For the roughly one in three people with depression who do not improve after multiple medications, the options narrow fast. Talk therapy helps some, electroconvulsive therapy helps others, but a sizable group stays stuck. A new registered trial in China is testing whether a brain–computer interface (脑机接口) approach — clever stimulation that listens to the brain before acting — can break that stall.
The design matters as much as the idea.
Key takeaways
- A multicenter trial registered as NCT07783906 will test precision closed-loop neuromodulation for treatment-resistant major depressive disorder in China.
- Sponsor is Shanghai Mental Health Center (上海市精神卫生中心); sites are in Shanghai, Beijing (Anding Hospital) and Beijing (Xuanwu Hospital).
- Two approaches are compared against sham: closed-loop deep-brain stimulation (闭环深部脑刺激) and transcutaneous auricular vagus nerve stimulation (经皮耳迷走神经刺激).
- Planned enrollment is 70 adults (18–70); the primary endpoint is HAMD-17 change at 8 weeks.
- Status is "not yet recruiting" as of late August 2026; it is a registered study, not an available treatment.
Why "closed-loop" changes the math
Open-loop brain stimulation fires on a fixed schedule. Closed-loop (闭环) stimulation, by contrast, reads the brain's own electrical signal and adjusts or triggers only when the pattern calls for it. The theory is that matching stimulation to a person's live neural state should work better and disturb normal function less than a constant pulse.
For depression, two closed-loop routes are in play. One is deep-brain stimulation (DBS), where electrodes are implanted in mood-related circuits and stimulation is tuned to neural biomarkers. The other is transcutaneous auricular vagus nerve stimulation (taVNS), a non-invasive method using ear electrodes — no surgery. The trial pits each against a sham version, so any benefit cannot be explained away by placebo or expectation.
What the trial actually does
The registry entry describes a multicenter, prospective interventional study. Eligible participants are adults 18 to 70 with major depressive disorder who meet strict treatment-resistance criteria: at least 24 months of depression, inadequate response to at least two antidepressants of adequate dose and duration, a HAMD-17 score above 17 at baseline, and a stable medication regimen for at least a month. Within each stimulation group, participants are randomized to active or sham stimulation.
The primary outcome is the change in the 17-item Hamilton Depression Rating Scale (HAMD-17) from baseline to week 8. Secondary assessments cover anxiety, cognition, social function, EEG activity, MRI, and adverse events. Estimated enrollment is 70; the responsible party is listed as Lin Sun of Shanghai Mental Health Center, and the study identifier is 2026-63.
Why a sham-controlled psychiatric BCI trial is notable
Brain–computer interface (脑机接口) news in China has been dominated by movement restoration — helping paralyzed patients move a cursor or a hand. Psychiatric applications are harder to validate because mood is subjective and placebo effects are strong. A sham-controlled design is the standard way to separate real signal from wishful thinking, and it is still rare in this corner of the field.
The three-site structure also helps. Shanghai Mental Health Center, Beijing Anding Hospital (北京安定医院), and Xuanwu Hospital (宣武医院) are major psychiatric and neurological centers; spreading the study across them reduces the chance that one team's technique or patient mix drives the result.
The honest caveats
This is not yet recruiting, and the posted timelines are estimates: a study start around August 2026 and estimated primary and study completion at the end of 2028. "Phase not applicable" is listed because the interventions are devices evaluated for their effect rather than dosed like a drug. Treatment-resistant depression is a serious condition with suicide risk, and the protocol explicitly includes patients with a history of self-harm or suicidal thoughts — appropriate for the population, but a reminder that such trials carry real ethical weight and need careful monitoring.
How it fits China's wider BCI momentum
The trial sits inside a broader national push. China's 2025 multi-ministry opinion on brain–computer interface (脑机接口) listed neurological and psychiatric application as a priority, and implantable devices received priority registration guidance in principle. But a registered, sham-controlled psychiatric trial is a different animal from a policy statement: it is the kind of evidence that, if positive, could eventually support reimbursement and clinical adoption rather than just demonstration.
Honest limitations
All facts here are drawn from the ClinicalTrials.gov record for NCT07783906 (first posted 25 August 2026, last updated 28 August 2026, status "not yet recruiting") and mirrored registry summaries. The trial has not started enrolling, so no efficacy or safety results exist yet; the 70-participant target and 8-week primary endpoint are planned, not observed. "Closed-loop DBS" and "taVNS" mechanisms are described by the registry, not independently verified in a published paper for this study. The "precision" claim about targeting specific circuits is the trial's hypothesis, not a proven outcome. No RMB amounts appear in the registry summary, so no currency conversion applies. This article is not medical advice.
What readers can do now
- If you or a loved one has treatment-resistant depression, discuss evidence-based options — including ECT and vagus-nerve stimulation — with a psychiatrist; this trial is not yet open for enrollment.
- When you read "BCI cures depression" headlines, check whether the claim comes from a controlled trial registry or a demo; controlled design is the difference.
- For researchers and funders, sham-controlled psychiatric neuromodulation is where the field needs proof — support replication and long-term follow-up, not just first results.