Article Text
PDF
Abstract
Introduction Stroke survivors often face motor dysfunction, increasing fall risk. Lower extremity muscle weakness is a key factor affecting walking ability. Tai chi (TC) has been shown to improve muscle strength and mobility in patients with stroke more effectively than traditional walking training. However, existing TC programmes for stroke rehabilitation are often too simplified and fail to fully use TC’s benefits. Additionally, subjective assessment scales are time-consuming and prone to bias. This study proposes integrating TC’s early movement features with neurodevelopmental therapy, using surface electromyography and inertial measurement unit (IMU) sensors to thoroughly analyse diverse TC movements. Tailored exercises, based on stroke-induced impairments, will be objectively assessed through biomechanical analysis.
Methods and analysis The study unfolds in two phases. The initial phase employs the IMU sensor and electromyography to objectively analyse TC’s biomechanics, informing personalised rehabilitation plans aligned with distinct movement impairments. The second phase adopts a randomised, single-blind, parallel controlled trial design involving 60 patients with stroke randomly assigned to either the intervention or control group. The intervention group undergoes biomechanics-based TC training alongside routine rehabilitation for 12 weeks, practicing the 24-form TC three times weekly. The control group engages in routine rehabilitation thrice weekly for the same duration. Primary and secondary outcomes, including kinematic/dynamic data, surface electromyography, motion analysis, comprehensive the international classification of functioning, disability and health Core Set for Stroke, Modified Barthel Index and Fugl-Meyer Assessment, will be evaluated at baseline and post-intervention.
Ethics and dissemination The study has received approval from the Ethics Committee of Zhongda Hospital Southeast University (2023ZDSYLL378-P01). All prospective participants will receive comprehensive information regarding the study protocol, and their informed consent will be obtained before their participation. Additionally, the trial will be registered with the Chinese Clinical Trial Registry to ensure transparency and compliance with research regulations. Results from this study will be disseminated through peer-reviewed journals, conference presentations and public databases to ensure wide accessibility and to contribute to the advancement of medical knowledge.
Protocol version 2.0 (14 June 2024).
Trial registration number www.chictr.org.cn, identifier ChiCTR2400080158.
- randomized controlled trial
- clinical trial
- stroke
- rehabilitation medicine
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
Uses surface electromyography and inertial measurement unit sensors for objective assessment.
Provides detailed biomechanical analysis of tai chi movements in patients with stroke.
Customises rehabilitation plans based on initial biomechanical data.
Difficulty in blinding participants and therapists due to the nature of physical exercise intervention.
The 12-week intervention period may not be sufficient to observe long-term effects.
Introduction
Stroke exhibits five prominent characteristics, including a high incidence rate, frequent recurrence, significant disability and elevated mortality. Following a stroke, individuals commonly experience various functional impairments, with motor function being the most profoundly affected.1 Muscle weakness and decreased balance capacity are critical factors contributing to the increased risk of falls in stroke survivors.2 Biomechanical analysis of walking in stroke survivors highlights the impact of insufficient muscle strength around the joints, leading to reduced ankle dorsiflexion, diminished knee or hip joint flexion and restricted range of joint motion (ROM) caused by muscle spasticity, ultimately culminating in a stiff-knee gait.3 Additionally, the combination of inadequate hip joint control and weakened lower limb muscles often results in stroke survivors adopting a trunk-bending walking pattern, leading to slower walking speed and inadequate propulsion during gait.4 5
Tai chi (TC) has been recognised as a safe and effective therapeutic exercise by the American Heart Association and the American Stroke Association.6 Engaging in long-term TC practice can significantly improve hip muscle strength, lateral stability and balance in patients with stroke.7 This slow-paced, bilateral exercise also targets anti-gravity muscles and lower-body muscle groups, such as quadriceps, knee extensors and flexors, which ultimately leads to enhanced walking ability.8 9 Despite the numerous benefits, conventional TC may pose challenges for patients with stroke, particularly those with impaired motor function, making participation more difficult.10 11 Therefore, when incorporating TC into patient who had a stroke rehabilitation, it is essential to make appropriate selections tailored to individual physical conditions and requirements, ensuring the prevention of overexertion and potential injuries.
Numerous TC improvement methods have been explored in research to address exercise rehabilitation for patient who had a stroke. Among these approaches, a simplified seated TC programme has shown promise in preventing falls during the training of patients with subacute stroke and effectively enhancing rehabilitation outcomes.12 In another study, TC stepping techniques were combined with body weight support to reduce lower limb loading, while upper limb movements were eliminated to assess their impact on balance ability of patients with stroke.13 However, it is worth noting that these simplified methods may be considered too basic and might not fully capitalise on TC’s potential benefits for overall coordination of the upper and lower limbs. Moreover, the selection of individual movements independently in these approaches lacks continuity.14 Currently, most research predominantly employs functional tests to evaluate the impact of TC on the motor function of patients with stroke. These assessments encompass tests such as the timed up-and-go test, the 2-minute walk test and quality of life evaluations.8 15 Nevertheless, this evaluation process is time-consuming and subject to some degree of subjectivity.16 Consequently, there is a pressing need for objective evaluations of motor function in patients with stroke to facilitate the development of tailored rehabilitation plans and to assess the effectiveness of TC in addressing motor impairments caused by stroke.17 18
Surface electromyography (sEMG) has emerged as a valuable tool in clinical practice and research for assessing the functionality of the muscle nervous system. It provides an objective and sensitive means of evaluating the status of muscles with impaired neural innervation following a stroke.19 20 To enhance accuracy and overcome subjective visual scoring errors, motion capture systems have been employed, enabling automated evaluation.21
In this study, a modified TC training programme will be developed by considering the characteristics of post-stroke motor dysfunction and through consultations with TC experts and rehabilitation medicine specialists. This training programme will incorporate the characteristics of the patients’ movement impairments and select appropriate movements from the 24-form TC for them to practice.22 This approach aimed to create a targeted and individualised TC training programme for patients with stroke. To objectively assess the effects of this modified TC programme in improving motor impairments caused by stroke, sEMG equipment will be used.
