Robot-aided Therapy Research Articles

An End-Effector Type Therapeutic Robot for Home-Based Upper Limb Rehabilitation

<h3>Research Objectives</h3> This research proposes a telerehabilitation framework with a developed desktop-mounted rehabilitation robot (DMRbot) and PTC's Industrial Internet of Things (IIoT) platform to remotely provide home-based passive rehabilitation therapies to individuals with upper limb dysfunctions. <h3>Design</h3> Experimental study. <h3>Setting</h3> In the BioRobotics Lab. <h3>Participants</h3> An experiment provided a passive mode of telerehabilitation to two healthy male human subjects (age: 26-30 years, weights: 70–80 kg, and height: 1.6–1.7 m). <h3>Interventions</h3> While providing robot-aided therapy, rehab robot data (for example, joint angle, torque) and therapists' joystick commands are transported by ThingWorx, which are then processed to get information such as achievement of cures. Participants were seated on a chair, wearing gripper gloves attached to the rehab robot's end-effector. The therapist leveraged the digital twin structure, facilitated by Vuforia studio, to visualize the physical robot motions happening in the remote place. <h3>Main Outcome Measures</h3> We evaluated the device and the ability of the therapist to provide telerehabilitation, meaning how well the patient hand followed the trajectory directed by the therapist remotely. An additional measure was the latency, the time difference between sending joint angle parameters, and the movement of physical joints of the network. <h3>Results</h3> The proposed end-effector type therapeutic robot for home-based upper limb rehabilitation resulted in 100% trajectory movement of the patient's hand, with a network latency of approximately 0.15 seconds. <h3>Conclusions</h3> Therefore, the proposed framework is inclined to make therapies more approachable to remote areas with convenience and affordability. <h3>Author(s) Disclosures</h3> The authors have no conflicts of interest to declare.

Upper-Limb Rehabilitation Robot: A Measure of Intent Detection for Robot Aided Active Therapy Exercise

<h3>Research Objectives</h3> In robot-aided rehabilitation of post-stroke patients with upper limb impairment, robotic device must allow patient's effort in performing exercises. To estimate such effort, user's force exerted on robotic device can be used during active exercises. To validate use of force exerted by user/wearer by comparing with EMG signals as a measure of intent detection in robot-aided active therapy on rehabilitation robot, in this research, user's interactive force with robot was measured and compared with user's muscle activity (EMG) during elbow flexion-extension performed by an upper-limb rehabilitation robot (u-Rob). This physiological measure will help design rehab robot centering patient's need and level of impairment. Furthermore, it can give an indication of spastic behavior during rehab exercises. <h3>Design</h3> Experimental study of robot-aided therapy to observe force values with EMG values, leading to use force as a measure of intent detection. <h3>Setting</h3> Protocol: Elbow flexion/extension (0-120 deg) motion. Experiments (passive and active therapy exercise, five trials/exercise) were conducted with healthy subject in seated position wearing the robot. <h3>Participants</h3> 2 healthy males and 1 healthy female. <h3>Interventions</h3> Elbow flexion/extension (0-120 deg) motion. Experiments (passive and active therapy exercise, five trials) were conducted subject in seated position wearing the robot. <h3>Main Outcome Measures</h3> Force exerted by the participant and participant's EMG <h3>Results</h3> Significant amount of force is exerted during active exercise and significant amount of EMG activity is observed during active exercise. Besides, looking at the torque, it is seen that during active rehab exercise, robot has to contribute lesser torque than in passive exercise. <h3>Conclusions</h3> User-robot interactive forcesdetected by sensor can be used as a measure of intent detection during robot-aided active exercise. Moreover, measured forces can also be observed from muscle activity. It may give an idea of subject's spastic behavior. <h3>Author(s) Disclosures</h3> N/A.

Engagement Enhancement Based on Bayesian Optimization for Adaptive Assist-as-Needed Controller

