Effects of electroacupuncture synchronised with motor imagery training on upper limb motor recovery and neural mechanisms in ischaemic stroke: protocol for a single-centre, randomised controlled trial

Abstract

Introduction

Approximately half of stroke survivors experience persistent upper limb dysfunction, which impairs self-care, reduces independence and lowers quality of life. Electroacupuncture is an established intervention with evidence supporting its role in improving upper limb motor function following ischaemic stroke. Motor imagery training (MIT), which activates the sensorimotor cortex through the mental rehearsal of movement, has shown promise as an adjunctive therapy in stroke rehabilitation. The concurrent application of electroacupuncture and MIT may enhance sensorimotor recovery by promoting the integration of central and peripheral neural pathways, potentially establishing a central–peripheral–central closed-loop circuit. However, empirical evidence supporting this integrative approach remains limited.

This study aims to investigate the effects of electroacupuncture synchronised with MIT on upper limb function in patients with ischaemic stroke. In addition, longitudinal analysis of multimodal neuroimaging data will be used to explore the associated neural mechanisms.

Methods and analysis

A total of 72 patients with ischaemic stroke will be enrolled and randomly assigned (1:1) to receive either electroacupuncture synchronised with MIT or electroacupuncture. Each group will undergo 20 treatment sessions over 4 weeks (5 times per week). All participants will also receive standardised conventional rehabilitation training.

The primary outcome is the Fugl-Meyer Assessment for the upper extremity. Secondary outcomes include the Modified Barthel Index for activities of daily living, the Modified Ashworth Scale (MAS) for spasticity, Brunnstrom stages, the 17-item Hamilton Depression Rating Scale, the Chinese version of the Massachusetts Acupuncture Sensation Scale and the Kinaesthetic and Visual Imagery Questionnaire. Assessments will be conducted at baseline, mid-treatment, post-treatment and at 8-week follow-up. In addition, functional connectivity of the cerebral cortex will be assessed using functional near-infrared spectroscopy and electroencephalography, which may serve as potential biomarkers of treatment response.

Ethics and dissemination

This study has been approved by the Ethics Committee of Shanghai Second Rehabilitation Hospital (approval number: 2025-18-01) and has been registered with the International Traditional Medicine Clinical Trial Registry (ITMCTR; registration number: ITMCTR2025001311). The study will be conducted in accordance with the Declaration of Helsinki, relevant local regulations and applicable clinical guidelines. Informed consent will be obtained from all participants or their legal guardians, where applicable. The results will be disseminated through peer-reviewed publications and presentations at scientific conferences.

Trial registration number

ITMCTR2025001311.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This is a parallel-group, assessor-blinded, randomised clinical trial with a prespecified protocol for electroacupuncture synchronised with motor imagery training.

• The electroacupuncture procedure is standardised using Jin’s three-needle technique, with fixed peripheral acupoint selection.

• Allocation separation is strictly maintained, with different researchers responsible for sequence generation, concealment, recruitment, intervention delivery and outcome assessment.

• This is a single-centre study, which may reduce the external validity of the findings.

Introduction

Stroke remains a major global public health challenge owing to its high incidence, disability, mortality and recurrence rates.¹ Although advances in emergency medicine have reduced mortality during the acute phase, 70%–80% of survivors continue to experience significant motor impairments,² often resulting in loss of independent functioning and increased psychological burden.³

Neuroimaging studies have demonstrated that stroke is frequently associated with diminished functional connectivity (FC) in the lesioned hemisphere, compensatory hyperconnectivity in the contralesional hemisphere and disrupted interhemispheric communication.4,6 Among these changes, FC between bilateral primary motor cortices (M1) has shown a positive correlation with Fugl-Meyer Assessment (FMA) scores,⁷ highlighting the importance of restoring connectivity within motor-related cortical networks for functional recovery. However, current rehabilitation strategies remain suboptimal, underscoring the need for more effective interventions. Electroacupuncture, recommended by the WHO as a peripheral intervention, has been widely recognised for promoting poststroke motor recovery.8,13 Clinical guidelines support its use as an adjunct to physical rehabilitation or pharmacotherapy.¹⁴ A large multicentre trial involving 862 patients across 40 hospitals demonstrated that a 3-week course of electroacupuncture significantly improved activities of daily living (ADLs) in patients with ischaemic stroke.¹⁵ Mechanistically, electroacupuncture modulates both central and peripheral systems: it enhances functional coupling within the primary motor cortex (M1),^{16 17} while also regulating neuromuscular excitability in the periphery.^{18 19} Despite these clinical benefits, electroacupuncture is limited by substantial interindividual variability and suboptimal recovery outcomes, highlighting the need for strategies to further enhance its efficacy. Indeed, the 2023 Yearbook of Neurorestoratology highlights that diverse forms of neurostimulation and neuromodulation have shown good therapeutic outcomes in clinical trials for neurological impairments, including stroke, but also emphasises that many of these interventions still lack confirmation through high-quality randomised controlled trials (RCTs).²⁰

In recent years, a central–peripheral–central (CPC) closed-loop rehabilitation framework has emerged as a promising approach to enhance the therapeutic effects of electroacupuncture.²¹ Grounded in Hebbian principles of neural plasticity,²² this model emphasises the dynamic coupling between central motor intention and peripheral sensory feedback. Within this context, the synchronised application of motor imagery training (MIT)²³ and electroacupuncture represents a concrete implementation of the CPC mechanism. MIT is a non-invasive technique that can activate ipsilesional M1 and its associated motor-assistive networks while inhibiting excessive contralesional compensation.^{24 25} Moreover, it induces event-related desynchronisation (ERD) in sensorimotor rhythms, which is closely associated with improved motor outcomes.²⁶ Clinical studies have further demonstrated that MIT enhances limb control, balance and movement precision.^{27 28}

