3-system, and it is compared to the solutions using only the minimum set of sensor information.

For haptic devices, compensation of the robot's gravity is a frequent strategy with the aim to reduce interaction forces between robot and human in zero-impedance control. However, a closer look at the composition of these interaction forces may reveal that the net effect of uncompensated gravitational components of the robot actually reduces interaction forces during dynamic movements, because inertial and gravitational components at least partially compensate each other. This is the case in lower extremity exoskeletons, where less user force is necessary to swing the robot's leg when gravity helps. Here, we go one step further by shaping optimal passive dynamics for arbitrary haptic devices. The proposed method of generalized elasticities uses conservative force fields to improve haptic transparency for certain movements types. In an example realization, these force fields are generated by elasticities spanning multiple joints. Practical experiments with the Lokomat lower extremity exoskeleton show the success of the proposed method in terms of reduced interaction torques and more physiological user motion compared to gravity compensation.

This paper presents the design of a new user-cooperative rope robot. This robot serves as a large-scale haptic interface in a multi-modal Cave environment used for sport simulation. In contrast to current rope robots, the configuration of the presented robot is adaptable to different simulation tasks what makes the robot more versatile. However, this adaptability and the high dynamics in sports lead to challenging requirements and specific design criteria of the hardware components. We present the requirements on the single robot components as well as the design of the entire setup optimized in terms of user-cooperativity and versatility. The setup includes sensors to measure the relevant parameters for user-cooperative control, i.e. position with a high resolution and the rope forces. Furthermore, an algorithm is introduced, which calculates the distance between the single ropes and the user in order to avoid collisions between the ropes and the user. Single points on the user's body are, therefore, tracked with a motion tracking system; the user's single body parts are then represented by geometrical objects whose distances to the ropes are calculated. The algorithm is programmed in such way that the collision detection runs in real-time. Both, the hardware and the algorithm, were evaluated experimentally in two applications, a rowing simulator and a tennis application. The hardware concept combined with the distance calculation allows the use of new kinematic concepts and expands the spectrum of realizable movement tasks that can be implemented into the Cave environment.

The rehabilitation robot Lokomat allows automated treadmill training for patients with neurological gait disorders. The basic position control approach for the robot has been extended to patient-cooperative strategies. These strategies provide more freedom and allow patients to actively influence their training. However, patients are likely to need additional support during patient-cooperative training. In this paper, we propose an algorithm based on iterative learning control that shapes a supportive torque field. The torque field is supposed to assist the patient as much as needed in performing the desired task. We evaluated the algorithm in a proof-of-concept experiment with 3 healthy subjects. Results showed that the amount of support was automatically adapted to the activity and the individual needs of the subjects. Furthermore, the support improved the performance of the subjects.

Task-oriented repetitive movements can improve motor recovery in patients with neurological or orthopaedic lesions. The application of robotics can serve to assist, enhance, evaluate, and document neurological and orthopaedic rehabilitation. ARMin II is the second prototype of a robot for arm therapy applicable to the training of activities of daily living. ARMin II has a semi-exoskeletal structure with seven active degrees of freedom (two of them coupled), five adjustable segments to fit in with different patient sizes, and is equipped with position and force sensors. The mechanical structure, the actuators and the sensors of the robot are optimized for patient-cooperative control strategies based on impedance and admittance architectures. This paper describes the mechanical structure and kinematics of ARMin II.

The rehabilitation robot LOKOMAT has been developed at University Hospital Balgrist to automate treadmill training of spinal cord injury and stroke patients. Current rehabilitation training on that robot consists of moving the patient's legs on predefined trajectories. However, this kind of training is not challenging, as patients are moved regardless of their efforts and do not see their advancement. To enhance rehabilitation training with the LOKOMAT, a virtual reality setup was installed. It consists of a passive stereo projection system (screen size 3 m times 2 m), a Dolby 5.1 sound system and an electric fan. With that setup an obstacle crossing scenario was implemented. The patients can see their advancement on the screen, as an animated figurine (avatar) moves along a path simultaneously with their own movements. Additionally they can hear sounds (e.g. environmental sounds, steps), feel the wind, and experience force feedback, provided by the orthosis, when hitting obstacles. The objective of a first study on visual feedback was to investigate which feedback suits best to perceive the obstacle distances and heights correctly. To answer this question, 14 healthy subjects walked in the actuated gait orthosis, received visual feedback and tried to avoid collisions with obstacles. Subjects could move freely within the gait orthosis and determine their own speed and step length. They had to cross the obstacles independently, with haptic feedback, indicating obstacle hits. Results show that the side view results in least obstacle hits and that 2D excels 3D display in this respect

We present the technical principle and evaluation of a multimodal virtual reality (VR) system for medical education, called a touch simulator. This touch simulator comes with an innovative three-dimensional (3-D) touch sensitive input device. The device comprises a six-axis force-torque sensor connected to a tangible object representing the shape of an anatomical structure. Information related to the point of contact is recorded by the sensor, processed, and audiovisually displayed. The touch simulator provides a high level of user-friendliness and fidelity compared to other purely graphically oriented simulation environments. In this paper, the touch simulator has been realized as an interactive neuroanatomical training simulator. The user can visualize and manipulate graphical information of the brain surface or different cross-sectional slices by a finger-touch on a brain-like shaped tangible object. We evaluated the system by theoretical derivations, experiments, and subjective questionnaires. In the theoretical analysis, we could show that the contact point estimation error mainly depends on the accuracy and the noise of the sensor, the amount and direction of the applied force, and the geometry of the tangible object. The theoretical results could be validated by experiments: applying a normal force of 10 N on a 120 mm x 120 mm x 120 mm cube causes a maximum error of 2.5 +/- 0.7 mm. This error becomes smaller when increasing the contact force. Based on the survey results, the touch simulator may be a useful tool for assisting medical schools in the visualization of brain image data and the study of neuroanatomy.

Task-oriented repetitive movements can improve motor recovery in patients with neurological or orthopaedic lesions. The application of robotics can serve to assist, enhance, evaluate, and document neurological and orthopaedic rehabilitation. ARMin is a new robot for arm therapy applicable to the training of activities of daily living in clinics. ARMin has a semiexoskeletal structure with six degrees of freedom, and is equipped with position and force sensors. The mechanical structure, the actuators and the sensors of the robot are optimized for patient-cooperative control strategies based on impedance and admittance architectures. This paper describes the mechanical structure, the control system, the sensors and actuators, safety aspects and results of a first pilot study with hemiplegic and spinal cord injured subjects