Assist-as-needed (AAN) control strategy is applied in robot-aided therapy to promote the subject's engagement with minimal assistance. Current methods are based on physical measurements or employ a model to estimate the subject's engagement, which indirectly reflects the physiology of the subject or needs a complicated modeling process. Besides, manual adjustments are required to tune the parameters of the AAN controller to adapt to the estimated engagement of subjects. Such two drawbacks of previous studies have limited their effectiveness for promoting the subject's engagement during the training. In this paper, we propose an engagement enhancement method based on Bayesian Optimization for the adaptive AAN controller. Firstly, we evaluate the subject's engagement by using sEMG-based muscle activation, which directly relates to the physiology of the subject and eliminates the need for complicated modeling of engagement. Next, Bayesian optimization, which is tolerant of noisy measurements and human adaptation, is used to search for the optimal parameter of the AAN controller for each training trial. Finally, we conduct a study with sixteen healthy human subjects to verify the effectiveness of the proposed method. The experimental results showed that subjects’ engagement in the experiment group improved more significantly and was maintained at a higher level than that in the control group during the training. The performance improvement after the training obtained in the experiment group was higher than that in the control group (45.42% v.s. 30.40%). The experimental results indicate that the proposed method may be promising for transferring to the rehabilitation of post-stroke patients.

Analysis of Compensatory Movements Using a Supernumerary Robotic Hand for Upper Limb Assistance.

Recently, extratheses, aka Supernumerary Robotic Limbs (SRLs), are emerging as a new trend in the field of assistive and rehabilitation devices. We proposed the SoftHand X, a system composed of an anthropomorphic soft hand extrathesis, with a gravity support boom and a control interface for the patient. In preliminary tests, the system exhibited a positive outlook toward assisting impaired people during daily life activities and fighting learned-non-use of the impaired arm. However, similar to many robot-aided therapies, the use of the system may induce side effects that can be detrimental and worsen patients' conditions. One of the most common is the onset of alternative grasping strategies and compensatory movements, which clinicians absolutely need to counter in physical therapy. Before embarking in systematic experimentation with the SoftHand X on patients, it is essential that the system is demonstrated not to lead to an increase of compensation habits. This paper provides a detailed description of the compensatory movements performed by healthy subjects using the SoftHand X. Eleven right-handed healthy subjects were involved within an experimental protocol in which kinematic data of the upper body and EMG signals of the arm were acquired. Each subject executed tasks with and without the robotic system, considering this last situation as reference of optimal behavior. A comparison between two different configurations of the robotic hand was performed to understand if this aspect may affect the compensatory movements. Results demonstrated that the use of the apparatus reduces the range of motion of the wrist, elbow and shoulder, while it increases the range of the trunk and head movements. On the other hand, EMG analysis indicated that muscle activation was very similar among all the conditions. Results obtained suggest that the system may be used as assistive device without causing an over-use of the arm joints, and opens the way to clinical trials with patients.

Motor improvement estimation and task adaptation for personalized robot-aided therapy: a feasibility study

BackgroundIn the past years, robotic systems have become increasingly popular in upper limb rehabilitation. Nevertheless, clinical studies have so far not been able to confirm superior efficacy of robotic therapy over conventional methods. The personalization of robot-aided therapy according to the patients’ individual motor deficits has been suggested as a pivotal step to improve the clinical outcome of such approaches.MethodsHere, we present a model-based approach to personalize robot-aided rehabilitation therapy within training sessions. The proposed method combines the information from different motor performance measures recorded from the robot to continuously estimate patients’ motor improvement for a series of point-to-point reaching movements in different directions. Additionally, it comprises a personalization routine to automatically adapt the rehabilitation training. We engineered our approach using an upper-limb exoskeleton. The implementation was tested with 17 healthy subjects, who underwent a motor-adaptation paradigm, and two subacute stroke patients, exhibiting different degrees of motor impairment, who participated in a pilot test undergoing rehabilitative motor training.ResultsThe results of the exploratory study with healthy subjects showed that the participants divided into fast and slow adapters. The model was able to correctly estimate distinct motor improvement progressions between the two groups of participants while proposing individual training protocols. For the two pilot patients, an analysis of the selected motor performance measures showed that both patients were able to retain the improvements gained during training when reaching movements were reintroduced at a later stage. These results suggest that the automated training adaptation was appropriately timed and specifically tailored to the abilities of each individual.ConclusionsThe results of our exploratory study demonstrated the feasibility of the proposed model-based approach for the personalization of robot-aided rehabilitation therapy. The pilot test with two subacute stroke patients further supported our approach, while providing encouraging results for the applicability in clinical settings.Trial registration This study is registered in ClinicalTrials.gov (NCT02770300, registered 30 March 2016, https://clinicaltrials.gov/ct2/show/NCT02770300)

A Novel End-Effector Robot System Enabling to Monitor Upper-Extremity Posture During Robot-Aided Planar Reaching Movements