Integrating MIT with electroacupuncture may enhance central–peripheral synergy, optimise functional brain reorganisation and promote superior recovery of upper limb function. However, high-quality RCTs comparing this combined approach with electroacupuncture remain lacking. More importantly, the underlying neural mechanisms—such as whether it leads to enhanced ipsilesional M1 activation or restoration of interhemispheric balance—have yet to be elucidated.^{29 30}

To address these gaps, neuroimaging techniques with high temporal and spatial resolution are required to characterise the dynamic reorganisation of brain function during intervention. Electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) are two non-invasive, complementary neuroimaging modalities that have gained increasing attention in neurorehabilitation research. EEG captures real-time cortical oscillatory dynamics, including ERD, which reflect sensorimotor integration and neural plasticity.^{31 32} In contrast, fNIRS measures localised haemodynamic responses via changes in oxyhaemoglobin and deoxyhaemoglobin concentrations, offering insight into regional metabolic activity with high spatial specificity.^{33 34}

The integration of EEG and fNIRS enables the simultaneous assessment of brain function from electrophysiological and haemodynamic perspectives, overcoming the limitations of single-modality approaches. Therefore, this study will employ EEG–fNIRS multimodal imaging to characterise changes in FC within motor-related networks and their associations with behavioural outcomes following electroacupuncture synchronised with MIT. We hypothesise that the synergistic effect of peripheral regulation via electroacupuncture and central activation via MIT will produce greater improvements in upper limb function and promote sensorimotor circuit reorganisation, thereby providing empirical and theoretical support for precision rehabilitation based on the CPC closed-loop theory.

Methods and analysis

Study design, setting and timeline

This single-centre, assessor-blinded, parallel-group, randomised clinical trial will be conducted at the Department of Neurorehabilitation, Shanghai Second Rehabilitation Hospital, Shanghai, China. 72 patients will be randomised to receive either electroacupuncture synchronised with MIT (n=36) or electroacupuncture (n=36) in a 1:1 ratio. Both interventions will be delivered five times per week for 4 weeks. Participants will undergo follow-up for an additional 4 weeks postintervention (figure 1 and table 1 outline the trial procedures in detail).

Figure 1. Study procedure flow diagram. C-MASS, Chinese version of the Massachusetts Acupuncture Sensation Scale; EEG-fNIRS, electroencephalography-functional near-infrared spectroscopy; FMA-UE, Fugl-Meyer assessment of upper extremity; HAMD-17, Hamilton Depression Scale-17; KVIQ-10, Kinaesthetic and Visual Imagery Questionnaire-10; MBI, modified Barthel index.

Table 1. Schedule of enrolment, interventions and assessments.

	Study period
	Enrolment	Allocation	Postallocation			Follow-up
Time point	0	0	First intervention	Week 2	Week 4	Week 8
Enrolment
Eligibility screen	×
Informed consent	×
Allocation		×
Intervention
Electroacupuncture synchronised with MIT			×	×	×
Electroacupuncture			×	×	×
Assessment
Primary outcome
FMA-UE	×			×	×	×
Secondary outcomes
MBI	×			×	×	×
Brunnstrom	×			×	×	×
MAS	×			×	×	×
HAMD-17	×			×	×	×
KVIQ-10	×			×	×
C-MASS			×
EEG	×				×
fNIRS	×				×
Treatment adherence						×
Safety
Adverse events			×	×	×	×

C-MASS, Chinese version of the Massachusetts Acupuncture Sensation Scale; EEG-fNIRS, electroencephalography-functional near-infrared spectroscopy; FMA-UE, Fugl-Meyer Assessment of Upper Extremity; HAMD-17, 17-item Hamilton Depression Rating Scale; KVIQ-10, Kinesthetic and Visual Imagery Questionnaire-10; MAS, Modified Ashworth Scale; MBI, Modified Barthel Index; MIT, motor imagery training.

The study protocol adheres to the principles of the Declaration of Helsinki and will be reported in accordance with the Standard Protocol Items: Recommendations for Interventional Trials 2013 Statement (online supplemental material 4). Prior to enrolment, potential participants will be provided with detailed study information. Written informed consent will be obtained from all participants (online supplemental material 2).

The primary aim of this trial is to determine whether electroacupuncture synchronised with MIT (online supplemental figure S1)—forming a CPC rehabilitation loop—provides additional benefits compared with electroacupuncture. To address this question ethically and practically, all participants will receive standardised conventional rehabilitation training (CRT), ensuring that no patient is deprived of evidence-based care.

A sham electroacupuncture or MIT-only group was not included for several reasons. First, the efficacy of both interventions for poststroke upper limb rehabilitation is already well established. Second, our focus is on the synergistic effect of the combined intervention within the CPC framework rather than the independent effects of each modality. Third, assigning patients to a sham or no-treatment group could raise ethical concerns by withholding established therapy.

Both groups will receive equal treatment duration and therapist contact. Electroacupuncture will be standardised, and MIT will be delivered via uniform video guidance. Outcome assessors will remain blinded. Objective neurophysiological measures, including EEG–fNIRS indices of cortical activation and FC, will provide evidence independent of subjective response.

This trial will provide supportive, but not confirmatory, evidence for the synergistic effects of synchronised MIT. Definitive confirmation of specific central mechanisms will require future sham-controlled or factorially designed RCTs.