Objectives
The study not only focused on evaluating the therapeutic outcomes but also delved into analysing the biomechanical mechanisms underlying the observed effects. By doing so, it aimed to provide scientific evidence and valuable references for optimising the application of TC in stroke rehabilitation.
Methods/design
Trial design
The part I of this study aims to observe and record the fundamental patterns and biomechanical characteristics of TC movements using EMG analysis. These findings will then be used to personalise TC training programmes for patients with stroke, taking into account their specific functional impairments at different stages of rehabilitation. This study follows an observational approach. The second phase of this study aims to investigate the impact of a modified TC programme on motor function improvement in patients with stroke, using a single-blind, randomised, parallel-controlled trial design. Eligible participants will be randomly assigned in a 1:1 ratio to either the TC training group or the conventional rehabilitation group. The study duration spans 12 weeks, with a 12-week intervention phase. Assessments of primary and secondary outcomes will be conducted at baseline and at the conclusion of the 12-week period. An overview of the study design is presented in figure 1.
An overview of the study design.
Sample size
Drawing on similar studies as references,23 24 part I of this research will involve recruiting 15 participants experienced in practicing TC. The goal is to gather and analyse the biomechanical characteristics specific to TC. The sample size of the second phase determination was conducted using the G*Power software (developed in Germany), leveraging insights from a precedent study. This antecedent investigation documented noteworthy advancements in upper limb functionality among participants engaged in a 12-week Tailored Sitting Tai Chi Programme. Specifically, the group practicing sitting TC exhibited a substantial enhancement in primary outcome measures encompassing performance time and functional ability.25 At a significance level of 0.05, employing a statistical power of 0.80 within a two-tailed test and accounting for a dropout rate of 10%, the calculated effect size stands at 0.78. Consequently, the projected sample size for each of the two groups is estimated to be 60 participants.
Participants
Participants for part I will be selected from the pool of students at both Nanjing Sport Institute and Nanjing University of Chinese Medicine. They will practice TC at designated locations, while the researchers simultaneously measure and record kinematic and kinetic data. At the same time, the study participants of part II will be recruited from the Rehabilitation Medicine Department of Zhongda Hospital Southeast University, China. They will undergo initial evaluation and assessment at this hospital. During the intervention phase, the participants will receive TC training or conventional rehabilitation at the same department. Following the intervention, the patients will continue their exercises at home under the supervision and guidance of the research team. This arrangement ensures that the participants receive continuous support and monitoring during their rehabilitation process.
Recruitment criteria of part I
Inclusion criteria
Has more than 1 year of experience in practicing TC.
Is in good health, with no history of lower limb or joint injuries in the past 6 months.
Agrees to participate in this experiment and signs the informed consent form.
Aged 18–70.
Exclusion criteria
Clinical imaging examinations suggest knee osteoarthritis, knee fractures, dislocation, tuberculosis, tumours, infections, etc.
Individuals with other factors that might affect lower limb movement function, such as a history of joint fractures or knee surgery.
Those with significantly abnormal physical structures, such as flat feet, high arches or X-shaped legs.
Individuals who are unable to cooperate with the testing due to other reasons.
Recruitment criteria of part II
Inclusion criteria
Patients diagnosed with cerebral infarction based on CT and/or MRI.
Patients presenting with primary or unilateral onset, or with previous episodes but no neurological dysfunction.
Patients with stable vital signs.
Participants aged between 18 and 70 years old.
The duration of the disease ranged from 2 weeks to 3 months.
Hemiplegia with affected lower limb muscle strength of grade 3 or above.
Informed consent was obtained from the selected patients or their family members.
Modified Rankin Scale score ≤3.
Exclusion criteria
History of epilepsy, including idiopathic epilepsy in first-degree relatives and use of epileptogenic drugs.
Severe cognitive and communication disorders that could hinder participants from cooperating with evaluation and treatment.
Posterior circulation cerebral infarction, presence of a pacemaker or intracranial metal implants.
Severe cervical spondylosis, including cases with severe cervical stenosis and instability.
Complete occlusion of the internal carotid artery.
Cranial defects.
Pregnancy in female participants.
