Based on evidence from the previous research in rehabilitation robot control strategies, we found that the common feature of the effective control strategies to promote subjects’ engagement is creating a reward–punishment feedback mechanism. This article proposes a reward–punishment feedback control strategy based on energy information. Firstly, an engagement estimated approach based on energy information is developed to evaluate subjects’ performance. Secondly, the estimated result forms a reward–punishment term, which is introduced into a standard model-based adaptive controller. This modified adaptive controller is capable of giving the reward–punishment feedback to subjects according to their engagement. Finally, several experiments are implemented using a wrist rehabilitation robot to evaluate the proposed control strategy with 10 healthy subjects who have not cardiovascular and cerebrovascular diseases. The results of these experiments show that the mean coefficient of determination (R2) of the data obtained by the proposed approach and the classical approach is 0.7988, which illustrate the reliability of the engagement estimated approach based on energy information. And the results also demonstrate that the proposed controller has great potential to promote patients’ engagement for wrist rehabilitation.
Stroke has become one of the major diseases that threaten people’s physical and mental health in the world.1 Loss of control of upper limbs is a common impairment underlying disability after stroke for patients, which seriously affects their daily activities.2 Traditional physical therapy is labor intensive and requires great energy of therapists.3 With the development of robotics, the emergence of rehabilitation robots provides a new way for rehabilitation.4 Rehabilitation robots are able to assist patients to complete training tasks without therapists. In addition, rehabilitation robots are capable of estimating patients’ rehabilitation status accurately through a variety of sensors, which helps therapists to develop a follow-up treatment plan for patients.
Control of rehabilitation robots, however, remains an open-ended research area. Control strategies, which target subjects ranging from the mildly impaired and severely impaired, are the most extensively investigated controller paradigm in the rehabilitation robotics community and have been proved to be the most promising techniques for promoting recovery after stroke.5,6 There is strong evidence that high engagement in rehabilitation training induces neural plasticity.7 Therefore, great attention is paid on investigating how to use robot control strategies to promote patients’ active engagement in robotic therapy.
Assist-as-needed (AAN) control strategy is one of the most popular research topics in the field of rehabilitation robots control strategies and is considered promising to promote patients’ engagement. As the name suggested, AAN control strategy emphasizes that robots only supply as much effort as a patient needs to accomplish training tasks by estimating his/her performance in real time.8 Impedance control first proposed by Hogon was applied in AAN control strategy primitively.9 Representatively, Krebs et al. proposed an AAN controller based on impedance control with MIT-Manus,10,11 which can update impedance parameters according to patients’ performance. In this case of robotic therapy, the robot provides assistance based on specific impedance parameters when the subject is not able to track the desired trajectory and does not provide assistance when the subject is able to track or exceed the desired trajectory so as to allow the subject to move voluntarily. This kind of mechanism encourages subjects to get rid of the limitations of the desired trajectory, which can be regarded as a reward and make them more active. But some subjects showed signs of slack behavior that they rely too much on the robot’s assistance to complete the task without any punishments.12 In other words, giving only rewards without punishments will cause subjects’ slackness in rehabilitation training. Therefore, it is necessary to develop control strategies exhibiting the reward–punishment feedback.
Wolbrecht et al. proposed an adaptive controller including a forgetting term to create the reward–punishment feedback mechanism.13 The adaptive law is made up of an error-based adaptive law and a forgetting law. The standard adaptive law dominates when there is a major tracking error so as to assist the subject to complete the task, while the forgetting law dominates when there is a minor tracking error so as to decay the assistance force to promote the subject’s active engagement, which forms a mechanism that gives a reward feedback to subjects by exhibiting a minor tracking error when they are highly engaged and gives a punishment feedback by exhibiting a major tracking error when they are slack. But the adaptive controller is model-based, it does not perform well when it is applied to wrist or finger rehabilitation because minor modeling deviations affect the wrist or finger much more than the upper limb. The tracking error will not change significantly regardless of the degree of the subject’s engagement.
