Application of a Dual Upper Limb Task-Oriented Robotic System for the Functional Recovery of the Upper Limb in Stroke Patients

Summary

This experimental protocol outlines the use of a dual upper limb task-oriented robotic system for stroke patients with upper limb dysfunction. The findings indicate that this system can significantly improve stroke patients' upper limb function and daily living activities.

Abstract

Highly repetitive and task-oriented training has been shown to promote the recovery of limb function in stroke patients. Additionally, bilateral arm training can help stroke survivors regain their upper limb function and improve their daily activities. The dual upper limb task-oriented robotic system is designed to assist the healthy side of the stroke patient in driving the affected side to perform bilateral arm training through the use of a robotic device. It can also guide the patient in carrying out dual upper limb coordinated movements and engage them in a task-oriented virtual game using force feedback and human-computer interaction technology. This study aimed to assess the efficacy of the system in enhancing upper limb function and activities of daily living in stroke patients. The assessment methods used included motor evoked potential (MEP), functional test for the hemiplegic upper extremity-Hong Kong (FTHUE-HK), Fugl-Meyer Assessment Upper Extremity Scale (FMA-UE), and modified Barthel index (MBI). The results of the study indicate that the dual upper limb task-oriented robotic system can significantly improve the corticospinal pathway, upper limb function, and activities of daily living in stroke patients after 6 weeks of treatment. This system can serve as an effective adjunct to upper limb functional rehabilitation in stroke survivors, reducing the dependence on rehabilitation therapists. In conclusion, the dual upper limb task-oriented robotic system provides a new strategy for post-stroke limb functional rehabilitation and holds great potential for application, as it offers certain social and financial benefits.

Introduction

Stroke is one of the major causes of disability and the second leading cause of death globally^1,2. Stroke patients often face various challenges, such as motor, sensory, and cognitive deficits³. Upper limb dysfunction is a common problem after stroke, characterized by muscle weakness, spasticity, and reduced motor ability of the upper limb on the hemiplegic side⁴. It is reported to be present in more than 70% of stroke patients, and only around 5% recover to normal, while 20% regain some upper limb capabilities⁵. More than half of human life requires the participation of the upper limbs⁶, and upper limb dysfunction after a stroke severely affects patients' activities of daily living⁷, significantly decreasing their quality of life⁸ and increasing their financial burden⁹. Therefore, it is particularly important to explore effective methods of upper limb functional rehabilitation.

Various clinical upper limb rehabilitation treatments, such as mirror therapy, constraint-induced movement therapy, functional electrical stimulation, and other active or passive training, are commonly utilized for stroke patients^3,10. In recent years, bilateral arm training has garnered increased attention^6,11,12. It has been demonstrated to enhance neural connectivity between the sensorimotor areas of both ipsilateral and contralateral hemispheres¹². This type of training helps correct abnormalities in interhemispheric inhibition, facilitates reorganization of brain functional networks, and ultimately leads to improvements in upper limb function^12,13. Furthermore, robot-assisted training has also been shown to assist patients in consistently executing accurate limb movements and engaging in task-specific training¹⁴. This process provides the brain with substantial feedback stimulation, ultimately boosting neuroplasticity and aiding in the restoration of upper limb function in individuals with hemiplegia^14,15. There is currently limited research on strategies utilizing robot-assisted dual upper limb training for stroke patients. This study employed a dual upper limb task-oriented robotic system to combine robot-assisted training with bilateral upper limb training. The robotic device was utilized to aid stroke patients in conducting dual upper limb task-oriented training with high repetitions in a proper movement pattern. The objective of the research was to evaluate the effects of this method on the corticospinal pathway, upper-limb function, and activities of daily living in stroke survivors, with the aim of discovering innovative strategies for upper limb functional rehabilitation.

Protocol

This study (Approval No. JXEY-2020SW038) was approved by the Medical Ethics Committee of the Second Hospital of Jiaxing, with all participants providing informed consent. It aimed to assess the feasibility and effectiveness of a protocol through a randomized, single-blind, controlled trial. Between January and December 2021, 60 stroke patients admitted to the Second Hospital of Jiaxing were enrolled.