Randomisation and allocation concealment
Part II of this study involves a clinical randomised controlled trial, and as such, randomisation methods and blinding methods will be employed before recruiting participants for the research, we will establish a randomisation table, with each entry corresponding to an individual participant. Eligible participants will be randomly assigned to either the experimental group (n=30) or the control group (n=30) in a 1:1 ratio. To ensure the credibility of the randomisation process, a third party independent of the research team will carry out the allocation. During the allocation process, computer-generated random numbers or random sampling methods will be used to achieve unbiased randomisation. The allocation results for both the experimental and control groups will be recorded and securely stored. This will enable us to verify the effectiveness of the randomisation process during subsequent data analysis, ensuring the validity of the study’s findings.
All enrolled individuals will be assigned unique identification numbers, replacing their personal identity information. The research personnel involved in conducting the experiment will be kept blinded to the participants’ random group assignment to avoid any potential subjective influence on the experimental results. Throughout the data collection process, all participant-related data will be labelled solely with their assigned identification numbers and no reference to their true identities will be used. Any data that contains personal identity information will be stored securely and kept in a restricted environment, accessible only to authorised researchers. Strict data protection protocols will be implemented to safeguard the confidentiality of participants’ information and comply with relevant data privacy regulations and ethical guidelines. These measures aim to maintain the integrity of the research and uphold the rights and privacy of the study participants.
Blinding
Given the nature of the intervention in this study, blinding cannot be implemented for the therapists teaching TC and the participants. However, to minimise bias, the therapists will not be involved in the assessment process. Additionally, the research personnel conducting the assessment and statistical analysis will not participate in participant screening and allocation. To maintain confidentiality and prevent potential bias, all outcome data will be collected and entered by personnel who are independent of the research team and will be responsible for the randomisation process. Personal identity information will be replaced with unique identification numbers, and participants will be assigned group codes (Group 1 and Group 2) to distinguish between the experimental and control groups.
Intervention
Part I involves solely the measurement and analysis of kinematic and kinetic parameters during TC exercise on the participants, without implementing any interventions. Part II will involve a total of 60 patients who are post-stroke who meet the inclusion criteria. These participants will be divided into two groups: the control group and the intervention group. The control group will receive conventional rehabilitation training, following the guidelines specified in the ‘Technical Specifications for Common Rehabilitation Therapies’ published by the Chinese Association of Rehabilitation Medicine in 2012.26 On the other hand, the intervention group will undergo TC training, which incorporates an optimised set of TC movements, in addition to the conventional rehabilitation training. The TC sessions will be conducted three times a week, based on the conclusions of a previously reported systematic review, it was found that TC and qigong performed for 12 weeks or less were effective in improving the mobility of stroke survivors.27 Considering this, and in conjunction with our own medical conditions, we decided on a total intervention duration of 12 weeks.
Design of training plan
The 24-form TC demands a higher level of strength, balance and coordination in both upper and lower limbs, which could present challenges for patients with stroke in the recovery period.28 In addition, a practitioner needs to memorise an elaborate sequence of TC forms and learn to plan their next movement while maintaining good balance.29 Research findings indicate that seven specific movements have been identified as particularly effective in engaging the lower limbs and strengthening muscles responsible for postural stability.30 Through sEMG analysis, it will be observed that TC movements can be tailored to induce specific motions in patients with stroke and activate targeted muscles. This will help activate specific muscles and movement patterns. As their functional capacity develops, the training movements will be adapted accordingly. In the first phase of our study, we will examine the relationships among 16 muscles during various movements. For the upper limbs, we selected the deltoid, biceps brachii, triceps brachii and extensor carpi radialis longus, as their contractions involve the movements of the shoulder, elbow and wrist joints. For the lower limbs, we selected the rectus femoris, biceps femoris, gastrocnemius and tibialis anterior, as their contractions cover the movements of the knee and ankle joints. Each TC movement has distinct muscle force characteristics, so we will analyse each movement individually.
After data collection, we will begin with preprocessing, including filtering, rectification and normalisation of the EMG signals. Next, we will apply the Hilbert transform to obtain the analytical signal of the EMG and compute its absolute value to derive the signal envelope. Data processing will involve two main components: muscle synergy feature extraction and mutual information index calculation.
First, we will use non-negative matrix factorisation (NNMF) in MATLAB 2022a to extract muscle synergy components from the processed EMG signals of each channel. Given that some error exists between the basis matrix and the coefficient matrix derived from NNMF and the original EMG matrix, we will use the variability accounted for to assess the quality of the reconstructed matrix.31 After normalising the contribution index of each channel, we will set a threshold of 0.1; if the contribution index exceeds 0.1, the muscle from that channel will be considered a primary contributor to the movement.