Another improvement to the AAN control strategy was proposed by Pehilivan et al., who introduced a minimum AAN control strategy, which relied on Kalman filter to estimate subjects’ capability.14,15 According to the estimated results, the controller updates the derivative feedback gain to modify the bounds of allowable error on the desired trajectory, which also reflects the idea of reward–punishment feedback. Subsequently, Kalman filter was replaced by nonlinear disturbance observer, and the electromyography (EMG) sensors were used to estimate the subjects’ engagement.16
To sum up, in order to promote the engagement of subjects, the common feature of the above control strategies is that they can create a reward–punishment feedback mechanism according to the subjects’ current engagement or performance. To the best of our knowledge, previous researchers have not specifically identified this mechanism. More control strategies for rehabilitation robots support this point of view.17–25
In this article, we proposed a reward–punishment feedback control strategy to promote subjects’ engagement for wrist rehabilitation. Firstly, we utilize the energy contributed by the subject to estimate his/her engagement. The energy can be obtained by calculating the integral of the torque contributed by the subject against the position. Secondly, an adaptive controller including a reward–punishment term was proposed. Unlike the adaptive controller above,13 the included term is not constant. Instead, it updates based on the estimated results so that the controller can give reward or punishment feedback to subjects by reflecting different tracking error, which is suitable for wrist rehabilitation. Finally, the control strategy was demonstrated through experiments on healthy subjects without cardiovascular and cerebrovascular diseases operating a wrist rehabilitation robot. The contributions of this work include the development of an engagement estimated approach without any extra sensors, which greatly reduces development costs. This work also proposed an improved adaptive controller including a reward–punishment term for wrist rehabilitation, which has great potential to promote subjects’ engagement.
This article is organized as follows. The second section presents an engagement estimated approach based on energy information and a human robot coupled system modeling. The third section proposes an adaptive controller including a reward–punishment term and details the Lyapunov stability analysis. The fourth section introduces the specific implementation methods of three experiments. The fifth section presents and analyzes experimental results. Eventually, the discussion and conclusion are presented in the sixth section.
Engagement estimated based on energy information
We have developed a wrist rehabilitation robot, a three degree-of-freedom (DOF) device, as shown in Figure 1(a). The device is capable of independently actuating all three DOFs of subject’s forearm and wrist. Relatively, the device has three joints: flexion/extension joint, radial/ulnar deviation joint, and pronation/supination joint can all be controlled. Each joint of the device employs both a brushless DC motor with a conveyor belt to drive. Therefore, the control methods of the three joints are similar, and this article only describes the control strategy of the flexion/extension joint.
Background: Robotic rehabilitation of stroke survivors with upper extremity dysfunction yields different outcomes depending on the robot type. Considering that excessive dependence on assistive force provided by robots may interfere with the patient’s active learning and participation, we hypothesized that the use of an active-assistive robot does not lead to a more meaningful difference with respect to upper extremity rehabilitation than the use of an active robot. Accordingly, we aimed to evaluate the differences in the clinical and kinematic outcomes between active and active-assistive robotic rehabilitation among stroke survivors.
Methods: In this single-blinded randomized controlled trial, we assigned 20 stroke survivors with upper extremity dysfunction (Medical Research Council scale score, 3 or 4) to the active (ACT) and active-assistive (ACAS) robotic rehabilitation groups in a 1:1 ratio and administered 20 sessions of 30-minute robotic intervention (5 days/week, 4 weeks). The primary (Wolf Motor Function Test [WMFT]-score and -time: measures activity), and secondary (Fugl-Meyer Assessment [FMA] and Stroke Impact Scale [SIS] scores: measure impairment and participation, respectively; kinematic outcomes) outcome measures were determined at baseline, after 2 and 4 weeks of the intervention, and 4 weeks after the end of the intervention. Furthermore, we evaluated the usability of the robotic devices by conducting interviews with the patients, therapists, and physiatrists.
Results: In both the groups, the WMFT-score and -time improved over the course of the intervention. Time had a significant effect on the WMFT-score and -time, FMA-UE, FMA-prox, and SIS-strength; group × time interaction had a significant effect on SIS-function and SIS-social participation (all, p <0.05). The ACT group showed better improvement in participation and smoothness than the ACAS group. In contrast, the ACAS group exhibited better improvement in mean speed.
Conclusions: There were no differences between the two groups regarding the impairment and activity domains. However, the ACT robots were more beneficial than ACAS robots regarding participation and smoothness. Considering the high cost and complexity of ACAS robots, ACT robots may be more suitable for robotic rehabilitation in stroke survivors who can perform voluntary movement.