NOTE: Inclusion criteria comprised: 1) confirmed diagnosis of cerebral infarction or hemorrhage via computed tomography (CT) or magnetic resonance imaging (MRI), 2) first-onset and unilateral lesion with a disease duration of 2 weeks to 3 months and a stable condition, 3) age 25-75 years, 4) absence of hemianopsia or unilateral spatial neglect, as well as no visual or auditory deficits, 5) conscious, compliant, and able to participate in rehabilitation treatment, 6) clear unilateral upper limb dysfunction with a modified Ashworth scale (MAS) grade ≤ 2¹⁶. Exclusion criteria included: 1) previous craniocerebral injury or other intracranial diseases, 2) severe myocardial infarction, angina pectoris, liver, kidney, lung, or other important organ diseases, malignant tumors, etc., 3) previous history of psychiatric disorders and epilepsy, 4) severe pain, numbness, or other sensory deficits on the hemiplegic side of the limbs, 5) significant limitation of movement in the bilateral upper limbs.

1. Study design

1. Randomly divide patients (n = 60) who met the specified criteria into two groups: an experimental group (n = 30) and a control group (n = 30).
2. Have a skilled occupational therapist complete the following functional assessments, who was unaware of the group assignments before and after a 6-week treatment period.
   1. Motor evoked potential (MEP):
      1. Elicit MEPs in patients using a magnetic stimulation therapy system following the guidelines established by Groppa et al.¹⁷.
      2. During the test, position the patient in front of the device in a stable and comfortable manner and place the recording electrode pads on the abductor pollicis brevis and wrist joint osseous process.
      3. Then, center the magnetic stimulation coil above the motor cortex on the injured side of the brain, with the coil handle positioned at a 45° angle to the sagittal plane.
      4. Carry out stimulation of the motor cortex area 10 times at 100% intensity and record the presence or absence of motor-evoked potentials, along with their latency and amplitude.
         NOTE: Due to the inability to detect motor evoked potentials in all patients, a thorough comparison and analysis of the latency and amplitude of the evoked potentials between the two patient groups was not feasible. Therefore, the study aimed to determine the presence or absence of MEPs and compare the percentage of detectable MEPs between two groups of patients. A higher percentage of detectable MEPs indicates a greater potential for enhancing corticospinal pathways in stroke patients.
   2. Perform functional test for the hemiplegic upper extremity-Hong Kong (FTHUE-HK).
      1. Utilize the scale to assess the patient's upper limb functionality, which includes 12 tasks, such as placing the hand on the knee and wringing a rag.
         NOTE: Each task must be completed within 3 min and can only be attempted up to 3 times. The scale comprises 7 levels, with higher levels indicating better upper limb functionality¹⁸.
   3. Use the Fugl-Meyer Assessment Upper Extremity Scale (FMA-UE).
      1. Utilize this scale to evaluate the motor function of the shoulder, elbow, forearm, wrist, and hand.
         NOTE: A score of 0 indicates inability to perform the specified movement, a score of 1 indicates partial completion, and a score of 2 indicates full completion. The scale has a maximum score of 66 points, with higher scores indicating better upper limb motor function¹⁹.
   4. Calculate the Modified Barthel index (MBI).
      1. Utilize this scale to assess the patient's performance in activities of daily living.
         NOTE: The scale consists of 10 items, including eating, dressing, bathing, etc., with a maximum score of 100 points. A higher score indicates greater independence in the patient's daily life²⁰.
3. Ensure that all patients are prescribed conventional medications, including anti-hypertensives, anti-diabetics, lipid regulators, etc., tailored to their individual conditions.
   NOTE: The medication selection for stroke patients is based on their unique circumstances and may differ from one patient to another.
4. Confirm that all patients received routine physical therapy, occupational therapy for the forearm and hand, and activities of daily living training for 6 weeks.
5. Ensure that patients in the control group received routine occupational therapy targeting upper limb function for 1 h per day for 6 weeks.
   NOTE: The routine occupational therapy targeting upper limb function includes motor control training for the shoulder and elbow joints, roller training, hoop training, and reaching for objects training.
6. Confirm that patients in the experimental group received routine occupational therapy targeting upper limb function for 30 min per day, in addition to dual upper limb task-oriented robotic system training for 30 min per day for 6 weeks.

2. Dual upper limb task-oriented robotic system training session

NOTE: Only the stroke patients in the experimental group received these training sessions.