An approach for elucidation of changes in intermuscular coordination is the measure of normalised mutual information (NMI). NMI will be used to measure the degree of association between two random variables. A higher NMI value indicates a stronger association, making NMI a non-linear extension of the traditional correlation coefficient.32 It will be used to construct functional networks of sEMG signals. In this study, a ‘connection’ refers to the mutual information index between two channels. We will first verify the normality of the NMI values and then use a t-test to determine the statistical significance threshold for the NMI values. Muscle pairs with NMI indices above the threshold will be identified as highly correlated, connected muscle pairs. In figure 2, we depict the muscular functional connectivity network of a single TC movement. In practical experimentation, we intend to incorporate various stages of limb coordination impairments and muscle strength conditions post-stroke. By identifying distinct muscle connectivity patterns and muscle contribution distributions corresponding to different TC movements, we aim to formulate personalised TC training protocols. For instance, we will select movements emphasising significant contributions from the tibialis anterior for individuals with foot drop post-stroke, and movements highlighting substantial contributions from the extensor carpi radialis longus for patients experiencing difficulties with wrist extension. Moreover, TC movements emphasising upper and lower limb coordination will be tailored to patients with diverse coordination impairments, addressing their specific needs effectively. Therefore, in this study, we will select appropriate training exercises based on the muscle strength and muscle tension of each limb of the patients. This will help avoid activating muscles with high muscle tension to prevent exacerbating abnormal movement patterns. Instead, we will focus on activating muscles with low muscle strength to promote their strength and functionality.
Muscle channel-related functional connectivity string diagram (A). It shows a muscle functional connectivity map for a specific movement based on mutual information values above the threshold. A connection line between channels indicates that the mutual information value between the two surface electromyography channels is high during this movement, signifying a strong correlation. For instance, if channel 9 is activated, channel 13 is also activated synchronously. In this figure, C9-C16 represent the eight muscles of the lower limbs. Therefore, the diagram indicates that this movement exhibits strong interconnections among the lower limb muscles, emphasising coordination in the lower limbs. Illustrative diagram of the correlation coefficients between the left triceps brachii muscle and other channels (B). C3 represents the left biceps brachii. In the figure, the darker the colour or the larger the connection area, the greater the mutual information index between the left biceps brachii and the corresponding muscle, indicating a higher correlation and stronger association.
Intervention group
The intervention group will participate in a comprehensive training programme that combines a modified 24-form TC routine (30 min) with conventional rehabilitation training (30 min). This combined intervention will be conducted three times a week and will span a total duration of 12 weeks. During the training, we will select appropriate modified TC exercises based on the patient’s functional impairment characteristics at different stages. If the required movements cannot be performed due to the patient’s motion condition, we will use rehabilitation robots to assist the patients in accomplishing the training tasks.
We are collaborating with the School of Instrument Science and Engineering to develop a TC training assistance robot, equipped with an intelligent TC rehabilitation system. This system includes advanced EMG analysis technology, virtual reality equipment, robotic assistance devices and personalised rehabilitation strategies. Based on the analysis of movement patterns in various TC postures, the system analyses surface EMG signals at different stages of the patient’s recovery. By integrating virtual reality technology, it provides personalised TC training tailored to the patient’s specific functional impairments and offers robotic assistance for movement support.
Additionally, we will implement remote guidance. Patients will initially receive rehabilitation plans at the rehabilitation centre and practice using the rehabilitation robot under the supervision of doctors. Once patients become proficient and are evaluated by doctors, they can transition to home-based rehabilitation training. Through the network rehabilitation robot cloud platform, patients can wear the rehabilitation robot at home for remote training. The robot will upload relevant motion data and physiological data measured by sensors to the database, allowing for the adjustment and optimisation of rehabilitation processes and plans. The system evaluates the patient’s recovery status based on real-time EMG signals, automatically selecting appropriate training movements and intensities for each rehabilitation stage. Hospitals can monitor patients’ training conditions and related data on the cloud platform in real time, providing remote guidance or suggestions. Additionally, family members will assist patients during training to ensure safety.
Each set of movements in the training takes approximately 6 min. During the initial 4-week intervention phase, participants in the intervention group will receive instructions in the correct modified TC movements from a person familiar with the modified TC and a professional TC instructor. Each training session will have a structured format, beginning with a 15 min warm-up period, followed by 30 min of continuous training. The training duration can be adjusted based on the individual patient’s condition, ensuring it is appropriately allocated. While the goal is for participants to complete the training continuously, those who find it challenging to do so will be allowed to perform the exercises intermittently. However, each training session must last for at least 15 min. After each training session, there will be a 5 min rest period provided to all patients. To maintain the quality of the modified TC training, a supervisor will be responsible for overseeing and managing the training sessions. Starting from the fifth week, participants will be encouraged to continue practicing the same exercises at home, three times a week, until the end of the 12-week intervention period. The intervention procedures for the control group and TC group are shown in table 1.
Intervention procedures for control group and tai chi group
Control group
The control group will receive a comprehensive conventional rehabilitation programme, which comprises various active exercises for the limbs and joints, proprioceptive neuromuscular facilitation, muscle resistance training, stretching exercises, sitting-to-standing exercises and walking. Participants in the control group will engage in rehabilitation training sessions three times a week, with each session lasting 60 min. During the initial 4 weeks of the intervention phase, the rehabilitation training will be conducted under the guidance and supervision of a professional therapist. The therapist will provide detailed instructions and correct any improper movements to ensure the effectiveness and safety of the exercises. Starting from the fifth week, participants in the control group will be encouraged to continue the same rehabilitation exercises at home, performing them three times a week until the completion of the 12-week intervention period. This approach aims to promote consistency and the continuation of rehabilitation efforts beyond supervised sessions.