The purpose of this paper is to design and develop a new robotic device for the rehabilitation of the upper limbs. The authors are focusing on a new symmetrical robot which can be used to rehabilitate the right upper limb and the left upper limb. The robotic arm can be automatically extended or reduced depending on the measurements of the patient’s arm. The main idea is to integrate electrical stimulation into motor rehabilitation by robot. The goal is to provide automatic electrical stimulation based on muscle status during the rehabilitation process.
The developed robotic arm can be automatically extended or reduced depending on the measurements of the patient’s arm. The system merges two rehabilitation strategies: motor rehabilitation and electrical stimulation. The goal is to take the advantages of both approaches. Electrical stimulation is often used for building muscle through endurance, resistance and strength exercises. However, in the proposed approach the electrical stimulation is used for recovery, relaxation and pain relief. In addition, the device includes an electromyography (EMG) muscle sensor that records muscle activity in real time. The control architecture provides the ability to automatically activate the appropriate stimulation mode based on the acquired EMG signal. The system software provides two modes for stimulation activation: the manual preset mode and the EMG driven mode. The program ensures traceability and provides the ability to issue a patient status monitoring report.
The developed robotic device is symmetrical and reconfigurable. The presented rehabilitation system includes a muscle stimulator associated with the robot to improve the quality of the rehabilitation process. The integration of neuromuscular electrical stimulation into the physical rehabilitation process offers effective rehabilitation sessions for neuromuscular recovery of the upper limb. A laboratory-made stimulator is developed to generate three modes of stimulation: pain relief, massage and relaxation. Through the control software interface, the physiotherapist can set the exercise movement parameters, define the stimulation mode and record the patient training in real time.
There are certain constraints when applying the proposed method, such as the sensitivity of the acquired EMG signals. This involves the use of professional equipment and mainly the implementation of sophisticated algorithms for signal extraction.
Functional electrical stimulation and robot-based motor rehabilitation are the most important technologies applied in post-stroke rehabilitation. The main objective of integrating robots into the rehabilitation process is to compensate for the functions lost in people with physical disabilities. The stimulation technique can be used for recovery, relaxation and drainage and pain relief. In this context, the idea is to integrate electrical stimulation into motor rehabilitation based on a robot to obtain the advantages of the two approaches to further improve the rehabilitation process. The introduction of this type of robot also makes it possible to develop new exciting assistance devices.
The proposed design is symmetrical, reconfigurable and light, covering all the joints of the upper limbs and their movements. In addition, the developed platform is inexpensive and a portable solution based on open source hardware platforms which opens the way to more extensions and developments. Electrical stimulation is often used to improve motor function and restore loss of function. However, the main objective behind the proposed stimulation in this paper is to recover after effort. The novelty of the proposed solution is to integrate the electrical stimulation powered by EMG in robotic rehabilitation.
Abstract: In view of the urgent need for intelligent rehabilitation equipment for some disabled people, an intelligent, upper limb rehabilitation training robot is designed by applying the theories of artificial intelligence, information, control, human-machine engineering, and more. A new robot structure is proposed that combines the use of a flexible rope with an exoskeleton. By introducing environmentally intelligent ergonomics, combined with virtual reality, multi-channel information fusion interaction technology and big-data analysis, a collaborative, efficient, and intelligent remote rehabilitation system based on a human’s natural response and other related big-data information is constructed. For the multi-degree of the freedom robot system, optimal adaptive robust control design is introduced based on Udwdia-Kalaba theory and fuzzy set theory. The new equipment will help doctors and medical institutions to optimize both rehabilitation programs and their management, so that patients are more comfortable, safer, and more active in their rehabilitation training in order to obtain better rehabilitation results.
In rehabilitation of patients who have lost their ability to move independently due to the paralysis of lower limbs, using exoskeletons is a perspective direction. In recent years a great number of robotic devices improving walking of people with lower paraparesis have been developed. However, their comparison is hindered since there are no standardized approaches to the assessment of their efficiency and safety. In this review, general principles of evaluating external robotic devices have been presented, and methods of determining safety and convenience of exoskeleton usage have been analyzed. Assessment of qualitative and quantitative parameters of exoskeleton-assisted walking has also been considered. The characteristic of the questionnaires, standard tests and biochemical investigations, which are used in approbation of exoskeletal devices in people with paraplegia has been presented. Possible ways of evaluating energy expenditure when moving in exoskeletons are shown. The need of elaborating a unified evaluation strategy of walking in exoskeletons has been substantiated.