Patients

Participants will be recruited via posters, leaflets and announcements in hospital wards. They may contact the clinical research coordinator (CRC) by telephone, QR code or in person. The CRC will explain the study objectives, procedures, time commitment and potential risks and benefits. Those who pass the preliminary telephone screening will be invited for a two‐stage assessment, comprising an online prescreening, followed by an on‐site evaluation at the hospital. Applicants who meet all inclusion criteria and have no exclusion criteria will then be invited to discuss enrolment. Confidentiality measures will be implemented to protect personal information. Written informed consent will be obtained before randomisation.

Participation in the EEG–fNIRS substudy will be entirely voluntary. Patients will be provided with detailed information regarding the additional procedures, including the application of conductive gel, restrictions on head and speech movements and the expected time commitment. A separate consent form will be obtained from those willing to participate. Eligibility for this substudy will not affect randomisation or access to the main intervention.

Eligibility criteria

Inclusion criteria

Eligible participants must meet all the following inclusion criteria: (1) Diagnosis of first-ever ischaemic stroke confirmed by CT or MRI, consistent with International Classification of Diseases, 11th Revision (ICD-11) code 8B25³⁵; (2) Stroke onset between one and 6 months prior to enrolment³⁵; (3) Age between 40 and 80 years, regardless of sex^{36 37}; (4) Right-handedness as determined by the Edinburgh Handedness Inventory; (5) Unilateral upper limb or hand motor impairment at Brunnstrom stage IV or below²⁴; (6) Adequate motor imagery (MI) ability, defined as a Kinaesthetic and Visual Imagery Questionnaire-10 (KVIQ-10) score ≥ 25^{38 39} and (7) Ability to understand the study and provide written informed consent, either personally or via a legal representative.

Exclusion criteria

Participants will be excluded from this trial if they meet any of the following conditions: (1) Aphasia or cognitive impairment affecting comprehension, defined as a Mini–Mental State Examination ≤26⁴⁰; (2) Severe spasticity in the affected upper extremity (Modified Ashworth Scale (MAS)>2)²⁴; (3) Major primary or secondary comorbidities, such as severe hepatic, renal or haematological disorders; significant cardiovascular disease; or severe pulmonary dysfunction⁴¹; (4) Conditions interfering with EEG or fNIRS recording, including scalp wounds, metallic implants or a history of epilepsy⁴²; (5) History of neurological or psychiatric disorders unrelated to stroke⁴²; (6) Contraindications to electroacupuncture, including needle phobia or the presence of an implanted pacemaker, prosthetic heart valve or neurostimulator; (7) Musculoskeletal disorders or severe pain in the affected upper limb that could compromise the intervention⁴²; (8) Pregnancy or breastfeeding and (9) Concurrent participation in another clinical trial.

Sample size

Based on a previous study,⁴³ the primary outcome of this trial was the FMA for the Upper Extremity (FMA-UE). After 4 weeks of treatment, the mean FMA-UE score in the control group was reported as 44.33 (SD=11.10), and in the intervention group as 54.13 (SD=10.98). Using an independent samples t-test with a significance level of 0.05 and a power of 90%, we initially estimated that 56 participants would be required. A total of 72 patients were recruited, with 36 in each group. This sample size accounted for an expected dropout rate of 20%, which was consistent with rates observed in previous non-pharmacological trials and in the research team’s earlier studies, as well as the requirements of stratified block randomisation with a fixed block size of four.

For the mechanistic substudy involving EEG–fNIRS assessments, we plan to recruit approximately 30 participants, representing about 40% of the total trial population. This sample size is considered sufficient for exploratory neuroimaging analyses and is comparable with previous stroke rehabilitation studies, in which only a subset of patients underwent EEG or functional MRI (fMRI) measurements.⁴⁴

Randomisation, allocation concealment and blinding

Baseline assessments will be conducted by CRC for all eligible participants, who will be required to sign a written informed consent form. Participants will be enrolled in sequence, and their identification numbers will be written on the cover of opaque sealed envelopes. Block randomisation will be used, with a fixed block size of four. A statistician independent of the study team will generate the random sequence using SPSS V.26.0, with a seed number (202412). Random numbers will be linked to group assignments and inserted in sequence into opaque envelopes labelled with consecutive serial numbers.

The envelopes will be distributed by the statistician to two independent research assistants. These assistants will open the envelopes in order and record the allocation results. The envelopes will remain under the statistician’s custody. The CRC will arrange treatment appointments with participants and provide their serial numbers and treatment times to the intervention team. Participant names will not be disclosed to the treating staff.

Research assistants will inform the acupuncturist (a registered traditional Chinese medicine practitioner) and the MI therapist (a rehabilitation physician trained in MI) of the assigned serial number and corresponding intervention. The acupuncturist will remain blinded to group allocation. The MI therapist will be aware of treatment assignment, as they are only involved in the experimental group. Each participant will receive treatment individually and will not be aware of other participants’ schedules or interventions.

During treatment, the acupuncturist will perform electroacupuncture and then leave the room. The MI therapist will subsequently enter the room to deliver the imagery training. A clinical physician will serve as a safety supervisor throughout the procedure, responsible for monitoring, assessing and managing adverse events (AEs). They will also remove the needles at the end of each session. All procedures will follow a standardised treatment protocol.

Outcome assessors will be blinded to group allocation. During the statistical analysis phase, group allocation will remain concealed until the primary analysis is completed. The study results will be reported in accordance with the Consolidated Standards of Reporting Trials guidelines.

