Development and feasibility of a motor imagery-based brain-computer interface-controlled closed-loop functional electrical stimulation system for swallowing rehabilitation.
Yu C, Dong X, et al. • Journal of neural engineering • 2026
A motor imagery-based brain-computer interface-controlled closed-loop swallowing functional electrical stimulation system demonstrated technical feasibility, acceptable short-term usability, and no major safety concerns in post-stroke dysphagia patients.
Key Findings
Results
During swallowing motor imagery, Fp1 spectral power ratios decreased relative to rest across all participant groups.
The δ/α ratio decreased significantly in all three groups (healthy controls, post-stroke dysphagia patients, and post-stroke patients without dysphagia).
The δ/(α + β) ratio and the (δ + θ)/(α + β) ratio decreased significantly in healthy controls and post-stroke patients without dysphagia, and showed the same downward trend in post-stroke dysphagia patients.
These EEG features were identified at electrode Fp1 and used as the basis for the decoder.
Groups included healthy controls (n=15), post-stroke dysphagia patients (n=15), and post-stroke patients without dysphagia (n=15).
Results
Post-stroke dysphagia patients exhibited higher θ-band power at electrode T3 compared to healthy controls and post-stroke patients without dysphagia.
The difference in θ-band power at T3 was statistically significant (P = 0.0382).
This neurophysiological difference was observed between PSD patients and both HC and PSND groups.
This finding suggests a potential neurophysiological marker distinguishing dysphagia from non-dysphagia post-stroke states during swallowing motor imagery.
Results
A threshold-based decoder using Fp1 spectral power ratios achieved a mean classification accuracy of 71.5% in an independent validation cohort.
The decoder was validated in an independent healthy control cohort of n=10 participants.
Mean classification accuracy was 71.5%.
The decoder was based on spectral power ratios at electrode Fp1 during swallowing motor imagery versus rest.
Results
The MI-BCI-controlled closed-loop swallowing FES intervention demonstrated 100% adherence across 10 sessions in post-stroke dysphagia patients.
10 patients with post-stroke dysphagia completed 10 sessions of MI-BCI-controlled closed-loop swallowing FES over 2 weeks.
Adherence was 100% across all participants and sessions.
Patients achieved 29.8 ± 6.2 successful closed-loop triggers per session.
No serious adverse events were recorded during the intervention period.
Results
The system received an acceptable usability rating with a mean System Usability Scale score of 72.8 ± 4.2.
The mean System Usability Scale (SUS) score was 72.8 ± 4.2 among the 10 PSD patients.
A SUS score above 70 is generally considered 'good' usability.
Usability was assessed alongside feasibility and safety as primary outcomes of Experiment 2.
Background
Conventional swallowing FES is typically delivered in open loop or triggered by peripheral signals, which may not align with voluntary swallowing intention.
The authors identified this misalignment as a key limitation motivating the development of a BCI-controlled closed-loop system.
The proposed system uses motor imagery EEG signals to detect swallowing intention and trigger FES accordingly.
The system was designed specifically for post-stroke dysphagia rehabilitation.
What This Means
This research describes the development and early testing of a new system for helping stroke survivors who have difficulty swallowing. The system works by detecting brainwave patterns when a patient imagines swallowing, then uses those signals to automatically trigger electrical stimulation of the swallowing muscles — creating a closed loop between the patient's intention and the therapy. The study found that specific brainwave patterns at a forehead electrode reliably change during swallowing imagination across both healthy people and stroke patients, and a computer decoder could correctly identify these patterns about 71.5% of the time in a validation group.
In a small clinical test with 10 stroke patients who had swallowing difficulties, the system was used for 10 therapy sessions over two weeks. All patients completed every session (100% adherence), successfully triggered the stimulation about 30 times per session on average, rated the system as reasonably easy to use, and experienced no serious side effects. The researchers also found that stroke patients with swallowing problems showed a distinct brainwave signature compared to stroke patients without swallowing problems, suggesting there may be measurable brain differences underlying the condition.
This research suggests that linking a patient's conscious swallowing intention — detected through brainwaves — directly to electrical muscle stimulation is technically feasible and appears safe and usable in a small group of stroke patients. This approach could potentially make swallowing rehabilitation more effective by ensuring the therapy is delivered precisely when the patient is actively trying to swallow, which is thought to promote better brain recovery. Larger studies with longer follow-up are needed to determine whether this approach improves swallowing outcomes compared to existing treatments.
Yu C, Dong X, Zhang Y, Wan X, Zhou Y, Zheng Y, et al.. (2026). Development and feasibility of a motor imagery-based brain-computer interface-controlled closed-loop functional electrical stimulation system for swallowing rehabilitation.. Journal of neural engineering. https://doi.org/10.1088/1741-2552/ae730b