Introduction. Bioengineering devices, enhancing functional capabilities of patients with pathology of the musculoskeletal apparatus, include, among others, exoskeletons, which are special constructions that are put on a man in the form of an external frame, reproduce the biomechanics of his movements, improve muscular power, and reduce metabolic expenditure for walking [1–7]. In rehabilitation medicine, the development of exoskeletons for patients who have lost the ability to ambulate due to paralyzation of the lower limbs, is the most grounded and perspective [8–15]. A sufficient number of models of robotic orthoses and exoskeletons enabling patients with lower paraplegia and paraparesis to stand up and sit down, walk along an even surface and ascend stairs [16–27]. Creation and improvement of such systems require assessment of their efficiency and safety. Nevertheless there are not so many publications on this topic. The majority of these works touch upon more simple robotic devices compared to skeletons [28–31] often using different metrical sets [32–34]. Assessing the efficiency of a new robotic device with functional electrostimulation for patients with lower limb paraparesis, Goldfarb et al. analyzed an average walking speed, heart rate (HR), arterial pressure (AP), gas exchange, variability of the angles in the pelvic and knee joints [30, 35]. An average walking speed and HR normalized relative to the walking speed served as criteria of evaluation of orthoses for people with paraplegia in the works of Nene, Harvey, Winchester et al. [36–39]. Ohta et al. assessed a robotic orthosis designed for patients with vertebral-cerebrospinal trauma (VCST) using walking speed, step length, amplitude of vertical and lateral displacement of the head in walking . In some cases, in addition to the walking speed, the authors took into consideration the maximal distance the patient could travel without rest using the device [40–42], or the kinematics of motions in the knee or pelvic joints . Kobetic et al. studied the efficiency of a robotic orthosis intended for restoration of the capacity of persons with paraparesis to standing, walking and ascending stairs by analyzing the kinematics of motions in the knee joint . A short analysis of biomechanical parameters was presented also in the work of Jung et al.: the investigators performed a comparative analysis of gaits of patients with spinal cord traumas using robotic devices and without their assistance . In order to evaluate the efficiency of using exoskeletons for rehabilitation of stroke patients Fan et al. analyzed indices of surface electromyogram . Quantitative characteristic of the efficiency of an active lower limb exoskeleton in the work of Neuhaus et al. was given on the basis of walking speed, and an extent of efforts expended on the exoskeleton control was evaluated by registration of the HR, respiration rate, color of the skin; the authors assessed also the stability of standing (ability of the patient to catch a ball), and cognitive efforts (ability to maintain a visual contact) . Apart from walking, of patient’s capacity to sit down and stand up in the exoskeleton was estimated in some cases; for this purpose angles in the pelvic and knee joints , as well as pressure of the arms on the wheel-chair handles during these maneuvers were used [20, 21].
On the whole, it should be noted that a generally accepted methodology of exoskeleton assessment has not been worked out so far [47–50]. But the analysis of the literature showed, that a great deal of investigations are devoted to the elaboration of the general principles of approbation of novel robotic device, exoskeletons for lower limbs in particular. Approbation of exoskeletons usually includes testing of the walk in the exoskeleton and determination of such indices as energy expenditure, safety, convenience and simplicity of using the external device [24, 51]. These approaches to approbation are valid for all types of lower limb exoskeletons, making it possible to compare different variants of exoskeleton devices [52, 53].[…]
Robot-mediated therapy is an innovative form of rehabilitation that enables highly repetitive, intensive, adaptive, and quantifiable physical training. It has been increasingly used to restore loss of motor function, mainly in stroke survivors suffering from an upper limb paresis. Multiple studies collated in a growing number of review articles showed the positive effects on motor impairment, less clearly on functional limitations. After describing the current status of robotic therapy after upper limb paresis due to stroke, this overview addresses basic principles related to robotic therapy applied to upper limb paresis. We demonstrate how this innovation is an evidence-based approach in that it meets both the improved clinical and more fundamental knowledge-base about regaining effective motor function after stroke and the need of more objective, flexible and controlled therapeutic paradigms.