Strategies to improve adherence to interventions
To encourage and support home-based practice, both groups of participants will be provided with informative flyers containing illustrations of the prescribed exercises. Additionally, the research team implemented weekly phone check-ins with participants from both the intervention and control groups. During these check-ins, participants’ physical health will be assessed, adherence to the exercise plan will be confirmed and any questions related to the programme will be addressed. To monitor and assess compliance with the home-based exercise regimen, each participant will be given an exercise diary to record their adherence to the prescribed exercises as scheduled. Additionally, participants will be requested to check in with a video recording in the Patient Management System of the Rehabilitation Medicine Department at Zhongda Hospital Southeast University, after completing their exercises. This video recording measure will be implemented to ensure the effectiveness of the training and help reduce dropout rates.
Outcome assessment
In this study, demographic information and relevant clinical data, such as age, gender, height, weight, stroke type and location, will be collected from all participants at the baseline. Primary and secondary outcomes will be assessed at two time points: baseline and at the end of the 12-week intervention. These outcomes include measurements of kinematic and kinetic parameters, activities of daily living and upper limb functional activity, as summarised in online supplemental table 1. To ensure accurate and reliable assessments, kinematic and kinetic parameters will be evaluated by experienced engineering professionals and therapists from the Rehabilitation Medicine Department of Zhongda Hospital Southeast University. Other outcomes, such as activities of daily living and upper limb functional activity, will be assessed by qualified therapists.
Supplemental material
Primary outcome
Surface electromyography
This study will use a 16-channel sEMG analysis feedback device (dno, China) to record the EMG activity levels of a total of 16 muscles in the 4 limbs. Before placing the electrodes, hair removal will be performed, followed by abrasion of the skin surface using an abrasive paste to remove dead skin cells and surface dirt. Subsequently, the skin will be cleaned with a 75% alcohol swab, and the test electrodes will be positioned on the muscle belly surface of the target muscles, aligned along the direction of the muscle fibres. In the study, we used two types of Butterworth filters to process the EMG signals. A high-pass Butterworth filter with a cut-off frequency of 30 Hz is designed to remove motion artefacts by allowing frequencies above 30 Hz to pass. Additionally, a low-pass Butterworth filter with a cut-off frequency of 600 Hz is used to remove high-frequency noise by allowing frequencies below 600 Hz to pass. Both filters are designed with a sampling frequency (Fs) of 2000 Hz.
First, maximum voluntary contractions (MVC) tests will be conducted, with three maximum muscle contractions for each test muscle. The highest filtered peak EMG amplitude (EMGmax) will be used for subsequent EMG normalisation.33 Each MVC trial lasted for 5 s, and EMG data will be recorded. After each MVC test, subjects will be instructed to rest for 5 min. All resistance will be manually applied by the researchers.34 During MVC testing, the positions of the subjects and researchers followed the method proposed by Kelln et al,35 which provides the tester with the greatest possible mechanical advantage relative to the subject. During the collection of surface EMG signals during the 10-metre walk, subjects will be instructed to complete the walking task three times, in order to average the results of the three attempts. We will collect data from a total of 16 key muscles involved in various major joint movements of both upper limbs. These muscles are the biceps brachii, extensor carpi radialis longus, deltoid and triceps brachii. Additionally, we will collect data from a total of eight joint muscles that cover various lower limb joint movements on both sides. These muscles include the rectus femoris, biceps femoris, tibialis anterior and gastrocnemius.
We will use a surface EMG analysis feedback device consisting of electrodes and an EMG main unit. The EMG main unit stores the collected data, while the electrodes capture the changes in potential difference during activity. This data is transmitted via Bluetooth to a cart, and the data is retrieved after each test. During data collection, we secure the EMG main unit with breathable athletic bandages to reduce sensor detachment caused by excessive sweating. Additionally, we will continuously monitor data quality; if poor data quality is detected due to inadequate sensor adhesion, we will discard that data set, reapply the sensors and retake the measurements. We aim to obtain three valid data sets. The device details and sensor fixation specifics are depicted in figure 3.
Surface electromyography (sEMG) device display (A) and electrode placement diagram (B, C). *Note: the person depicted is not a patient and was taken with the participant’s knowledge.
We will calculate the levels of EMG activity throughout the entire 10-metre walking process before and after the intervention. Ultimately, we will compare these levels to assess the effectiveness of TC in improving muscle strength in patients. First, the raw EMG signals undergo dual filtering with high-pass and low-pass filters to remove motion artefacts. Subsequently, the absolute values of the filtered signals are taken, and finally, normalisation is applied to eliminate amplitude differences across different individuals and conditions.