Intervention

All participants will receive CRT, including physiotherapy and occupational therapy, aimed at promoting motor function recovery. Details of the CRT protocol are provided in the online supplemental material 1. Each session will last 60 min, five times per week, for a total duration of 4 weeks.

The acupuncture prescription is based on clinical experience and previous research⁴⁵ and employs Jin’s three-needle prescription, which includes the following acupoints: Quchi (LI11), Waiguan (TE5) and Hegu (LI4) (figure 2). Acupoint locations follow the international standardised nomenclature defined by the WHO.⁴⁶

Figure 2. Locations of acupoints LI11, TE5 and LI4. LI11, on the lateral aspect of the elbow, at the midpoint of the line connecting LU5 with the lateral epicondyle of the humerus (LU5, on the anterior aspect of the elbow, at the cubital crease, in the depression lateral to the biceps brachii tendon); TE5, on the posterior aspect of the forearm, midpoint of the interosseous space between the radius and the ulna, 2 B-cun proximal to the dorsal wrist crease; LI4, on the dorsum of the hand, radial to the midpoint of the second metacarpal. Proportional bone cun (B-cun), this method divides the height of the human body into 75 equal units. Using joints on the surface of the body as the primary landmarks, the length and width of every body part are measured by such proportions. The specific method is to divide the height of the human body into 75 equal units, according to such an estimate, and then estimate the length and width of a certain part of the body. One unit is equal to one cun.

All acupuncture procedures will be performed by certified acupuncturists with at least 3 years of clinical experience. Electroacupuncture will be delivered for 20 min per session, 5 days per week, for 4 weeks.

Control group

Participants in the control group will receive electroacupuncture. Electroacupuncture will be delivered for 20 min per session, 5 days per week, for 4 weeks. Electroacupuncture will be presented using paradigms developed and delivered via E-Prime 3.0.

Before the intervention, participants will be instructed to relax and adopt a comfortable position. Sterile stainless-steel needles (Huatuo, Suzhou Medical Supplies Factory, Suzhou, China), 40 mm in length and 0.30 mm in diameter, will be used for acupuncture. The acupuncturist will disinfect their hands and the skin at the needle sites using 75% alcohol before insertion.

Needling techniques will follow standardised Jin’s three-needle⁴⁷ protocol, supplemented by clinical experience. Quchi (LI11) will be needled obliquely towards Shaohai (HT3) to a depth of 25–35 mm. Waiguan (TE5) will be inserted perpendicularly towards Neiguan (PC6) to a depth of 20–30 mm. Hegu (LI4) will be inserted towards Houxi (SI3) to a depth of 30–40 mm. Manual stimulation will be applied with a uniform reinforcing–reducing technique, aiming to elicit the characteristic sensation of Deqi, which is described as soreness, numbness, heaviness or distension, and is considered to be biologically relevant to acupuncture efficacy.⁴⁸

After needle placement, electrodes from a Huatuo SDZ-III electroacupuncture device (Huatuo, Suzhou, China) will be connected. The stimulation parameters will be set to a discontinuous wave at 2 Hz for 20 min.^{49 50}

Following the intervention, the device will be powered off. The acupuncturist will disinfect their hands, gently press the needle site with a dry sterile cotton ball using the left hand, and swiftly withdraw the needle with the right hand while applying pressure to prevent bleeding. The needles will be removed in the order: Quchi, Waiguan, then Hegu. The procedure is illustrated in figure 3.

Figure 3. Experimental paradigm of electroacupuncture.

Intervention group

Participants in the intervention group will receive electroacupuncture synchronised with MIT. The synchronised electroacupuncture and MIT intervention will be delivered for 20 min per session, five sessions per week, for 4 weeks. Except for the addition of synchronised MIT, all other procedures are consistent with the control group.

Electroacupuncture will follow the same protocol as in the control group. However, in this group, MIT and electroacupuncture commenced and terminated simultaneously, ensuring complete temporal alignment throughout the 20 min session. Detailed procedures for MIT preparation and delivery are outlined in online supplemental material 1. MIT and electroacupuncture will be presented using paradigms developed and delivered via E-Prime 3.0. The MIT video stimuli include the following movements: ‘left elbow flexion’, ‘left-hand grasping’, ‘right elbow flexion’ or ‘right-hand grasping’. The laterality of the video is selected according to the hemiparetic side of the patient. Each MIT session will consist of 30 trials. One trial includes a 3 s cue, 17 s of MI and 20 s of rest. The presentation order of the movements is randomised. Throughout the 30 MIT trials, electroacupuncture will be continuously delivered at 2 Hz, providing sustained peripheral stimulation during imagery tasks. Participants will be instructed to focus on the video cues while performing imagery of the depicted movements. The procedure is illustrated in figure 4.

Figure 4. Experimental paradigm of electroacupuncture synchronised with motor imagery training.

Outcome assessments

Clinical assessments will be conducted at baseline, at the first intervention, at 2 and 4 weeks during the intervention, and at a 4-week follow-up after the intervention ends. Outcome measures will be evaluated and collected by trained assessors (training protocol is detailed in online supplemental material 1), who will remain blinded to treatment allocation.