Robot-mediated rehabilitation is an innovative exercise-based therapy using robotic devices that enable the implementation of highly repetitive, intensive, adaptive, and quantifiable physical training. Since the first clinical studies with the MIT-Manus robot (1), robotic applications have been increasingly used to restore loss of motor function, mainly in stroke survivors suffering from an upper limb paresis but also in cerebral palsy (2), multiple sclerosis (3), spinal cord injury (4), and other disease types. Thus, multiple studies suggested that robot-assisted training, integrated into a multidisciplinary program, resulted in an additional reduction of motor impairments in comparison to usual care alone in different stages of stroke recovery: namely, acute (5–7), subacute (1, 8), and chronic phases after the stroke onset (9–11). Typically, patients engaged in the robotic therapy showed an impairment reduction of 5 points or more in the Fugl-Meyer assessment as compared to usual care. Of notice, rehabilitation studies conducted during the chronic stroke phase suggest that a 5-point differential represents the minimum clinically important difference (MCID), i.e., the magnitude of change that is necessary to produce real-world benefits for patients (12). These results were collated in multiple review articles and meta-analyses (13–17). In contrast, the advantage of robotic training over usual care in terms of functional benefit is less clear, but there are recent results that suggest how best to organize training to achieve superior results in terms of both impairment and function (18). Indeed, the use of the robotic tool has allowed us the parse and study the ingredients that should form an efficacious and efficient rehabilitation program. The aim of this paper is to provide a general overview of the current state of robotic training in upper limb rehabilitation after stroke, to analyze the rationale behind its use, and to discuss our working model on how to more effectively employ robotics to promote motor recovery after stroke.
Upper Extremity Robotic Therapy: Current Status
Robotic systems used in the field of neurorehabilitation can be organized under two basic categories: exoskeleton and end-effector type robots. Exoskeleton robotic systems allow us to accurately determine the kinematic configuration of human joints, while end-effector type robots exert forces only in the most distal part of the affected limb. A growing number of commercial robotic devices have been developed employing either configuration. Examples of exoskeleton type include the Armeo®Spring, Armeo®Power, and Myomo® and of end-effector type include the InMotion™, Burt®, Kinarm™ and REAplan®. Both categories enable the implementation of intensive training and there are many other devices in different stages of development or commercialization (19, 20).
The last decade has seen an exponential growth in both the number of devices as well as clinical trials. The results coalesced in a set of systematic reviews, meta-analyses (13–17) and guidelines such as those published by the American Heart Association and the Veterans Administration (AHA and VA) (21). There is a clear consensus that upper limb therapy using robotic devices over 30–60-min sessions, is safe despite the larger number of movement repetitions (14).
This technic is feasible and showed a high rate of eligibility; in the VA ROBOTICS (9, 11) study, nearly two thirds of interviewed stroke survivors were enrolled in the study. As a comparison the EXCITE cohort of constraint-induced movement therapy enrolled only 6% of the screened patients participated (22). On that issue, it is relevant to notice the admission criteria of both chronic stroke studies. ROBOTICS enrolled subjects with Fugl-Meyer assessment (FMA) of 38 or lower (out of 66) while EXCITE typically enrolled subjects with an FMA of 42 or higher. Duret and colleagues demonstrated that the target population, based on motor impairments, seems to be broader in the robotic intervention which includes patients with severe motor impairments, a group that typically has not seen much benefit from usual care (23). Indeed, Duret found that more severely impaired patients benefited more from robot-assisted training and that co-factors such as age, aphasia, and neglect had no impact on the amount of repetitive movements performed and were not contraindicated. Furthermore, all patients enrolled in robotic training were satisfied with the intervention. This result is consistent with the literature (24).
The main outcome result is that robotic therapy led to significantly more improvement in impairment as compared to conventional usual care, but only slightly more on motor function of the limb segments targeted by the robotic device (16). For example, Bertani et al. (15) and Zhang et al. (17) found that robotic training was more effective in reducing motor impairment than conventional usual care therapy in patients with chronic stroke, and further meta-analyses suggested that using robotic therapy as an adjunct to conventional usual care treatment is more effective than robotic training alone (13–17). Other examples of disproven beliefs: many rehabilitation professionals mistakenly expected significant increase of muscle hyperactivity and shoulder pain due to the intensive training. Most studies showed just the opposite, i.e., that intensive robotic training was associated with tone reduction as compared to the usual care groups (9, 25, 26). These results are shattering the resistance to the widespread adoption of robotic therapy as a therapeutic modality post-stroke.