1. Start the robotic system equipment, turn on the system's computer screen, open the ULCOT Rehab application, and enter the main interface of the system.
2. During the initial training session, click Register to establish a personal file for each patient, mainly including name, gender, age, case number, diagnosis, affected side, and other relevant medical contents.
3. Click Login on the main interface of the system, select the patient who needs training from the list, and enter the interface of the training system for that patient.
4. Assist the patient in positioning themselves in front of the robotic device, ensuring a safe and comfortable distance.
5. Click Adjustment on the interface of the patient training system to enter the interface of equipment parameter adjustment and set the appropriate parameters for the patient.
   NOTE: It is not necessary to set parameters for each training session. Upon logging into the patient's training system interface, the system automatically adjusts to the parameters established during the patient's previous training session. The therapist can then modify the corresponding parameters in accordance with the therapeutic goals. If no changes to the parameters are needed, the user can click Training in the training system interface to access the training program setting interface.
   1. Click on + or - to increase or decrease the height of the platform in the Platform Height Adjustment module. Adjust the height of the equipment platform based on the patient's height.
   2. Click + or - to increase or decrease the tilt angle of the system's robot arm in the Arm Tilt Angle Adjustment module. Adjust the tilt angle of the robotic arm according to the patient's shoulder flexion and extension training goals (the higher the target, the greater the angle).
   3. Click + or - to increase or decrease the angle between the two robot arms in the Arm Angle Adjustment module. Adjust the angle between the robotic arms according to the patient's upper limb adduction and abduction training goals (the higher the goal, the greater the angle).
6. Click Training in the patient training system interface to enter the training program setting interface.
   1. Select an appropriate training program based on the patient's upper limb functional status. When the upper limb on the hemiplegic side is unable to actively manipulate the mechanical handle through the full range of motion, opt for the assisted training program.
   2. Conversely, if the upper limb on the hemiplegic side is capable of actively manipulating the mechanical handle to complete the full range of motion, choose the resistance training program.
7. Explain and demonstrate the training methods of the selected items and inform relevant precautions to ensure that patients know how to perform the training session safely and accurately.
8. Help the patient to fix their hands on the handles at the end of the two robotic arms (Figure 1).
9. Conduct dual upper limb task-oriented robot system training.
   1. For patients who are unable to actively manipulate the mechanical handle to achieve a full range of motion on the hemiplegic side of the upper limb, click Assistance in the training program setting interface to enter the assisted training mode interface.
      NOTE: The therapist may select Air Flying game or Ping-Pong game for the patient in assisted training mode. It should be noted that patients may only select one game per training session.
      1. Set the time to 30 min in the Training Time module, and select the level set for the patient in the Assisted Level module.
         NOTE: This mode offers 6 levels of assistance, with level 6 involving the affected upper limb being driven by both the robot and the healthy upper limb during bilateral upper limb training. On the other hand, level 1 entails the affected upper limb participating in bilateral upper limb training directly without external force. The training session commences at level 6, and the patient can progress to the next level after achieving a full score at each level. Once the patient attains a full training score at assistance level 1, they are deemed ready for resistance mode training.
      2. Click Air Flying or Ping-Pong, then click Start to enter the game interface.
      3. Air Flying game: Instruct the patient to control a virtual airplane displayed on the computer screen by maneuvering the affected upper limb through the healthy side with the assistance of a robotic device, enabling the patient to optimize their efforts in guiding the virtual airplane along the designated flight trajectory while simultaneously capturing virtual gold coins (Figure 2).
      4. Ping-Pong game: With the assistance of the robot, instruct the patient to use the unaffected side to drive the affected side upper limb to control the virtual table tennis racket and move the racket to catch the flying ping-pong (Figure 3).
   2. For patients who are able to actively manipulate the mechanical handle to achieve a full range of motion on the hemiplegic side of the upper limb, click Resistance in the training program settings interface to access the resistance training mode interface.
      NOTE: In the resistance training mode, participants can choose from five available games: Air Flying, Ping-Pong, Bridge & Road, Weight-Lifting, and Pop Matching. Only one game may be selected for each training session.
      1. Set the time to 30 min in the Training Time module, and select the resistance levels of the healthy side and the affected side, respectively, in the Healthy Level and Affected Level modules.
         NOTE: In the resistance training mode, resistance levels can be individually set for the patient's healthy and affected sides based on upper limb muscle strength. Levels range from 1 (lowest resistance) to 10 (highest resistance). The initial treatment involved the selection of Level 1 resistance, with patients permitted to progress to the subsequent level upon achieving a perfect score on each level of training.
      2. In the Healthy Side Resistance Direction and Affected Side Resistance Direction modules, select the resistance direction indicated by the system for the patient's healthy side and the affected side of the upper limb during resistance training, respectively.
         NOTE: The direction of resistance is selected for the patient according to the purpose of the exercise, including push and pull.
      3. Select the amount of time the target needs to be held in the Holding Time module.
         NOTE: The time is determined based on the patient's upper limb function, ranging from 1 to 10 s. The longer the time, the more challenging it becomes. If the set holding time is 10 s and the training score is perfect, the resistance level will be increased for the next session. The Air Flying and Ping-Pong games do not include this step.
      4. Click to select one of the following games: Air Flying, Ping-Pong, Bridge & Road, Weight-Lifting, and Pop Matching. Click Start to enter the game interface.
      5. Air Flying game: Instruct the patient to control the virtual airplane by resisting the resistance given by the robotic arm on both the healthy and affected upper limbs, enabling the patient to optimize their efforts in guiding the virtual airplane along the designated flight trajectory while simultaneously capturing virtual gold coins.
      6. Ping-Pong game: Instruct the patient to control the virtual table tennis racket by resisting the resistance given by the robotic arm on both the healthy and affected upper limbs and move the racket to catch the flying ping-pong.
      7. Bridge & Road game: Have the patient control both ends of a wooden bridge on the screen by resisting the resistance given by the robotic arm on both the healthy and affected upper limbs, move two ladder platforms of different heights, and hold them for a certain time to allow the virtual character to pass (Figure 4).
      8. Weight-Lifting game: Have the patient control the ends of a weight-lifting barbell displayed on a screen by resisting the resistance given by the robotic arm on both the healthy
         and affected upper limbs, adjusting its position to reach a target location by varying the distance and maintaining that position for a specified duration (Figure 5).
      9. Pop Matching game: Have the patient control two virtual
         fingers located at the left and right ends of the screen by resisting the resistance given by the robotic arm on both the healthy and affected upper limbs, select identical items from the left and right columns of pictures through the virtual fingers and maintain this position for a
         designated duration (Figure 6).
         NOTE: The system verifies whether the selected pictures on both sides are the same; if they are, the selected pictures are eliminated. If they do not match, the patient is prompted to re-select.