Root mean square
This index reflects the muscle strength of the tested muscle, and its peak represents the amplitude and contraction intensity. As an objective data measure, it is more sensitive and reliable in reflecting changes in muscle strength compared with the clinical muscle strength assessment that only divides into 0–5 grades,36 a higher root mean square value represented the recruitment of more motor units and better muscle strength.35
Median frequency
It will be used to observe muscle fatigue. It is defined as the frequency at which the power spectrum of the EMG signal is divided into two equal parts. In other words, median frequency (MF) is the frequency below which 50% of the total power of the signal resides. As muscle fatigue progresses during sustained contractions, there is a shift in the power spectrum of the EMG signal towards lower frequencies. This shift results in a decrease in the MF.37
Kinematic parameters
An inertial measurement unit (IMU) sensors is employed to analyse the gait characteristics of subjects while walking.
In this study, we use a sophisticated motion capture and assessment system that employs IMU sensors to evaluate and record subjects’ kinematic data. We intend to collect kinematic data during the motion process and subsequently analyse various metrics, including step length, stride, walking speed, step frequency and the trajectory of the centre of mass movement. Furthermore, our attention will be directed towards quantifying the maximal range of motion for the upper limb joints (shoulder, elbow, wrist) as well as the lower limb joints (hip, knee, ankle).
A total of 13 IMUs will be installed to collect the three-dimensional kinematic data of the participants during a 10-metre walk. These IMUs will be attached to the head, upper arms, forearms, hands, lumbar spine, thighs, shins and feet. The 13 IMU sensors can display the skeletal motion of the human body in real time. The motion capture device we use (MVN, Xsens Technologies, Enschede, Netherlands; sampling rate: 60 Hz) does not rely on cameras for capturing, meaning it does not restrict the activity range and is not affected by the environment, thus avoiding issues with signal blockage and marker confusion.
First, the wireless full-body motion capture system is calibrated to ensure that the simulated image presented by the system matches the subject’s physique and movements. According to the instructions, measurements begin simultaneously with EMG. At the end of each measurement, data is stopped, saved and named. This process is repeated three times. The sensors placement diagram is shown in figure 4.
IMU sensors placement diagram. *Note: the person depicted is not a patient and was taken with the participant’s knowledge. IMU, inertial measurement unit.
In this study, the processing algorithm for calculating gait features using IMU data primarily involves the fusion of accelerometer, gyroscope and magnetometer data. First, angles are calculated using accelerometer data by applying the arctangent function to the tri-axial accelerometer data to determine the tilt angles in various directions. Gyroscope data is used to calculate angular velocity, and angles are obtained by integrating the angular velocity.
In gait analysis, the Zero Velocity Update method is employed to identify zero velocity points by detecting the stationary state of the foot, resetting the velocity at these points to reduce cumulative errors in position and velocity. Specifically, zero velocity points are determined by detecting the magnitude of the velocity, and the velocity is reset to zero at these points, thus improving the accuracy of position estimation. Subsequently, position is calculated by integrating the velocity, and step length is determined by calculating the difference in position. To ensure data consistency and accuracy, coordinate transformation is performed to convert the local coordinate system data of the sensors to a global coordinate system. This involves computing rotation matrices to align all sensor data within a unified reference frame for integration and processing. A classical complementary filter is used to fuse accelerometer and gyroscope data, with a high-pass filter applied to the gyroscope data and a low-pass filter applied to the accelerometer data, resulting in a smoother and more reliable attitude estimation.
The calculation of joint angles first involves coordinate transformation and integration of gyroscope data. For the lower limbs, the angle changes of the thigh, calf and foot are calculated, ultimately yielding the angles of the knee and ankle joints. For the upper limbs, the angle changes of the upper arm, forearm and hand are calculated, resulting in the angles of the elbow and wrist joints.
Finally, step length is calculated through a calibration method, which involves computing the difference between the initial and final positions. The calculation of the centre of mass involves double integration of the acceleration of each part to obtain velocity and position, followed by averaging the centre of mass of each part.
Secondary outcomes
Comprehensive the international classification of functioning, disability and health (ICF) Core Set for Stroke: The Comprehensive ICF Core Set for stroke has proven to be highly valuable in stroke rehabilitation, displaying robust correlations with widely used assessment tools such as the Fugl-Meyer Assessment and the Barthel Index. This core set serves as a comprehensive and reliable rehabilitation assessment tool, enabling a thorough evaluation of patients with stroke functional status, activities and participation levels, while also considering the influence of environmental factors.38 When integrated with clinical assessment scales, the Comprehensive ICF Core Set offers a holistic approach to evaluating patients with stroke. By incorporating various domains of functioning and considering the impact of the environment, this assessment tool provides a more comprehensive understanding of the patient’s overall condition and the factors influencing their rehabilitation progress.
Modified Barthel Index (MBI): MBI is a tool to assess the patient’s ability to perform activities of daily living regarding the following categories personal hygiene, bathing, eating, toileting, dressing, bladder control, bowel control, stair climbing, chair/bed transfer and walking. Scores range from 0 to 100, and higher scores indicate a higher level of functional independence.39
Fugl-Meyer Assessment for upper extremity (FMA-UE): The FMA-UE score will be measured by an experienced physical therapist while patients are seated. The FMA-UE is used internationally and is considered the criterion standard for evaluating upper extremity function in patients with stroke. The scores range from 0 to 126 (66 points for motor function, 12 points for sensation, 24 points for passive joint motion and 24 points for joint pain). The FMA-UE is clinically feasible and shows excellent reliability and validity.40
Patient satisfaction and perception: To assess patients’ satisfaction and perception of TC training, we designed a questionnaire based on previous studies. This questionnaire consists of 10 questions covering various aspects of the programme design, such as duration, intensity, intervention components, selection of exercises, TC instruction, outcome measures and support provided for home practice. The questions are scored on a Likert scale ranging from 1 (completely dissatisfied) to 5 (very satisfied). Additionally, the questionnaire includes three open-ended questions to capture participants’ subjective opinions, such as difficulties encountered during training and any suggestions for improving the programme.41 42 The questionnaire set-up is shown in online supplemental table 2.