That said, not all is rosy. Superior changes in functional outcomes were more controversial until the very last years as most studies and reviews concluded that robotic therapy did not improve activities of daily living beyond traditional care. One first step was reached in 2015 with Mehrholz et al. (14), who found that robotic therapy can provide more functional benefits when compared to other interventions however with a quality of evidence low to very low. 2018 may have seen a decisive step in favor of robotic as the latest meta-analysis conducted by Mehrholz et al. (27) concluded that robot-assisted arm training may improve activities of daily living in the acute phase after stroke with a high quality of evidence However, the results must be interpreted with caution because of the high variability in trial designs as evidenced by the multicenter study (28) in which robotic rehabilitation using the Armeo®Spring, a non-motorized device, was compared to self-management with negative results on motor impairments and potential functional benefits in the robotic group.
The Robot Assisted Training for the Upper Limb after Stroke (RATULS) study (29) might clarify things and put everyone in agreement on the topic. Of notice, RATULS goes beyond the Veterans Administration ROBOTICS with chronic stroke or the French REM_AVC study with subacute stroke. RATULS included 770 stroke patients and covered all stroke phases, from acute to chronic, and it included a positive meaningful control in addition to usual care.[…]
An FRR is developed for active and passive training using two PMs and an MR damper.
An underactuated mechanism is proposed for independent training of all finger joints.
Modelling of kinematics, statics and dynamics of the FRR is presented.
The motion and force properties of the FRR are experimentally evaluated.
This paper presents the development of a finger rehabilitation robot (FRR) for active and passive training to fulfill the requirements of different rehabilitation stages. In the design, an antagonistic pair of pneumatic muscles (PMs) are utilized to exert a bidirectional force for passive training, and a controllable magnetorheological (MR) damper is used to provide a damping force for active training. In this paper, first, a detailed illustration of the mechanical design of the FRR, including the driving, transmission and actuating mechanisms, and the damping device, is presented. Subsequently, the kinematic analysis and simulation are described, followed by the static and dynamic analysis of the designed FRR. This paper details the static force transfer of the transmission mechanism, and the establishment of dynamic equations for the passive training system. Finally, an experimental set-up is established, and several passive and active training experiments are conducted for the performance evaluation of the FRR prototype. The results validate the feasibility and stability of the developed FRR.
According to the structural design requirements of upper limb rehabilitation robot, the static analysis and harmonic response analysis are made in this paper. The structure of the upper limb rehabilitation robot is simplified in different ways according to the importance of each part of the structure. Selecting different element types and establishing finite element model Based on its analysis results, some unnecessary parts are removed and simplified, and the main forced parts are optimized by defining parameters to optimize and re-design the structure. The static analysis and harmonious response analysis of the modified model, comparing the structure before and after optimization, show that the modified structural mechanical properties are obviously improved and the design requirements are fully met.
A frequent consequence of stroke is a limited hand function. Numerous studies have shown, that repetitive passive training enhances the rehabilitation process. As there are high anthropometric variances in hand and finger anatomy, this contribution presents a custom-made hand rehabilitation robot. The individual design is proposed to ensure an ergonomic interface which allows long-time wearing. To provide a cost-effective production, we present an automated design process. The individual fingers are manufactured monolithically using the selective laser sintering of polyamide. The presented device is portable and can be used for training as well as for grasping things.
Proprioception or body awareness is an essential sense that aids in the neural control of movement. Proprioceptive impairments are commonly found in people with neurological conditions such as stroke and Parkinson’s disease. Such impairments are known to impact the patient’s quality of life. Robot-aided proprioceptive training has been proposed and tested to improve sensorimotor performance. However, such robot-aided exercises are implemented similar to many physical rehabilitation exercises, requiring task-specific and repetitive movements from patients. Monotonous nature of such repetitive exercises can result in reduced patient motivation, thereby, impacting treatment adherence and therapy gains. Gamification of exercises can make physical rehabilitation more engaging and rewarding. In this work, we discuss our ongoing efforts to develop a game that can accompany a robot-aided wrist proprioceptive training exercise.