Primary outcome

The primary outcome is the FMA-UE score. The FMA-UE is a widely used, performance-based measure recommended for assessing motor impairment in stroke survivors with varying degrees of dysfunction.⁵¹ It is also commonly employed to monitor recovery in patients with hemiplegic stroke.⁵² The scale evaluates upper limb performance, including reflex activity, muscle strength and motor control.⁵¹ It comprises 33 items assessing the shoulder, elbow, forearm, wrist and fingers. The total score ranges from 0 to 66, with higher scores indicating better functional recovery.⁵³ The Chinese version of the FMA-UE has demonstrated good reliability and validity for assessing upper limb function.⁵⁴

Secondary outcomes

Modified Barthel Index

The Modified Barthel Index (MBI) is commonly used to assess disability or dependence in ADL among stroke patients.⁵⁵ It is considered one of the most suitable tools for measuring functional response in ADL.⁵⁵ The MBI consists of 10 items evaluating independence in basic daily activities, including grooming, bathing, eating, bowel and bladder control, toileting, stair climbing, dressing, walking on level ground and chair-bed transfers. The maximum score is 100, with higher scores indicating better ADL performance.⁵⁶ The Chinese version of the MBI has demonstrated validity and reliability for assessing elderly stroke patients.⁵⁷

Brunnstrom

The Brunnstrom Recovery Stages (BRS) classify motor recovery of the upper limb, including the arm and hand, into six distinct stages. These are: stage 1, flaccidity; stage 2, minimal or no voluntary movement; stage 3, movement within synergy patterns without voluntary control; stage 4, some movements deviating from synergy; stage 5, complex movements with partial voluntary control outside synergy; and stage 6, disappearance of synergy and near-normal movement. Higher stages indicate better motor recovery.⁵⁸

Modified Ashworth Scale

The MAS is used to assess increased muscle tone following central nervous system lesions, reflected by resistance to passive joint movement.⁵⁹ The scale ranges from 0, indicating normal tone, to 4, indicating rigid muscle contracture, with intermediate scores of 1, 1+, 2 and 3.⁶⁰ The MAS demonstrates excellent intrarater reliability when assessing spasticity in the upper limbs of patients with hemiplegia.^{61 62}

Hamilton Depression Scale-17

The Hamilton Depression Scale-17 (HAMD-17) consists of 17 items, each rated on a scale from 0 to 4.⁶³ The total score ranges from 0 to 53, with higher scores indicating greater severity of depression.⁶⁴ The Chinese version of the HAMD-17 has demonstrated good reliability.⁶⁵

Chinese version of the Massachusetts General Hospital Acupuncture Sensation Scale

The Massachusetts Acupuncture Sensation Scale is a revised version of a subjective acupuncture sensation scale originally developed for healthy subjects.⁶⁶ It includes 12 descriptors such as soreness, aching, deep pressure, heaviness, fullness/distension, tingling, numbness, sharp pain, dull pain, warmth, coldness and throbbing, all considered relevant to acupuncture treatment. Participants rate the intensity of each sensation on a scale from 0 to 10, where 0 indicates no sensation and 10 indicates an unbearable sensation. The Chinese version of the scale has demonstrated good reliability and validity.^{67 68}

Kinesthetic and Visual Imagery Questionnaire-10

The KVIQ-10 consists of five distinct movements, including two upper limb, one trunk and two lower limb actions.⁶⁹ It distinguishes between two primary sensory modalities used during imagery—visual and kinesthetic—and assesses MI ability by evaluating vividness. This differentiates visual MI, reflecting image clarity, from kinesthetic MI, reflecting sensation intensity.⁷⁰ The KVIQ-10 can detect poor or good imagery ability, enabling optimisation of MI-based therapy by identifying suitable candidates for mental practice training and tailoring treatment strategies accordingly. The Chinese version of the KVIQ-10 has demonstrated good construct validity.⁷¹

The KVIQ-10 will primarily serve as a screening tool to ensure that participants possess adequate MI ability (score ≥25) for inclusion.⁷² Although it is also listed as a secondary outcome, it is acknowledged that the KVIQ-10 was originally developed for screening rather than longitudinal follow-up. Therefore, in this study, KVIQ-10 scores will not be interpreted as a reliable efficacy endpoint but will provide supplementary information on imagery vividness across the intervention period.

EEG-fNIRS system

To enable simultaneous acquisition of EEG and fNIRS signals, this study integrates EEG electrodes and fNIRS optodes into a custom-designed cap. All participants undergo EEG-fNIRS assessment using an integrated device at the Department of Neurorehabilitation, Shanghai Second Rehabilitation Hospital. The device combines a 64-channel wireless EEG system (NeuSen.W64, Neuracle, China) with a multichannel fNIRS system (Nirsmart, Huichuang, Danyang, China) (figure 5).

Figure 5. EEG–fNIRS system. (A) Layout of the 64-channel EEG system; (B) Layout of the 40-channel fNIRS system; (C) Combined configuration of the integrated EEG–fNIRS system. EEG–fNIRS, electroencephalography-functional near-infrared spectroscopy.

EEG electrodes are arranged according to the international 10–20 system,⁶⁸ with 64 channels evenly distributed across the scalp. CPz serves as the reference electrode and AFz as the ground electrode to ensure signal stability. Signals will be sampled at 1000 Hz to allow high-resolution acquisition of cortical dynamics. Electrode impedance will be continuously monitored throughout the experiment. Conductive gel (GREENTEK GT10; Wuhan GreenTech Technology, Wuhan, China) will be applied to maintain impedance below 10 kΩ, thereby ensuring data quality.

fNIRS employs dual-wavelength detection at 730 nm and 850 nm to measure changes in oxyhaemoglobin (HbO₂) and deoxyhaemoglobin (HbR) concentrations in the cerebral cortex. Sixteen sources and eighteen detectors are arranged over the scalp following the international 10–20 system,⁶⁸ forming 40 channels with a fixed source–detector separation of 3 cm to optimise signal penetration and depth sensitivity. The sampling rate will be set at 11 Hz, which is sufficient to capture haemodynamic fluctuations. A 3D positioning system (NirSpace) will be used to record the spatial coordinates of each source and detector in Montreal Neurological Institute (MNI) space, enabling accurate spatial localisation for subsequent cortical mapping.