Supplemental material
Safety assessments
In the event of any intervention-related adverse events (AEs), a thorough and detailed documentation will be carried out. Prompt actions will be taken by the therapists and physicians to provide immediate rescue and appropriate treatment measures. If a participant experiences severe adverse reactions, the ethics committee will be involved in determining whether the participant needs to withdraw from the study for their safety and well-being. The documentation of AEs will include information such as the severity of symptoms, time of occurrence, duration and the treatment measures employed. Furthermore, a causal relationship analysis will be conducted to assess the potential link between the AEs and the intervention being studied. To ensure patient safety during TC training, we plan to implement the following measures:
Involvement of family members or caregivers: We have incorporated the involvement of participants’ family members or caregivers to oversee the home practice sessions. They will be instructed on the proper techniques and safety precautions to monitor the participants effectively.
Comprehensive training sessions: Prior to the commencement of home practice, participants and their caregivers will undergo thorough training sessions. These sessions will cover the correct execution of TC movements, potential risks and safety protocols to mitigate any accidents or injuries.
Regular check-ins: Throughout the home practice period, our research team will conduct regular check-ins with the participants and their caregivers. These check-ins will be carried out via phone calls or virtual meetings to ensure that the exercises are being performed correctly and safely.
Emergency contact information: Participants and their caregivers will be provided with emergency contact information, including the contact details of the research team and local medical services, in case immediate assistance is required.
Feedback mechanism: We will establish a feedback mechanism where participants and their caregivers can report any issues, concerns or incidents during the home practice sessions. This will enable us to promptly address any safety-related matters and make necessary adjustments to the intervention protocol.
Statistical analysis
A statistician who is independent of the trial will be responsible for the statistical analysis. The statistical analysis will be performed using SPSS V.20.0 software. Continuous data are presented as mean±SD and will be subjected to normality and homogeneity of variance tests. If the data followed a normal distribution, the independent samples t-test will be used for between-group comparisons. If the data did not follow a normal distribution, the Wilcoxon rank-sum test will be employed. Categorical data will be analysed using the χ2 test. Descriptive statistical analysis of qualitative data will be conducted by frequency and composition ratio. Comparisons will also be made between and within groups. Χ2 or Fisher’s exact test will be used for intergroup comparison, and the Wilcoxon rank-sum test will be used for rank data. A p value<0.05 indicates a statistically significant difference.
Patient and public involvement
No patient is involved. The development of the research question and outcome measures was informed by clinical observations and existing literature on TC’s benefits for motor function. Patients were not directly involved in the study’s design, recruitment or conduct. The study design relied on previous protocols and expert consultations. Results will be shared with participants through follow-up appointments. The intervention’s burden was not formally assessed by patients, but the protocol aimed to minimise the burden based on prior studies. There are no patient advisers to thank in the contributorship statement or acknowledgements.
Discussion
This study endeavours to explore the impact of a modified TC programme on the motor function of patients who are post-stroke through a single-blind, randomised, parallel-controlled trial. To achieve this, sEMG and motion analysis techniques will be employed to objectively and accurately measure the effectiveness of TC training in improving post-stroke motor impairments. By delving into the biomechanical mechanisms underlying the therapeutic effects, the research aims to gain a deeper understanding of how these improvements are achieved. By combining traditional Chinese practices, such as TC, with modern rehabilitation training, the study aims to provide scientific evidence and a standardised approach for integrating these practices into stroke rehabilitation. The ultimate objective is to contribute to the advancement of the field and enhance the efficacy of rehabilitation strategies for patients who are post-stroke.
TC training has proven to be beneficial for the functional recovery of patients with post-stroke and Parkinson’s disease. It has demonstrated the ability to improve balance function, increase lower limb muscle strength, prevent falls and enhance walking ability.43 44 The bilateral coordination and rhythmic movements in TC, involving both upper and lower limbs, make it similar to the principles of Bobath therapy and proprioceptive neuromuscular facilitation (PNF) techniques, making TC a feasible option for stroke rehabilitation.45 During the preliminary phase of this study, modern techniques such as and sEMG analysis will be used to examine the biomechanical characteristics of the 24-form TC and the activation states of various muscle groups. Based on a comprehensive understanding of the coordination and movement characteristics of TC, and considering the classic Brunnstrom stages in post-stroke motor impairments, a targeted TC training programme will be designed. This programme addresses the challenges posed by the difficulty of conventional TC exercises. Additionally, the study employed sEMG and motion analysis to quantitatively assess changes in gait performance, muscle tension and muscle activation states following TC intervention. These objective parameters allow for a precise and quantitative evaluation of the effectiveness of TC in stroke rehabilitation.