Proprioception, the sense of body awareness, is essential for normal motor function. Proprioceptive deficits are common in neurological conditions [Coupar et al. 2012; Konczak et al. 2009]. Such deficits cause a decline in precision of goal-directed movements, and altered postural and spinal reflexes resulting in balance and gait problems [Rothwell et al. 1982]. Proprioceptive training is an intervention aiming to improve proprioceptive function [Aman et al. 2015]. Previous work has established the efficacy of a robot-aided proprioceptive training using WristBot [Elangovan et al. 2017, 2018, 2019]. The WristBot (Figure 1. Left) is a three degrees-of-freedom (3-DoF) exoskeleton robot that allows full range of motion (ROM), delivers precise haptic, position, and velocity stimuli at the wrist, and accurately encodes wrist position across time. Additional details about the WristBot can found in [Cappello et al. 2015].
Nevertheless, while the WristBot has demonstrated its efficacy, it shares a limitation that is often encountered in rehabilitation settings. In a clinical setting, patients are often required to perform task-specific and repetitive movements [Kwakkel et al. 1999]. Initial patient enthusiasm to complete such activities rapidly declines as a result of the monotonous nature of movements. Patient engagement can be improved by complementing therapy with a virtual environment (VE). Prior research has shown that users have favored exercises complemented with a VE rather than conventional approaches [Hoffman et al. 2014]. Thus, our project objective is to turn these tedious movements into an interactive VE experience.
2 GAMIFICATION OF PROPRIOCEPTIVE TRAINING
Gamification process accounted for two key considerations: (1) the game should foster patient motivation and attention (2) and be clinically meaningful. To address these objectives, we reviewed the literature on game development [Bond 2014; Fullerton 2018] and identified four essential components: (1) Variability, (2) Feedback, (3) Rewards, and (4) a Compelling Purpose. The user will be gradually exposed to increasing levels of difficulty, which will likely reduce user frustrations. The user will receive meaningful feedback on concurrent metrics (e.g., Optimal ROM), as well as on previous treatment sessions. During game progress, the user will be alerted about deviations from the target movement requirements. Achievement badges will be rewarded to the user upon reaching therapy milestones, such as target ROM. Lastly, to encourage game completion, we establish an interesting backstory and a meaningful character arc for our virtual avatars. The developed game will be adaptable based on the user’s current clinical status, thus, making the game clinically meaningful. The clinician will have the ability to prescribe exercises based on user needs such as 1 DoF vs 3 DoF movements, continuous vs discrete movements, and strength training vs mobility training. WristBot will provide supportive forces aiding the user to achieve therapy milestones.
Gamified exercise is being developed using the Unity Game Engine, Python and libraries which interface with the WristBot. The game closely resembles an endless runner type game (Figure 1. Right) and utilizes the WrsitBot’s 3-DoF functionality to interact with the VE. Wrist flexion, extension, and abduction can be used to traverse their environment. The remaining 3 movements will allow interactions with their VE in unique ways, such as opening/closing doors, crouching, and pulling levers. In the VE, coins are strategically placed to maximize and improve the use of available ROM. Upon contact with either a wall or obstacle, visual feedback will be provided in the form of avatar damage and coin deduction. Consequently, users achieve improved mobility.
In Python, the connection between Unity and the WristBot library is managed through the use of a local WebSocket, a protocol for two-way communication over a single Transmission Control Protocol (TCP) connection [Fette and Melnikov 2011]. Through the WebSocket, reciprocal data are transferred between the WristBot and Unity. For example, wrist kinematic data will be streamed to the game while game progress is being relayed to the WristBot library. Game progress data will be utilized to compute and deliver haptic feedback to the user. Haptic feedback provided in the form of haptic assistance will aid users to improve their available ROM, while haptic resistance will improve muscle strength within the desired ROM. The clinical motive of the game is to transition the user from use of haptic assistance to resistance during game play. WristBot will adapt haptic feedback based on time spent and progress achieved in game play.
3 USABILITY TESTING
Usability testing will be conducted to ensure proper game usage by the clinical population and healthcare professionals. Specifically, the usability testing will evaluate areas such as 1) ease of game play, 2) game efficiency, and 3) user engagement. We will test the assumptions in each of these areas are accurately depicted in game development and met during game play. For example, we expect online visual feedback of deviations from target to help user focus on achieving the movement requirements. The users will be asked to verify the benefits of visual feedback in modifying their movements. Similarly, other assumptions such as performance badges and coins as rewards, and increase in difficulty levels will be evaluated. A common pitfall of usability studies involving physical rehabilitation setting is not recruiting from the representative population, most notably elderly population [Laver et al. 2017] as age has been shown to interfere with interactions in VE [Meldrum et al. 2012]. Therefore, to ensure our game is intuitive, we will recruit representative users from our patient populations.