End-effector type robots have been popularly applied to robot-aided therapy for rehabilitation purpose. However, those robots have a key drawback for the purpose: lack of the user's posture (joint angle) information. This letter proposes a novel end-effector rehabilitation robot system that contains a contactless motion sensor to monitor upper- extremity posture during robot-aided reaching exercise. The sensor allows the posture estimation without complicated procedures but has an inaccuracy problem such as occlusion and an unreliable segment length. Therefore, we developed a posture monitoring method, which is an analytical method without training procedure, based on the combined use of the information obtained from the sensor and the robot. Eight healthy subjects participated in the experiment with planar reaching exercise for validation. The results of joint angle estimation, high correlation coefficient (0.95 ± 0.03) and small errors (3.55 ± 0.70 deg), show that the proposed system can provide affordable upper-extremity posture estimation.

Clinical and kinematic evaluation of the H-Man arm robot for post-stroke upper limb rehabilitation: Preliminary findings of a randomised controlled trial

The H-Man robot, a table-top, portable, 2D planar, end-effector with virtual reality feedback was designed to deliver self-paced, repetitive reaching arm movements. Preliminary results of a randomized clinical trial of 26/44 strokes with hemiparetic arm weakness are presented. Inclusion criteria included first-stroke, > 4 months duration with Fugl Meyer Assessment Scale (FMA) 20–50/66) without contraindications to robot-aided therapy. Following informed consent, subjects were randomized into 2 groups: H-Man-conventional (HCT) group received 18 sessions over 6 weeks of 60 minutes of H-man training then 30 minutes of conventional therapy (CT), while control group (CG) received a similar intensity of 90 minutes of CT. Blinded outcome assessments at weeks 0 (baseline), 3, 6 (end-training), 12 and 24 (follow-up). The primary outcome measure was FMA change at week 6. Parametric analysis was used and level of significance was P < 0.05. Altogether, data from 26 out of 44 subjects were analyzed. (13 HCT, 13 CG). Mean age was 54.0 years (SD 10.9), 14/26 were male, 15/26 had hemorrhagic strokes, mean stroke duration 227.2 days (SD 207.2), and mean baseline FMA 38.6 (SD 11.1). The HCT group achieved significantly better FMA gains compared to CG (4.15 HCT vs. 1.69 CG, P = 0.03) at week 6 (post-training), and at week 24 (5.77 HCT vs 2.61 CG, P = 0.03). There were no adverse side effects or drop outs. Robotic kinematic measures of line and circle temporal tracing correlated with FMA scores at Week 0. Combinatory arm rehabilitation with H-Man robot was superior to CT and well tolerated.

The effect of robot therapy assisted by surface EMG on hand recovery in post-stroke patients. A pilot study

Background: Hemiparesis caused by a stroke negatively limits a patient’s motor function. Nowadays, innovative technologies such as robots are commonly used in upper limb rehabilitation. The main goal of robot-aided therapy is to provide a maximum number of stimuli in order to stimulate brain neuroplasticity. Treatment applied in this study via the AMADEO robot aimed to improve finger flexion and extension. Aim: To assess the effect of rehabilitation assisted by a robot and enhanced by surface EMG. Research project: Before-after study design. Materials and methods: The study group consisted of 10 post-stroke patients enrolled for therapy with the AMADEO robot for at least 15 sessions. At the beginning and at the end of treatment, the following tests were used for clinical assessment: Fugl-Meyer scale, Box and Block test and Nine Hole Peg test. In the present study, we used surface electromyography (sEMG) to maintain optimal kinematics of hand motion. Whereas sensorial feedback, provided by the robot, was vital in obtaining closed-loop control. Thus, muscle contraction was transmitted to the amplifier through sEMG, activating the mechanism of the robot. Consequentially, sensorial feedback was provided to the patient. Results: Statistically significant improvement of upper limb function was observed in: Fugl-Meyer (p = 0.38) and Box and Block (p = 0.27). The Nine Hole Peg Test did not show statistically significant changes in motor skills of the hand. However, the functional improvement was observed at the level of 6% in the Fugl-Meyer, 15% in the Box and Block, and 2% in the Nine Hole Peg test. Conclusions: Results showed improvement in hand grasp and overall function of the upper limb. Due to sEMG, it was possible to implement robot therapy in the treatment of patients with severe hand impairment.

Treadmill vs. overground walking: different response to physical interaction.