3. Follow-up procedure

1. Utilize statistical software to analyze the assessment data collected, determining appropriate analysis methods based on the data type.
2. Elucidate the significance of the data results and evaluate the impact of dual upper limb task-oriented robot system training on upper limb function in stroke patients.

Results

A total of 60 stroke patients were divided into a control group (n = 30) and an experimental group (n = 30) for this study. Upon comparing age, gender, stroke type, disease duration, side of hemiplegia, and other general information between the two groups, no statistically significant differences were found (P > 0.05), indicating their comparability (Table 1). Patients in the experimental group, who underwent training with a dual upper limb task-oriented robotic system, showed greater improvements in MEPs, FMA-UE, FTHUE-HK, and MBI compared to those receiving conventional treatment.

After 6 weeks of training, the detection ratio of motor evoked potentials (MEPs) in the experimental group surpassed that of the control group (P < 0.05) (Table 2). Following the training period, both groups of patients exhibited improvements in FTHUE-HK compared to pre-treatment levels (P < 0.05), with the experimental group showing more pronounced improvement than the control group (P < 0.05) (Table 3). Moreover, improvements in FMA-UE and MBI scores were observed in both groups of patients compared to pre-treatment levels (P < 0.05), with the experimental group experiencing more significant enhancements than the control group (P < 0.05) (Table 4). These findings highlight the effectiveness of the dual upper limb task-oriented robotic system in promoting the recovery of upper limb function in stroke patients.

Statistical analysis was conducted using appropriate software, with a significance level set at P < 0.05 for a two-tailed test. The measurement data were verified to adhere to a normal distribution and display homogeneous variances. Paired t-tests were utilized for comparisons within groups before and after treatment for normally distributed continuous variables, while two independent samples t-tests were employed for comparisons between groups. Count data were assessed using the χ2 test, ranking variables within groups were evaluated using the Wilcoxon signed-rank test, and between-group analysis was performed using the Mann-Whitney test.

Figure 1: Dual upper limb task-oriented robotic system. This system assists stroke patients with bilateral upper limb training to promote the recovery of upper limb function. Please click here to view a larger version of this figure.

