BCIwiki — Brain-computer interface news, research and industry database

Why it matters Psychological interventions for adolescent depression often struggle with adherence and vary widely in effectiveness. This trial brings a brain-computer interface into mindfulness training, using EEG to track brain activity during sessions, aiming to make the intervention responsive to a patient's brain state rather than a fixed protocol. Only the title and registration number are public so far; recruitment has not begun, and the design, sample size and sponsor remain undisclosed, so it is too early to judge feasibility.

Why it matters The study turns flexible-electrode claims into human single-neuron yield and failure data, while leaving a large gap between stable intraoperative recording and chronic implantation.

Why it matters This is an invasive BCI trial registered by Johns Hopkins University, testing an implant that both records and stimulates brain areas in people with paralysis from conditions such as tetraplegia and ALS. Now recruiting, the trial will assess whether the device can help patients control assistive devices and receive feedback. It is among the first studies of an implant that both records and stimulates, potentially enabling bidirectional interaction. Still at an early feasibility stage with limited sample size, it remains far from clinical application.

Why it matters EEG decoding has defaulted to bigger models. Finding that the Transformer's attention collapses on RSVP and that removing it costs nothing is a direct prompt for teams building lightweight, deployable BCI decoders, and a reminder that paradigm-specific priors can outweigh architecture.

Why it matters Almost every rehabilitation BCI decodes an imagined limb movement. Swallowing is not one, which makes this China's first test of how far motor imagery carries past the arm, into a complication that is common after stroke and short on effective therapy. Recruitment has not opened.

Why it matters The RMB 2 billion round raises the capital benchmark for Chinese BCI companies and makes revenue growth, medical registrations and manufacturing delivery more measurable expectations.

Why it matters Selecting source domains dynamically instead of pooling them naively targets negative transfer — the failure mode driven by low source-target similarity and inter-subject variability that undercuts cross-subject MI-BCI transfer.

Why it matters The notable part is that a hospital is leading it, not a university or a company. Provincial innovation centres in China are usually headed by a dominant firm or a university; putting the BCI one under a clinical institution suggests Hubei treats the clinical entry point as the critical position in the chain rather than treating hospitals as trial sites. Among the centre's stated priorities is domestic sourcing of core components, one of the harder constraints on Chinese invasive BCI work. The same announcement notes that Tongji is the first hospital in central China to pursue both implanted and non-implanted approaches systematically, that it has assembled optical, electrical, magnetic and focused-ultrasound modalities, and that it had run more than 50 BCI-related clinical trials by April 2026 — an unusual volume for a single Chinese hospital.

Why it matters The 83.93% and 88.38% results on the benchmark Competition IV datasets are competitive, but the more telling number is 67.67% on real stroke-patient data — a step toward validating MI decoders on the clinical population where neurorehabilitation BCIs would actually deploy.

Why it matters Training-free, real-time channel pruning targets a practical bottleneck — bad channels that degrade online performance — without per-user calibration, making SSVEP systems easier to deploy.