Rehabilitation of human motor function is an issue of growing significance, and human-interactive robots offer promising potential to meet the need. For the lower extremity, however, robot-aided therapy has proven challenging. To inform effective approaches to robotic gait therapy, it is important to better understand unimpaired locomotor control: its sensitivity to different mechanical contexts and its response to perturbations. The present study evaluated the behavior of 14 healthy subjects who walked on a motorized treadmill and overground while wearing an exoskeletal ankle robot. Their response to a periodic series of ankle plantar flexion torque pulses, delivered at periods different from, but sufficiently close to, their preferred stride cadence, was assessed to determine whether gait entrainment occurred, how it differed across conditions, and if the adapted motor behavior persisted after perturbation. Certain aspects of locomotor control were exquisitely sensitive to walking context, while others were not. Gaits entrained more often and more rapidly during overground walking, yet, in all cases, entrained gaits synchronized the torque pulses with ankle push-off, where they provided assistance with propulsion. Furthermore, subjects entrained to perturbation periods that required an adaption toward slower cadence, even though the pulses acted to accelerate gait, indicating a neural adaptation of locomotor control. Lastly, during 15 post-perturbation strides, the entrained gait period was observed to persist more frequently during overground walking. This persistence was correlated with the number of strides walked at the entrained gait period (i.e., longer exposure), which also indicated a neural adaptation.NEW & NOTEWORTHY We show that the response of human locomotion to physical interaction differs between treadmill and overground walking. Subjects entrained to a periodic series of ankle plantar flexion torque pulses that shifted their gait cadence, synchronizing ankle push-off with the pulses (so that they assisted propulsion) even when gait cadence slowed. Entrainment was faster overground and, on removal of torque pulses, the entrained gait period persisted more prominently overground, indicating a neural adaptation of locomotor control.

Exercise as Medicine for the Treatment of Brain Dysfunction: Evidence for Cortical Stroke, Cerebellar Ataxia, and Parkinson’s Disease

This review addresses the role of exercise as an intervention for treating neurological disease. It focuses on three major neurological diseases that either present in acute or neurodegenerative forms—Parkinson’s disease, cerebellar ataxia, and cortical stroke. Each of the diseases affects primarily different brain structures, namely the basal ganglia, the cerebellum, and the cerebrum. These structures are all known to be involved in motor control, and the dysfunction of each structure leads to distinct movement deficits. The review summarizes current knowledge on how exercise can aid rehabilitation or therapeutic efforts. In addition, it addresses the role of robotic devices in enhancing available therapies by reviewing how robot-aided therapies may promote the recovery for stroke survivors. It highlights recent scientific evidence in support of exercise as a treatment for brain dysfunction, but also outlines the still open challenges for unequivocally demonstrating the benefits of exercise.

A Kinematic Study of Combined Robot-Aided Arm Therapy and Constraint-Induced Therapy in Chronic Stroke

A Kinematic Study of Combined Robot-Aided Arm Therapy and Constraint-Induced Therapy in Chronic Stroke

Electroencephalographic markers of robot-aided therapy in stroke patients for the evaluation of upper limb rehabilitation.

Stroke is the leading cause of permanent disability in developed countries; its effects may include sensory, motor, and cognitive impairment as well as a reduced ability to perform self-care and participate in social and community activities. A number of studies have shown that the use of robotic systems in upper limb motor rehabilitation programs provides safe and intensive treatment to patients with motor impairments because of a neurological injury. Furthermore, robot-aided therapy was shown to be well accepted and tolerated by all patients; however, it is not known whether a specific robot-aided rehabilitation can induce beneficial cortical plasticity in stroke patients. Here, we present a procedure to study neural underpinning of robot-aided upper limb rehabilitation in stroke patients. Neurophysiological recordings use the following: (a) 10-20 system electroencephalographic (EEG) electrode montage; (b) bipolar vertical and horizontal electrooculographies; and (c) bipolar electromyography from the operating upper limb. Behavior monitoring includes the following: (a) clinical data and (b) kinematic and dynamic of the operant upper limb movements. Experimental conditions include the following: (a) resting state eyes closed and eyes open, and (b) robotic rehabilitation task (maximum 80 s each block to reach 4-min EEG data; interblock pause of 1 min). The data collection is performed before and after a program of 30 daily rehabilitation sessions. EEG markers include the following: (a) EEG power density in the eyes-closed condition; (b) reactivity of EEG power density to eyes opening; and (c) reactivity of EEG power density to robotic rehabilitation task. The above procedure was tested on a subacute patient (29 poststroke days) and on a chronic patient (21 poststroke months). After the rehabilitation program, we observed (a) improved clinical condition; (b) improved performance during the robotic task; (c) reduced delta rhythms (1-4 Hz) and increased alpha rhythms (8-12 Hz) during the resting state eyes-closed condition; (d) increased alpha desynchronization to eyes opening; and (e) decreased alpha desynchronization during the robotic rehabilitation task. We conclude that the present procedure is suitable for evaluation of the neural underpinning of robot-aided upper limb rehabilitation.