Figure 2: Air Flying game. With the robot's assistance, the patient is guided to control the virtual airplane on the computer screen to make the virtual airplane fly along the set flight path. At the same time, the virtual airplane captures the virtual gold coins. Please click here to view a larger version of this figure.

Figure 3: Ping-Pong game. With the robot's assistance, the patient is instructed to control the virtual table tennis racket and move the racket to catch the flying ping pong. Please click here to view a larger version of this figure.

Figure 4: Bridge & Road game. The patient is guided to control the two ends of the wooden bridge on the screen and move it at different distances. The two ladders with different heights should be connected and maintained for a certain period of time so that the virtual villain can pass smoothly. Please click here to view a larger version of this figure.

Figure 5: Weight-Lifting game. The patient should control the two ends of the weight-lifting barbell on the screen, move it to different distances, press the barbell to the target position, and hold it for the specified time. Please click here to view a larger version of this figure.

Figure 6: Pop Matching game. The patient should control the two virtual fingers on the left and right ends of the screen through the healthy side and the affected side. The upper limbs need to use virtual fingers to select the same items in the left and right columns of pictures and maintain this position for the specified time. Please click here to view a larger version of this figure.

Group	n	Sex (n)		Age (x ± s, y )	Course of the disease (x ± s, d)	Type of stroke (n)		Hemiplegic side (n)
		Male	Female			Ischemic	Hemorrhagic	Left	Right
Control group (n=30)	30	16	14	56.70±7.60	38.77±15.71	14	16	14	16
Experimental group (n=30)	30	17	13	57.17±6.93	39.47±16.23	17	13	17	13
P		>0.05		>0.05	>0.05	>0.05		>0.05

Table 1. Baseline characteristics between the two groups. It comprehensively compares the baseline characteristics of the control and experimental groups. This includes demographic and clinical data, ensuring comparability between groups.

Group	n	Pre-treatment		Post-treatment
		response	no-response	response	no-response
Control group (n = 30)	30	8	22	10	20
Experimental group (n = 30)	30	7	23	18	12
P		>0.05		<0.05

Table 2. Comparison of MEPs response between the two groups. It demonstrates the effect of a dual upper limb task-oriented robotic system on corticospinal pathways in stroke patients.

Group		FTHUE-HK (M(P25, P75))
Control group (n = 30)	Pre-treatment	3(2,3)
	Post-treatment	3(3,4)*
Experimental group (n = 30)	Pre-treatment	3(2,3)
	Post-treatment	4(3,5)*#
*P < 0.05, compared to pre-treatment; #P < 0.05, compared to the control group

Table 3. Comparison of FTHUE-HK between the two groups. It describes the impact of the dual upper limb task-oriented robotic system on upper limb function in stroke patients.

Group		FMA-UE  (x ± s)	MBI (x ± s)
Control group (n=30)	pre-treatment	25.33±11.72	44.27±13.21
	Post-treatment	34.63±13.06*	51.03±12.55*
Experimental group (n=30)	pre-treatment	25.93±11.87	44.93±14.10
	Post-treatment	42.37±15.20*#	59.73±14.63*#
*P < 0.05, compared to pre-treatment; #P < 0.05, compared to the control group

Table 4. Comparison of FMA-UE and MBI between the two groups. It illustrates the impact of the dual upper limb task-oriented robotic system on upper limb function and activities of daily living in stroke patients.

Discussion

Bilateral training has been shown to normalize intercortical inhibition in stroke patients, facilitate brain functional network reorganization, and ultimately enhance upper limb function²¹. This study presents a program for upper limb functional training in stroke patients utilizing a dual upper limb task-oriented robotic system. The program integrates bilateral upper limb movement, task-oriented activities, and robot-assisted training to enhance the rehabilitation of upper limb function in stroke patients.

Several key steps warrant attention in implementing the dual upper limb task-oriented robotic system training. First, the therapist should promptly adjust the tilt angle of the robotic arm and the angle between the two arms based on the functional status of the patient's upper limb and the therapeutic goals. Second, the level of assistance or resistance provided by the system must be accurately selected in accordance with the patient's upper limb muscle strength. When the patient's training score reaches the maximum, it should be adjusted to the next level without delay. Third, in the resistance training mode, the therapist should establish the resistance levels for both the healthy and affected sides, as well as the direction of resistance (including push and pull), depending on the muscle strength of the patient's upper limbs on each side.