To improve compliance and ensure the successful acquisition of cortical signals before and after the intervention, EEG-fNIRS assessment will be conducted only on participants willing to undergo this procedure. It will not be mandatory for all subjects.

Statistical analyses

Clinical data analysis

Descriptive statistics will be performed using IBM SPSS V.26.0. Normality of continuous variables will be assessed using the Shapiro-Wilk test. Normally distributed data will be presented as mean±SD, while non-normally distributed data will be expressed as median with IQR. Categorical variables will be summarised as frequencies and percentages. Independent samples t-tests will be applied for normally distributed continuous variables, whereas the Mann-Whitney U test will be used for skewed data. χ² or Fisher’s exact tests will be employed for categorical variables.

Missing data for both primary and secondary outcomes will be handled using multiple imputation under the assumption of missing at random (MAR). Sensitivity analyses will be conducted to assess the robustness of results under different imputation strategies.

The primary analysis will follow the intention-to-treat (ITT) principle, with all randomised participants analysed according to their allocated groups. Per-protocol (PP) analyses will be conducted as secondary and sensitivity analyses to confirm the robustness of the main findings. Prespecified exploratory subgroup analyses will be performed based on clinically relevant factors, including time since stroke onset and the affected hemisphere; these will be interpreted cautiously, with interaction terms introduced into mixed-effects models as appropriate.

An interim analysis will be conducted when approximately 50% of participants (n=28) have completed the primary outcome assessment. The analysis will be performed by an independent statistician blinded to group allocation and reviewed by the Data Monitoring Committee (DMC). To maintain the overall two-sided type ie,rror rate at 0.05, an O’Brien-Fleming alpha-spending function will be applied.⁷³ The nominal significance level is set at p<0.005 for the interim analysis and p<0.048 for the final analysis. Early termination for efficacy will be considered if the interim results cross the predefined boundary; otherwise, the trial will continue until the planned sample size is reached.

Intervention effects on the primary outcome (FMA-UE) and secondary outcomes (MBI, BRS, MAS, HAMD-17) will be assessed using generalised estimating equations. A nominal p<0.048 will indicate statistical significance, accounting for the planned interim analysis using the O’Brien-Fleming α-spending function.

Imaging data analysis

To minimise potential selection bias in the EEG–fNIRS substudy, several measures will be undertaken. First, all eligible participants will be informed at enrolment that EEG–fNIRS assessments are optional, entirely voluntary and require a separate informed consent. Second, randomisation will be completed prior to the invitation for imaging, ensuring that willingness to participate does not influence group allocation. Third, baseline demographic and clinical characteristics will be compared between participants who do and do not undergo EEG–fNIRS to assess representativeness. Finally, imaging outcomes will be considered exploratory mechanistic indicators rather than primary endpoints, thereby preventing any potential bias from affecting the principal trial conclusions.

fNIRS data processing

fNIRS measures changes in haemoglobin concentration, indirectly reflecting cerebral blood flow and oxygen saturation, which indicate metabolic activity in brain regions. The raw fNIRS data represent only light intensity signals and contain considerable noise.

Data preprocessing will be performed using MATLAB-based NirSpark 1.7.5 (Huichuang, Danyang, China) or MATLAB software. The preprocessing pipeline will include the following steps: (1) automatic removal of motion artefacts caused by head movement; (2) conversion of light intensity signals to optical density; (3) application of a band-pass filter (0.01 Hz<f < 0.2 Hz) to eliminate physiological noise such as cardiac pulsation (1–1.5 Hz), respiration (0.2–0.5 Hz) and low-frequency drift (<0.09 Hz)⁷⁴; (4) conversion of optical density data into concentrations of HbO₂ and HbR based on the modified Beer-Lambert law.⁷⁵ Further analysis will focus on HbO₂ signals.

The general linear model is widely used in fNIRS studies to quantify cortical activation by estimating the magnitude and significance of haemodynamic responses.⁷⁶

To explore the mechanisms of electroacupuncture, FC analysis will be conducted to assess the effects on cortical network connectivity. Pearson’s correlation coefficients between HbO₂ concentration signals across channels will be used to characterise FC.⁷⁷ Since correlation coefficients (r values) cannot be directly averaged, Fisher’s r-to-z transformation will convert r values to z scores.⁷⁸ These z scores will represent the strength of FC between channels and will be averaged across regions of interest. An independent samples t-test will be conducted to assess differences in network connectivity between the two electroacupuncture conditions. A p<0.05 will indicate statistical significance.

EEG data processing

At the initial stage of EEG data preparation, raw signals will be sampled at 1000 Hz. A band-pass filter (0.5–45 Hz) will then be applied to attenuate high- and low-frequency noise. EEG signals will be referenced to the average potential across all electrodes to minimise bias from any single reference electrode. The data will subsequently be subjected to independent component analysis (ICA) to remove artefacts related to ocular movements and muscle activity. EEG signal processing and analysis will be conducted using MATLAB 2022b.

EEG activity is typically analysed across four canonical frequency bands: delta (δ, 0.5–4 Hz), theta (θ, 4–8 Hz), alpha (α, 8–12 Hz) and beta (β, 12–30 Hz),⁷⁹ which are frequently investigated for their relevance to neural synchronisation. Resting-state EEG data will be analysed in the frequency domain using fast Fourier transform (FFT) to obtain spectral characteristics within these bands for each participant.