Kinematic reconstruction of the human arm joints in robot-aided therapies with Hermes robot.

This paper presents a kinematic reconstruction algorithm for the variables of the human arm joints in robot-aided neurorehabilitation therapies. The presented algorithm uses the end effector of a rehabilitation robot and an accelerometer placed onto the upper arm to compute accurate values of the human arm chain. The goal of this algorithm is to obtain the joint values of the patient's arm to provide objective information to the therapist about the progress of the patient and to study the effectiveness of these kind of therapies.

Proof of Concept for Robot-Aided Upper Limb Rehabilitation Using Disturbance Observers

This paper presents a wearable upper body exoskeleton system with a model-based compensation control framework to support robot-aided shoulder–elbow rehabilitation and power assistance tasks. To eliminate the need for EMG and force sensors, we exploit off-the-shelf compensation techniques developed for robot manipulators. Thus, target rehabilitation tasks are addressed by using only encoder readings. A proof-of-concept evaluation was conducted with five able-bodied participants. The patient-active rehabilitation task was realized via observer-based user torque estimation, in which resistive forces were adjusted using virtual impedance. In the patient-passive rehabilitation task, the proposed controller enabled precise joint tracking with a maximum positioning error of 0.25 $^\circ$ . In the power assistance task, the users’ muscular activities were reduced up to 85% while exercising with a 5 kg dumbbell. Therefore, the exoskeleton system was regarded as being useful for the target tasks, indicating that it has a potential to promote robot-aided therapy protocols.

Auto-adaptative Robot-aided Therapy based in 3D Virtual Tasks controlled by a Supervised and Dynamic Neuro-Fuzzy System

This paper presents an application formed by a classification method based on the architecture of ART neural network (Adaptive Resonance Theory) and the Fuzzy Set Theory to classify physiological reactions in order to automatically and dynamically adapt a robot-assisted rehabilitation therapy to the patient needs, using a three-dimensional task in a virtual reality system. Firstly, the mathematical and structural model of the neuro-fuzzy classification method is described together with the signal and training data acquisition. Then, the virtual designed task with physics behavior and its development procedure are explained. Finally, the general architecture of the experimentation for the auto-adaptive therapy is presented using the classification method with the virtual reality exercise.

Wrist rehabilitation in chronic stroke patients by means of adaptive, progressive robot-aided therapy

Despite distal arm impairment after brain injury is an extremely disabling consequence of neurological damage, most studies on robotic therapy are mainly focused on recovery of proximal upper limb motor functions, routing the major efforts in rehabilitation to shoulder and elbow joints. In the present study we developed a novel therapeutic protocol aimed at restoring wrist functionality in chronic stroke patients. A haptic three DoFs (degrees of freedom) robot has been used to quantify motor impairment and assist wrist and forearm articular movements: flexion/extension (FE), abduction/adduction (AA), pronation/supination (PS). This preliminary study involved nine stroke patients, from a mild to severe level of impairment. Therapy consisted in ten 1-hour sessions over a period of five weeks. The novelty of the approach was the adaptive control scheme which trained wrist movements with slow oscillatory patterns of small amplitude and progressively increasing bias, in order to maximize the recovery of the active range of motion. The primary outcome was a change in the active RoM (range of motion) for each DoF and a change of motor function, as measured by the Fugl-Meyer assessment of arm physical performance after stroke (FMA). The secondary outcome was the score on the Wolf Motor Function Test (WOLF). The FMA score reported a significant improvement (average of 9.33±1.89 points), revealing a reduction of the upper extremity motor impairment over the sessions; moreover, a detailed component analysis of the score hinted at some degree of motor recovery transfer from the distal, trained parts of the arm to the proximal untrained parts. WOLF showed an improvement of 8.31±2.77 points, highlighting an increase in functional capability for the whole arm. The active RoM displayed a remarkable improvement. Moreover, a three-months follow up assessment reported long lasting benefits in both distal and proximal arm functionalities. The experimental results of th- s preliminary clinical study provide enough empirical evidence for introducing the novel progressive, adaptive, gentle robotic assistance of wrist movements in the clinical practice, consolidating the evaluation of its efficacy by means of a controlled clinical trial.

