17 March 2026 : Review article
Application and Research Progress of BCI in Post-Stroke Psychiatric Disorders: A Narrative Review
Zekai HuDOI: 10.12659/MSM.951399
Med Sci Monit 2026; 32:e951399
Table 2 Summary of representative BCI studies on post-stroke psychiatric and cognitive applications.
| Target disorder | BCI modality & paradigm | Technical features | Key findings | Study |
|---|---|---|---|---|
| Mental stress/anxiety | EEG (passive monitoring) | Deep learning (LSTM-RNN) for state classification | Achieved high accuracy in detecting stress states; demonstrated potential for breathing entrainment interventions | Sundaresan et al []58 |
| Depression & anxiety | EEG-based neurofeedback | Audio-visual entrainment | Significant reduction in depression (HAM-D) and anxiety scores in RCT; modulation of beta/theta rhythms | Pino []61 |
| Cognitive impairment | P300-BCI (cognitive training) | Event-Related Potential (ERP) based training | RCT showed significant improvements in MoCA scores (attention, naming) compared to control group | Fateeva et al []64 |
| Attention deficits | BCI + fMRI (multimodal) | Resting-state functional connectivity analysis | BCI attention training induced rapid strengthening of functional connectivity in attention networks | Mayorova et al []65 |
| Anxiety states | EEG (closed-loop) | Affective BCI with real-time feedback | Successfully classified anxious states and provided objective feedback for self-regulation | Chen et al []71 |
| BCI – brain computer interface; EEG – electroencephalography; LSTM – long short-term memory; RNN – recurrent neural networks; HAM-D – Hamilton Rating Scale for Depression; RCT – randomized clinical trials; MoCA – Montreal cognitive assessment test; fMRI – functional magnetic resonance imaging. | ||||