The dual upper limb task-oriented robotic system training involves movements of the upper limbs across various planes and directions. However, random switching between these planes and directions during training is not feasible, as each switch necessitates halting the current training session to recondition the system. Some researchers have employed two identical robots to assist patients in bilateral upper extremity training across three dimensions⁴. While this approach enables patients to engage in multiple directions of motion during training, it poses challenges in effectively transferring forces between the healthy and affected limbs. As the dual upper limb task-oriented robotic system is refined in subsequent stages, it is essential to enhance the degrees of freedom of movement of the robotic arm to accommodate the multi-directional movement training of the upper limb. Additionally, it is crucial to address the issue of compensatory trunk movements that some patients exhibit during training with the dual upper limb task-oriented robotic system. Such compensatory movements can diminish the range of motion of the upper extremities and may lead to the development of faulty movement patterns. To mitigate the impact of this issue, therapists should promptly remind patients to maintain proper sitting posture and adhere to correct movement patterns during training.

Most traditional bilateral upper limb training methods involve the healthy hand holding the affected hand or connecting the two hands with a device (e.g., a wooden stick). In contrast, the dual upper limb task-oriented robotic system training utilized in this study offers significant advantages. Research indicates that recovery of limb function in stroke patients is enhanced by precise and highly repetitive rehabilitation training²². However, following a stroke, patients often exhibit diminished muscle strength in the affected limb and reduced motor function in the healthy limb^23,24. Consequently, during traditional bilateral upper limb training, it becomes challenging for patients to maintain normal movement patterns continuously and repetitively over extended periods. Furthermore, to perform a specific movement, the healthy upper limb may exert considerable force while the affected upper limb applies minimal force, thereby compromising the full engagement of the affected limb. The dual upper limb task-oriented robotic system training can modulate the force transmitted from the healthy upper limb to the affected limb based on the muscle strength of the patient's affected upper limb, thereby facilitating gradual and structured participation of the affected limb. This training also employs robotic assistance to enable patients to execute highly repetitive and precise movements, which has been shown to provide constant feedback to the brain, which promotes functional reorganization and ultimately enhances limb function^14,22. Additionally, the virtual games incorporated into the dual upper limb task-oriented robotic system training are task-oriented, and studies have demonstrated that such training can improve upper limb function and the ability to perform activities of daily living in stroke patients^25,26.

In this study, the MEPs in patients were based solely on the presence or absence of detectable MEP. This decision was made because a comprehensive comparative analysis of MEP latency and amplitude was not possible, as MEP could not be detected in some patients. The study included patients with varying disease durations ranging from 2 weeks to 3 months, potentially impacting the results due to differences in spontaneous recovery. Patient selection criteria focused solely on stroke type and hemiplegic lateral condition without considering the specific brain lesion areas, thus affecting the comparative analysis of efficacy. Additionally, there are other limitations identified in this study. Firstly, patients with high muscle tone (MAS > 2) were excluded from the experiment, as their condition could potentially impact the training outcomes. Secondly, the evaluation of the experiment's efficacy was only conducted up to 6 weeks post-intervention, lacking long-term follow-up data. Thirdly, all participants were within 3 months of disease onset, leaving uncertainty regarding the effectiveness of this training approach for patients beyond the 3-month mark. Moreover, the study's sample size was small, highlighting the necessity for future research with a larger and more diverse sample. In response to the issues mentioned above, we will implement further enhancements and optimizations during the subsequent stages of the study.

In conclusion, the dual upper limb task-oriented robotic system has shown promise in enhancing upper limb function and activities of daily living for stroke patients. This approach warrants broader adoption in clinical settings for upper limb functional rehabilitation post-stroke.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Dual upper limb task-oriented robotic system	Auckland Tongji Rehabilitation Medical Equipment Research Center, Tongji Zhejiang College	N/A	The dual upper limb task-oriented robotic system can aid stroke patients in bilateral upper limb virtual game training by regulating force transmission between the healthy and affected upper limbs.
Magnetic stimulation therapy system	Sichuan Junjian Wanfeng Medical Equipment Co.,Ltd.	http://www.jjwf-med.com	This system can be used to measure the Motor evoked potential (MEP)
SPSS 25.0	IBM	Version 25.0	https://www.ibm.com/support/pages/downloading-ibm-spss-statistics-25