Clinical experimental research on adaptive robot-aided therapy control methods for upper-limb rehabilitation

SUMMARYThis study presents novel robotic therapy control algorithms for upper-limb rehabilitation, using newly developed passive and progressive resistance therapy modes. A fuzzy-logic based proportional-integral-derivative (PID) position control strategy, integrating a patient's biomechanical feedback into the control loop, is proposed for passive movements. This allows the robot to smoothly stretch the impaired limb through increasingly rigorous training trajectories. A fuzzy adaptive impedance force controller is addressed in the progressive resistance muscle strength training and the adaptive resistive force is generated according to the impaired limb's muscle strength recovery level, characterized by the online estimated impaired limb's bio-damping and bio-stiffness. The proposed methods are verified with a custom constructed therapeutic robot system featuring a Barrett WAM™ compliant manipulator. Twenty-four recruited stroke subjects were randomly allocated in experimental and control groups and enrolled in a 20-week rehabilitation training program. Preliminary results show that the proposed therapy control strategies can not only improve the impaired limb's joint range of motion but also enhance its muscle strength.

Anthropomorphic mechanical design and Lyapunov-based control of a new shoulder rehabilitation system

Article history: Received January 20, 2014 Received in Revised form April, 10, 2014 Accepted 7 May 2014 Available online 9 May 2014 Stroke is one of the main causes of disability. It affects millions of people worldwide. One symptom of stroke is disabled arm function. Restoration of arm function is necessary to resuming activities of daily living (ADL). Along with traditional rehabilitation techniques, robot-aided therapy has emerged recently. The control schemes of rehabilitation robots are designed for two reasons. First they are designed for passive rehabilitation in which the robot guides the patient's limb through a predefined path and second for active rehabilitation in which the patient initiates the movement and is partially assisted or resisted by the robotic device. This paper introduces a new robot for shoulder rehabilitation. The Shoulder Rehabilitation System (SRS) has three degrees of freedom (DOFs) for three rotational DOFs of the shoulder but additional translational DOFs of the shoulder are also allowed to avoid discomfort to the patient. A new open circular mechanism is proposed for the third joint that solves the known issues for rehabilitation robots such as long wiring and discomfort associated with closed mechanisms. Lyapunov-based controller with integral action is proposed to guide the robot through a predefined trajectory. Simulation results proved that the proposed controller can track the desired trajectory; reject constant bounded disturbance to the system and is robust due to its nonlinear nature. The proposed controller is designed to be used in passive rehabilitation. © 2014 Growing Science Ltd. All rights reserved.

Auto-adaptive robot-aided therapy using machine learning techniques

This paper presents an application of a classification method to adaptively and dynamically modify the therapy and real-time displays of a virtual reality system in accordance with the specific state of each patient using his/her physiological reactions. First, a theoretical background about several machine learning techniques for classification is presented. Then, nine machine learning techniques are compared in order to select the best candidate in terms of accuracy. Finally, first experimental results are presented to show that the therapy can be modulated in function of the patient state using machine learning classification techniques.

The effect of involuntary motor activity on myoelectric pattern recognition: a case study with chronic stroke patients

Objective. This study investigates the effect of the involuntary motor activity of paretic-spastic muscles on the classification of surface electromyography (EMG) signals. Approach. Two data collection sessions were designed for 8 stroke subjects to voluntarily perform 11 functional movements using their affected forearm and hand at relatively slow and fast speeds. For each stroke subject, the degree of involuntary motor activity present in the voluntary surface EMG recordings was qualitatively described from such slow and fast experimental protocols. Myoelectric pattern recognition analysis was performed using different combinations of voluntary surface EMG data recorded from the slow and fast sessions. Main results. Across all tested stroke subjects, our results revealed that when involuntary surface EMG is absent or present in both the training and testing datasets, high accuracies (>96%, >98%, respectively, averaged over all the subjects) can be achieved in the classification of different movements using surface EMG signals from paretic muscles. When involuntary surface EMG was solely involved in either the training or testing datasets, the classification accuracies were dramatically reduced (<89%, <85%, respectively). However, if both the training and testing datasets contained EMG signals with the presence and absence of involuntary EMG interference, high accuracies were still achieved (>97%). Significance. The findings of this study can be used to guide the appropriate design and implementation of myoelectric pattern recognition based systems or devices toward promoting robot-aided therapy for stroke rehabilitation.