To quantify FC, the HERMES toolbox (http://hermes.ctb.upm.es/)⁸⁰ will be employed to compute pairwise connectivity between electrodes over predefined cortical regions. FC metrics will include coherence, phase lag index (PLI), phase locking value, weighted PLI and Granger causality,⁸¹ which capture both phase-based and causality-based interactions. All analyses will be performed under identical resting-state conditions across participants. Group-level comparisons will be made between (a) intervention versus control and (b) preintervention versus postintervention resting-state FC matrices. Multiple comparisons will be corrected using the Network-Based Statistic method.⁸²

Data collection, management and monitoring

Data collection will be conducted by trained personnel who are independent of the randomisation and intervention procedures. All collected data will be promptly entered into a designated case report form (CRF). Subsequently, data entry staff will digitise all information, ensuring patient confidentiality and data security through the use of unique participant identifiers.

DMC will be responsible for verifying whether the conduct, recording and reporting of the trial comply with the protocol and regulatory requirements, based on a predefined monitoring plan and standard operating procedures. The DMC will regularly review the collected data for consistency, logic and validity, including checks for adherence to predefined criteria and guidelines for data entry and storage. Source data verification will be performed by comparing entered data against original records to ensure accuracy and consistency.

Any discrepancies or anomalies identified will prompt immediate investigation and correction by the DMC. Communication with data collectors, investigators or other relevant personnel may be undertaken to clarify and resolve any issues. The primary objective of data monitoring is to minimise data errors and bias, and to ensure high data quality that underpins accurate statistical analyses and reliable research conclusions.

Withdrawal

Participants may voluntarily withdraw from the study at any time by submitting a written request, due to reasons such as intolerance to the intervention (eg, discomfort from electroacupuncture or anxiety triggered by MIT), or for personal reasons (eg, hospital discharge or scheduling conflicts). In addition, withdrawal may be initiated by the investigator in the event of serious AEs, such as skin infection or nerve injury caused by electroacupuncture, or significant psychological distress or marked fluctuations in blood pressure or heart rate induced by MIT.

All withdrawal events will be recorded by the safety supervisor, including the participant’s details and time of withdrawal, and reported to the DMC. The final decision on withdrawal will be made following DMC review.

Where possible, a final clinical assessment and relevant data collection will be conducted in accordance with the participant’s preferences. Participants’ autonomy will be fully respected, and appropriate medical follow-up will be ensured. All withdrawn cases will be included in the ITT analysis to maintain data integrity. Serious AEs will be reported promptly in accordance with ethical guidelines and regulatory requirements.

Adverse events

All AEs, concomitant medications and concurrent interventions will be accurately recorded throughout the trial. The use of concomitant medications or additional interventions will not be restricted. Participants on long-term medication will be required to maintain their regular medication regimens from baseline through follow-up. Any changes in medications or interventions will be documented by the outcome assessors.

Although both electroacupuncture and MIT are widely regarded as safe therapeutic approaches, the occurrence of AEs cannot be entirely excluded. Potential AEs and corresponding management strategies are detailed in the online supplemental material 1.

Quality control

Before the trial commences, all researchers will receive training on participant recruitment, intervention delivery, outcome assessment and data entry. Acupuncturists performing electroacupuncture must hold a certified Traditional Chinese Medicine practitioner licence and have at least 3 years of clinical acupuncture experience. Personnel delivering MIT must be licensed rehabilitation physicians with a minimum of 3 years of rehabilitation training experience. They are required to complete specialised MIT training and pass the KVIQ-10 scale evaluation, demonstrating accurate interpretation of scale scores relative to imagery ability. Rehabilitation physicians must conduct at least 10 supervised mock sessions and pass an assessment—evaluating clarity of task instructions and achieving a participant imagery task completion rate of ≥85%—before commencing official training.

The Jin’s three-needle protocol will be strictly followed, targeting Quchi (LI11), Waiguan (TE5) and Hegu (LI4). Stimulation parameters are set to 2 Hz intermittent wave, with intensity adjusted to patient tolerance, maintaining treatment intensity variability within 10% for each patient per session. Needles will remain in situ for 20 min, with electroacupuncture device output stability monitored every 5 min (current intensity error ≤±0.5 mA).

MI videos, produced using E-Prime 3.0, include two standardised movement modules—elbow flexion and grasping—each randomly played 15 times. Each imagery lasts 17 s, followed by a 20 s rest interval.

The DMC will oversee adherence to protocols throughout the study. The DMC comprises two independent stroke rehabilitation experts and one statistician not involved in the project. Meetings will be held quarterly to review recruitment progress, protocol adherence, CRF completion, data quality, follow-up loss and safety data. In the event of a serious AE, the DMC will convene an emergency meeting to decide whether to terminate the trial early.

Patient and public involvement

Patient and public involvement will be actively integrated during the design phase of this study, with the aim of optimising the intervention protocol and research procedures for electroacupuncture synchronised with MIT in the rehabilitation of poststroke upper limb motor dysfunction. A total of six stroke survivors with upper limb impairment, three caregivers and three clinical professionals will be recruited to participate in semistructured interviews and focus group discussions.