This project was supported by National Science Foundation Partnerships For Innovation Technology Translation Award to Jürgen Konczak (1919036). Christopher Curry was supported by National Research Trainee-Understanding the Brain: Graduate Training Program in Sensory Science: Optimizing the Information Available for Mind and Brain (1734815).
Joshua E Aman, Naveen Elangovan, I Yeh, Jürgen Konczak, et al. 2015. The effectiveness of proprioceptive training for improving motor function: a systematic review. Frontiers in human neuroscience 8 (2015), 1075.
Jeremy Gibson Bond. 2014. Introduction to Game Design, Prototyping, and Development: From Concept to Playable Game with Unity and C. Addison-Wesley Professional.
Leonardo Cappello, Naveen Elangovan, Sara Contu, Sanaz Khosravani, Jürgen Konczak, and Lorenzo Masia. 2015. Robot-aided assessment of wrist proprioception. Frontiers in human neuroscience 9 (2015), 198.
Fiona Coupar, Alex Pollock, Phil Rowe, Christopher Weir, and Peter Langhorne. 2012. Predictors of upper limb recovery after stroke: a systematic review and meta-analysis. Clinical rehabilitation 26, 4 (2012), 291–313.
Naveen Elangovan, Leonardo Cappello, Lorenzo Masia, Joshua Aman, and Jürgen Konczak. 2017. A robot-aided visuo-motor training that improves proprioception and spatial accuracy of untrained movement. Scientific reports 7, 1 (2017), 17054.
Naveen Elangovan, Paul Tuite, and Jürgen Konczak. 2018. Somatosensory training improves proprioception and untrained motor function in Parkinson’s disease. Frontiers in neurology 9(2018), 1053.
Naveen Elangovan, I-Ling Yeh, Jessica Holst-Wolf, and Jürgen Konczak. 2019. A robot-assisted sensorimotor training program can improve proprioception and motor function in stroke survivors. In 2019 IEEE 16th International Conference on Rehabilitation Robotics (ICORR). IEEE, 660–664.
I. Fette and A. Melnikov. 2011. The WebSocket Protocol. Technical Report.
Tracy Fullerton. 2018. Game design workshop: a playcentric approach to creating innovative games. AK Peters/CRC Press.
Hunter G Hoffman, Walter J Meyer III, Maribel Ramirez, Linda Roberts, Eric J Seibel, Barbara Atzori, Sam R Sharar, and David R Patterson. 2014. Feasibility of articulated arm mounted Oculus Rift Virtual Reality goggles for adjunctive pain control during occupational therapy in pediatric burn patients. Cyberpsychology, Behavior, and Social Networking 17, 6(2014), 397–401.
Jürgen Konczak, Daniel M Corcos, Fay Horak, Howard Poizner, Mark Shapiro, Paul Tuite, Jens Volkmann, and Matthias Maschke. 2009. Proprioception and motor control in Parkinson’s disease. Journal of motor behavior 41, 6 (2009), 543–552.
Gert Kwakkel, Boudewijn J Kollen, and Robert C Wagenaar. 1999. Therapy impact on functional recovery in stroke rehabilitation: a critical review of the literature. Physiotherapy 85, 7 (1999), 377–391.
Kate E Laver, Belinda Lange, Stacey George, Judith E Deutsch, Gustavo Saposnik, and Maria Crotty. 2017. Virtual reality for stroke rehabilitation. Cochrane database of systematic reviews11 (2017).
Dara Meldrum, Aine Glennon, Susan Herdman, Deirdre Murray, and Rory McConn-Walsh. 2012. Virtual reality rehabilitation of balance: assessment of the usability of the Nintendo Wii® Fit Plus. Disability and rehabilitation: assistive technology 7, 3 (2012), 205–210.
JC Rothwell, MM Traub, BL Day, JA Obeso, PK Thomas, and CD Marsden. 1982. Manual motor performance in a deafferented man. Brain 105, 3 (1982), 515–542.