ReHAND-BCI
(2025)Objective
To assess the efficacy and neuroplastic effects of a brain-computer interface (BCI) intervention using the ReHand-BCI system with robotic hand orthosis feedback for upper extremity stroke neurorehabilitation, compared to a sham-BCI control.
Study Summary
• The experimental group showed trends toward more pronounced ipsilesional cortical activity, higher ipsilesional corticospinal tract integrity (rFA), and lasting corticospinal excitability at follow-up, though differences were not statistically significant, likely due to the small sample size
Intervention
BCI therapy using the ReHand-BCI system: 30 sessions (5 per week for 6 weeks) of EEG-based brain-computer interface controlling a robotic hand orthosis via motor intention decoding, compared to sham-BCI where the orthosis activated randomly independent of motor intention
Inclusion Criteria
Adults >18 years; any sex; first ischemic or hemorrhagic stroke in either hemisphere confirmed by neuroimaging; 3-24 months since stroke onset; hand paresis (Motricity index 0-22); right-handed before stroke; normal or corrected vision; at most mild alterations in attention and memory per NEUROPSI; no previous neurological diseases
Study Design
Arms: Experimental Group (BCI therapy with ReHand-BCI) vs. Control Group (sham-BCI therapy)
Patients per Arm: Experimental: 12 allocated, 10 analyzed; Control: 11 allocated, 9 analyzed
Outcome
• Only the experimental group had significantly different ARAT scores at post-treatment (EG: baseline 8.5 to post-treatment 20) while CG showed significant improvement only at follow-up (baseline 3 to follow-up 15)
• No significant intergroup or intragroup differences in secondary outcomes (laterality coefficient, laterality index, rFA); experimental group showed trends toward more ipsilesional cortical activity and higher corticospinal tract integrity
Bottom Line
Both BCI and sham-BCI groups showed clinically meaningful improvements in upper extremity sensorimotor function (FMA-UE and ARAT), but no statistically significant intergroup differences were observed. The experimental group showed non-significant trends toward greater ipsilesional cortical activity, higher corticospinal tract integrity, and more lasting corticospinal excitability, suggesting potential neuroplasticity advantages of closed-loop BCI feedback that the study was underpowered to detect.
Major Points
- Both groups demonstrated significant intragroup improvements in FMA-UE from baseline to post-treatment and follow-up, with changes within the range of the minimal clinically important difference (5.25 points)
- Only the experimental group showed significant ARAT improvement at post-treatment (baseline 8.5 to post-treatment 20), while the control group achieved significance only at follow-up (baseline 3 to follow-up 15)
- No statistically significant intergroup differences were found for any primary or secondary outcome at any timepoint
- The experimental group showed trends toward more pronounced ipsilesional cortical activity measured by alpha LC during affected hand motor intention, and higher ipsilesional corticospinal tract integrity measured by rFA
- BCI classification accuracy was similar between groups (experimental 67.3% +/- 8.97% vs. control perceived 69.28% +/- 4.1%, p=0.557), indicating comparable engagement with the system
- Target sample size of 30 was not reached due to conclusion of study funding; only 19 patients were analyzed, making the study likely underpowered to detect intergroup differences
- This is the first controlled clinical trial to assess such a wide range of physiological effects (EEG, fMRI, DTI, TMS) across a long BCI intervention for stroke rehabilitation
Study Design
- Study Type
- Randomized Controlled Trial
- Randomization
- Yes
- Blinding
- Triple-blinded (patients, clinicians performing outcome assessments, personnel performing statistical analyses)
- Sample Size
- 23
- Follow-up
- 6 months after baseline measurement (T3)
- Centers
- 1
- Countries
- Mexico
Primary Outcome
Definition: Fugl-Meyer Assessment for the Upper Extremity (FMA-UE) and Action Research Arm Test (ARAT) measured at T0 (baseline), T1 (after 15 sessions), T2 (end of 30 sessions), and T3 (6 months post-baseline)
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| FMA-UE: T0 median 26 [16,37.5], T2 median 34 [17.3,46.5]; ARAT: T0 median 3 [1.8,30.5], T3 median 15 [2.5,40.8] | FMA-UE: T0 median 24.5 [20,36], T2 median 28 [23,43]; ARAT: T0 median 8.5 [5,26], T2 median 20 [7,36] | - | FMA-UE intergroup: T0 p=0.656, T1 p=1, T2 p=0.656, T3 p=0.656; ARAT intergroup: T0 p=1, T1 p=1, T2 p=1, T3 p=0.656 (all non-significant) |
Limitations & Criticisms
- Target sample size of 30 patients was not reached (only 23 enrolled, 19 analyzed) due to conclusion of study funding, making the study underpowered to detect intergroup differences
- Heterogeneity of stroke lesions among participants (cortical, subcortical, cortical-subcortical) may have contributed to variability in recovery profiles
- Most participants were in the chronic stage with some in late subacute stage, introducing variability from different stroke etiologies and stages following distinct recovery trajectories
- No significant intergroup differences were found for any primary or secondary outcome, so the primary hypothesis that the experimental group would have higher recovery was not confirmed
- The sham-BCI intervention provided similar robotic feedback (random activation 55-85%) which may have elicited similar therapeutic effects as actual BCI, making differentiation difficult
- Conventional therapy was not restricted between T2 and T3 follow-up period, potentially confounding long-term outcome comparisons
- MEPs could not be recorded in all patients due to severely affected corticospinal tracts, reducing the TMS analysis sample further
Citation
Cantillo-Negrete J, Rodriguez-Garcia ME, Carrillo-Mora P, Arias-Carrion O, Ortega-Robles E, Galicia-Alvarado MA, Valdes-Cristerna R, Ramirez-Nava AG, Hernandez-Arenas C, Quinzanos-Fresnedo J, Pacheco-Gallegos MdR, Marin-Arriaga N and Carino-Escobar RI (2025) The ReHand-BCI trial: a randomized controlled trial of a brain-computer interface for upper extremity stroke neurorehabilitation. Front. Neurosci. 19:1579988. doi: 10.3389/fnins.2025.1579988