Ethics and dissemination

This study has been approved by the Ethics Committee of Shanghai Second Rehabilitation Hospital (approval number: 2025-18-01) and has been registered with the International Traditional Medicine Clinical Trial Registry (ITMCTR; registration number: ITMCTR2025001311). The trial will be conducted in accordance with the Declaration of Helsinki, relevant national regulations and the policies of the ethics committees at each participating centre. Written informed consent will be obtained from all participants or their legal guardians before enrolment. The participant consent form is available as online supplemental material 2. Data collected during the trial will be used solely for this clinical research. Participants’ privacy will be strictly protected, and no identifiable personal information will be disclosed. It will not be possible to identify individual participants from any disseminated results.

Discussion

This single-centre, RCT is based on the CPC closed-loop rehabilitation theory and aims to evaluate the efficacy of electroacupuncture synchronised with MIT in improving upper limb motor function after stroke. Electroacupuncture is expected to stimulate peripheral acupoints such as LI4 and LI11, enhancing somatosensory afferent input and modulating cerebral perfusion.⁸³ Concurrently, MIT is intended to engage central motor circuits by internally simulating movement, thereby activating the ipsilesional M1 and supplementary motor area.⁸⁴ According to the CPC theory,^{9 85} central interventions enhance synaptic plasticity and neural remodelling in perilesional regions, whereas peripheral inputs facilitate synaptogenesis.^{86 87} The integration of both mechanisms creates a positive feedback loop that amplifies neuroplasticity and promotes functional recovery.

The CPC closed-loop theory proposes that central neuromodulation facilitates synaptic plasticity and neural remodelling in perilesional regions, while peripheral sensory input promotes functional synaptogenesis. The combination of these mechanisms forms a positive feedback loop that may amplify neuroplastic changes and support functional restoration.²¹ Existing evidence lends support to this model. For example, Miao et al⁸⁸ demonstrated that virtual reality (VR) combined with functional electrical stimulation (FES) significantly increased β-band ERD, suggesting enhanced recruitment of motor networks.⁸⁹ Similarly, Kaneko et al⁹⁰ observed that MI-based brain–computer interface (MI-BCI) paired with FES elicited cortical excitability comparable to actual movement, as reflected in elevated motor-evoked potential amplitudes. EEG studies by Yakovlev et al⁹¹ further indicated that MI-BCI combined with FES enhanced μ-rhythm suppression in sensorimotor regions, consistent with improved central motor network activation. These findings support the use of closed-loop approaches to reinforce central neuroplasticity.

The bidirectional regulatory potential of acupuncture forms a core innovation of this combined strategy. Centrally, prior evidence suggests that electroacupuncture activates bilateral prefrontal cortices and the right M1, thereby enhancing cognitive–motor coupling.^{16 17} Preclinical studies also indicate that electroacupuncture may reduce infarct volume and strengthen inter-regional FC within the motor network.⁹² Peripherally, a clinical study¹⁸ showed that electroacupuncture improved forearm torque and movement velocity in stroke patients, and animal models suggest accelerated neuromuscular repair and functional gains within 3–7 days.¹⁹ Compared with other closed-loop interventions such as VR–FES, electroacupuncture not only provides consistent somatosensory input but also modulates the release of neurotransmitters—such as dopamine and GABA—through neuroanatomical pathways associated with specific acupoints.⁹³ By combining peripheral acupoint stimulation with MIT, the proposed intervention targets both afferent and cognitive–motor systems, potentially expanding the scope and depth of neuromodulation beyond conventional protocols.

This protocol also incorporates several methodological refinements relative to previous studies. For example, one earlier trial³⁹ recruited 17 participants for a 2-week intervention combining acupuncture and MIT, limiting the generalisability of their findings. In contrast, this study will enrol 72 participants and extend the intervention duration to 8 weeks (including follow-up), using standardised electroacupuncture parameters and a customised MIT video paradigm. Furthermore, unlike previous research that relied solely on fMRI to examine postacupuncture brain activity,⁹⁴ this trial will adopt a multimodal EEG–fNIRS approach to construct an electrophysiology–metabolism coupling model. Data will be collected during both resting and task states, allowing spatiotemporal characterisation of underlying neurophysiological mechanisms.

If effective, this synchronised intervention could provide new theoretical support for cognitive–motor integrated rehabilitation, promoting a shift from motor-centric to neurocognitive-based strategies. Furthermore, the standardised intervention protocol and multimodal analytical framework proposed here may inform future research and contribute to more precise investigation of poststroke brain plasticity.

However, this study has limitations. First, the study population is restricted to subacute stroke patients (1–6 months postonset), excluding patients outside the subacute phase or those with significant cognitive impairment, which may limit the generalisability of findings. Second, this trial will be conducted at a single centre in Shanghai, introducing geographical bias. Third, individual differences in MI capacity and adherence to the protocol may influence the intervention’s effectiveness. Therefore, future multicentre trials with wider recruitment criteria and extended follow-up periods are warranted to confirm the feasibility and scalability of the proposed intervention approach.

Taken together, this trial will explore a rehabilitation strategy based on the CPC closed-loop model, combining electroacupuncture and MIT to promote functional recovery after stroke. Although limitations exist, the theoretical grounding, methodological rigour and clinical accessibility of this approach may offer valuable insights for future neurorehabilitation research and practice.

Trial status

The protocol version number and date: V.1.0, 1 April 2025. The study protocol was approved on 14 April 2025 by the Ethics Committee of Shanghai Second Rehabilitation Hospital (approval number: 2025-18-01). The trial was registered on 1 July 2025 with the International Traditional Medicine Clinical Trial Registry (ITMCTR; registration number: ITMCTR2025001311), and the latest update was submitted on the same date. Participant recruitment commenced on 10 July 2025 and is expected to be completed by 20 March 2026. The findings of this study will be disseminated through publication in peer-reviewed journals.