Noninvasive BCI-based cognitive rehabilitation in poststroke cognitive impairment: study protocol for a randomized controlled trial

Comprehensive Summary

This article presents a randomized, double-blind, sham-controlled clinical trial protocol investigating the efficacy of non-invasive brain–computer interface (BCI)-based cognitive rehabilitation in patients with post-stroke cognitive impairment (PSCI). PSCI is a common and disabling consequence of stroke, affecting attention, memory, and executive function, and is associated with reduced functional independence, lower quality of life, and increased long-term mortality. Conventional pharmacological and cognitive rehabilitation strategies are limited by heterogeneous pathology, modest efficacy, poor accessibility, and low adherence. Against this background, the authors propose BCI-based cognitive training as a novel intervention that enhances neuroplasticity through real-time neurofeedback and closed-loop brain–behavior interactions. The trial plans to enroll 66 stroke survivors with confirmed cognitive impairment (MoCA ≤ 24), randomly assigning them to either a BCI intervention group or a sham-controlled group, alongside standardized conventional cognitive therapy. The intervention consists of EEG-based neurofeedback training delivered five days per week for four weeks, in which attention-related neural signals are translated into real-time feedback during interactive cognitive tasks. The primary outcome is improvement in global cognitive function as measured by the Montreal Cognitive Assessment (MoCA) at four weeks, with secondary outcomes including executive function (Stroop test), electrophysiological markers (theta/beta power ratio and phase lag index), inflammatory and metabolic biomarkers (CRP and homocysteine), hippocampal volume, and functional follow-up at three months. Methodologically, the study is designed in accordance with SPIRIT and CONSORT guidelines, incorporating stratified block randomization, assessor blinding, intention-to-treat analysis, and rigorous quality control procedures. By integrating behavioral, electrophysiological, biochemical, and structural neuroimaging outcomes, the protocol aims not only to evaluate clinical efficacy but also to explore the neurophysiological mechanisms underlying BCI-induced cognitive recovery, particularly changes in cortical connectivity and neural oscillatory dynamics.

Outcomes and Implications

If successful, this trial could provide clinically meaningful evidence supporting non-invasive BCI-based neurofeedback as a viable therapeutic option for post-stroke cognitive impairment, a condition for which effective treatments remain limited. The use of EEG-driven feedback to directly engage impaired neural networks offers a mechanistically grounded alternative to conventional cognitive training, with the potential to enhance attention, executive control, and memory through activity-dependent neuroplasticity. Such an approach may be particularly valuable for stroke survivors who struggle with traditional rehabilitation due to physical disability, limited access to care, or poor long-term adherence. From a broader clinical perspective, integrating electrophysiological markers (theta/beta ratio, phase lag index) and biological indicators (CRP, homocysteine, hippocampal volume) could help identify objective biomarkers of cognitive recovery and treatment responsiveness. This may facilitate personalized neurorehabilitation strategies, allowing clinicians to tailor interventions based on neural and inflammatory profiles. Additionally, the non-invasive and relatively low-risk nature of BCI training enhances its feasibility for widespread clinical adoption, including outpatient and community-based rehabilitation settings. Overall, the study has the potential to advance post-stroke care by bridging cognitive rehabilitation with neurotechnology-driven precision medicine.