This research employs a rigorous randomised parallel-controlled design to investigate the efficacy of TC training in patients who are post-stroke and explore the biomechanical mechanisms underlying its effects on motor impairments. By integrating TC’s biomechanics with biomechanical treatment principles, the study aims to enhance our understanding of how TC movements can positively impact post-stroke rehabilitation and how this traditional practice can be applied as a therapeutic approach for patients with stroke. Through this trial, we will generate new data and insights into the effectiveness of TC training and its influence on various clinical and functional aspects of patients who are post-stroke. This research will introduce a novel treatment approach that can benefit medical institutions, physicians and patients, ultimately leading to an improved quality of life for stroke survivors and easing the burden on healthcare resources.
Indeed, this study acknowledges several limitations that may impact the interpretation of its findings. One significant limitation is the inability to implement blinding for both the participants and instructors due to the nature of the TC intervention. This may introduce potential bias in the participants’ subjective expectations and influence the study’s outcomes. Additionally, the motor limitations of patients who are post-stroke may present challenges in completing the TC training, leading to variations in exercise intensity and potential errors. The researchers are aware of this issue and will adjust the exercise intensity based on each participant’s abilities to address this limitation. However, it may still introduce some variability in the study’s results. Furthermore, the relatively short duration of the trial may limit the ability to observe long-term intervention effects, especially in terms of health outcomes. Longer follow-up periods could provide a more comprehensive understanding of the sustained effects of TC training on patients who are post-stroke rehabilitation and overall well-being. To mitigate these limitations, the research team will strive to standardise experimental procedures as much as possible to minimise errors and enhance the reliability of the study’s findings.
In conclusion, this research endeavours to assess the impact of TC intervention on post-stroke motor impairments. The potential benefits of TC training as a cost-effective and valuable rehabilitation approach are significant, offering the potential to reduce hospitalisation days and medical expenses for patients while alleviating the burden on healthcare resources. By providing evidence to support the beneficial effects of TC exercise on post-stroke motor impairments, this study may facilitate its wider clinical application and integration into stroke rehabilitation programmes. Moreover, the exploration of biomechanical mechanisms underlying the influence of TC on patients with stroke motor function is a crucial aspect of this research. By delving into the physiological processes at play, valuable insights into the therapeutic effects of TC can be gained, informing and guiding its implementation as a complementary therapeutic approach in stroke rehabilitation.
Ethics and dissemination
The study has received approval from the Ethics Committee of Zhongda Hospital Southeast University (2023ZDSYLL378-P01). All prospective participants will receive comprehensive information regarding the study protocol, and their informed consent will be obtained before their participation. Additionally, the trial will be registered with the Chinese Clinical Trial Registry to ensure transparency and compliance with research regulations.
The dissemination of the study findings will be carried out through multiple channels to ensure wide accessibility and impact. Results will be published in peer-reviewed journals and presented at relevant conferences to reach the scientific community. Additionally, data will be shared in public databases, making the findings available to researchers, healthcare professionals and the general public. This approach aims to contribute to the broader body of medical knowledge and inform future research and clinical practices.
Trial status
The project received approval in July 2023, and the initial planning phase of the research will be carried out until December 2024. During this period, the main tasks will focus on optimising and sEMG analysis techniques. The research team will establish experimental procedures for motion analysis and sEMG biometric identification. Data collection and analysis of 24 TC movement actions will also be completed, providing valuable insights into the motion analysis of TC movements in the limbs, muscle function and their characteristic changes. Building on the understanding of post-stroke motor impairments and incorporating neurodevelopmental therapies such as Brunnstrom stages and PNF, the research team will design a modified TC movement plan. Following the planning phase, the recruitment of patients who are post-stroke will commence in December 2024.
Ethics statements
Patient consent for publication
Acknowledgments
I am very grateful to Wang Tao for his help, he has given me a lot of spiritual support during my writing.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors Li Zhi is the first author and was primarily responsible for the conceptualisation and drafting of this protocol. Li Xiaoyi, Fu Xueming, Zhou Ting, Wang Pei, Fang Leiwen, Sun Zihan and Wang Hongxing contributed to the development of the study design and the writing of the manuscript. Fu Xueming provided significant guidance and critical revisions to enhance the intellectual content and overall quality of the protocol. All authors have reviewed and approved the final version of the manuscript. Wang Hongxing is the guarantor of this work and takes responsibility for the integrity and accuracy of the study protocol.
Funding This work was supported by the National Key Research and Development Program of China (2022YFC2009700), Jiangsu Province Capability Improvement Project through Science, Technology and Education, Jiangsu Provincial Medical Key Discipline Cultivation Unit (JSDW202202), Jiangsu Province Capability Improvement Project through Science, Technology and Education (ZDXYS202210), the leader project of geriatrics clinical technology application research of Jiangsu Health Commission (LR2021036), Key Project of Jiangsu Province's Key Research and Development Program (No.BE2023023-4), Key Project of Jiangsu Province's Key Research and Development Program（BE2023034).
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.