Papers

P300 BCI spellers are often built by switching on every component that works on its own, on the assumption that more is better. A preprint study tested that assumption with a four-component full-factorial experiment and found that a component's value is conditional, not additive. Subject calibration was the strongest single contributor; Euclidean Alignment made up for the lack of calibration in zero-calibration settings; and stacking components that are each useful on their own could lower performance, an effect the authors call component anti-synergy. Language model support was not universally beneficial either: its effect depended strongly on how strong the underlying EEG pipeline was.

Running a motor-imagery BCI while transcranial alternating current stimulation (tACS) is switched on has been difficult, because stimulation artifacts swamp the brain rhythms the decoder relies on and push accuracy close to chance. In a controlled test with 14 healthy participants, a real-time spatial filtering pipeline combining spatio-spectral decomposition with beamforming held accuracy at 76 ± 9% during stimulation, while a standard Laplacian filter managed only 58 ± 9%. That makes closed-loop "stimulate while reading" operation feasible at the signal level; whether it actually improves BCI performance or produces neuroplastic changes, the paper explicitly leaves to future research.

Ten native Mandarin speakers read aloud six single-vowel rhymes while optically pumped magnetometer magnetoencephalography (OPM-MEG) and EEG recorded the same task. In the 150-500 ms window after stimulus onset, OPM-MEG decoded significantly more accurately than EEG with all five classifiers tested. The best-performing combination, common spatial pattern features with a linear support vector machine, reached a mean accuracy of 60.3% for OPM-MEG versus 56.7% for EEG. That suggests speech BCIs that need no implanted electrodes and do not restrict head movement may have a clearer signal path than EEG.

The study proposes coating the gold pillar sidewalls of the PRIMA retinal prosthesis with titanium dioxide or platinum while protecting surrounding device regions. The approach combines photolithography and electrodeposition to address tissue contact with exposed gold while preserving surrounding implant functionality. The authors report that both coatings were biocompatible in in-vivo studies and prevented adverse retinal tissue reactions to gold. Titanium dioxide had a lower specific capacitance than platinum, which the authors say better limits sidewall current and directs charge injection mainly through the pillar tops. These findings concern electrode materials and interfaces, not demonstrated improvements in patients' vision from the new coatings.

Motor-imagery brain-computer interfaces vary widely between users and have little calibration data, and federated learning lets them share a model without uploading raw EEG. NEXUS-MI treats gateway synchronization as a joint learning-and-communication control problem: raw EEG and classifier heads stay local, while an edge coordinator maintains the shared backbone network. On the BCICIV-2a and OpenBMI datasets, communication-aware coordination cut server-to-client backbone traffic by about 42%, with small, implementation-dependent differences in cohort-average accuracy. Those averages masked individual vulnerability: on BCICIV-2a, losses relative to an ideal-link reference reached about 12 percentage points.

Participants steering a virtual avatar with a brain-computer interface reported a stronger sense that the avatar's body was their own when they imagined jogging than when they imagined opening their right hand, even though jogging imagery produced weaker EEG signals. Forward movement was driven by motor imagery-related sensorimotor rhythm event-related desynchronization from scalp EEG and direction by eye gaze, as participants guided a jogging avatar along a curved course before rating their embodiment in questionnaires. The researchers say neural signal strength and embodied experience can diverge, so the congruence between imagery and action should be weighed alongside standard decoding metrics.

Researchers at the University of California, San Francisco (UCSF), a public research university known for its neuroscience work, had two Parkinson's disease patients train at home with a brain-computer interface (BCI) airplane-simulator game, learning to down-regulate cortical beta activity and thereby control the intensity of their own deep brain stimulation (DBS). The work, posted as a preprint on medRxiv on August 17, 2026, points to BCI applications in neuromodulation and to more personalized treatment for Parkinson's and other conditions; its conclusions have not yet been peer reviewed.

Researchers have mapped the brain-computer interface (BCI) attack surface along five dimensions (forged neural signals, desynchronization-based evasion, replay hijacking, "Vein Tapping" eavesdropping and embedded backdoors), which they group as "NERVE Attacks" and describe as orthogonal and together spanning the full BCI stack. Their EEGle framework, which the team is releasing to the community for building and verifying device security, surfaced 17 new neuro-specific attack instances and a stealth-versus-effectiveness trade-off in backdoor design. The authors warn that generative AI is lowering the barrier for non-expert attackers, with risks to cognitive autonomy, mental privacy and physical safety, from neural data exfiltration to malicious control of BCI-connected devices. The preprint, by Zahra Tarkhani, Georgios Akkogiounoglou, Lorena Qendro, Isabel Tscherniak and Anil Madhavapeddy, has not been peer reviewed.

Even when the eyes stay fixed on one point, the brain tracks moving objects in a video, and that tracking grows stronger with attention. But researchers at KU Leuven found that EEG tracking of an object's motion weakens the farther the object sits from the fixation point, meaning decoding methods that read attention from tracking strength may mistake where an object is for where attention is.

Sleep spindles, brief bursts of brain rhythm during deep sleep linked to memory consolidation, further lowered the excitability of the motor pathway beyond the drop already seen in spindle-free N2/N3 sleep. Researchers timed magnetic pulses to the motor cortex using live EEG readings. Contrary to their hypothesis, a measure of local GABA-A inhibition did not change with spindles or their phase. The study is a preprint and has not been peer reviewed.

A team at the Institute of Biomedical Engineering, Chinese Academy of Medical Sciences and Peking Union Medical College, has built a hybrid brain-computer interface (BCI) that combines surface electromyography (sEMG) with steady-state visual evoked potentials (SSVEP), encoding 480 targets with 120 flicker frequencies and four hand gestures. In online experiments it achieved a mean classification accuracy of 84.55 ± 7.23% and a mean information transfer rate (ITR), a key measure of practical BCI performance, of 260.07 ± 30.41 bits/min. Its command set matches the largest target counts in existing systems, and the authors place its performance among the top, offering a technical reference for large-command-set BCIs. The work appeared in Cognitive Neurodynamics on September 7, 2026.

To move from lab prototypes to long-term clinical systems, implantable brain-computer interfaces must place thousands to millions of electrodes several millimeters to centimeters deep in brain tissue without heating it by more than about 1 degree Celsius. This review benchmarks inductive, mid-field, RF, ultrasonic, magnetoelectric, optical, UWB and electro-quasistatic wireless links against three clinical axes (depth, size and data rate), noting that almost every clinically relevant implant is weakly coupled, with coupling coefficients of only 10^-3 to 10^-1. Within a communication budget of about 10 mW, narrowband high-Q links suit power transfer and low-speed data, but at 1-10 nJ/b they cannot deliver the more than 10 Mbps to tens of Gbps uplinks that interfaces with a thousand or more channels require. The study is a preprint and has not been peer reviewed.

A transformer neural network trained on single-neuron spikes, recorded with intracranial microelectrodes while participants watched an audiovisual episode once, predicted which specific people or places they would name during later free recall. The brain regions carrying that information shifted across sleep: medial temporal lobe signals decoded memories before sleep, frontal cortex signals after. The authors say such decoding could point toward brain-computer interfaces for episodic memory disorders. The study is a preprint and has not been peer reviewed.

Researchers have added a dedicated readout neuron that outputs decision confidence to an echo state network, a neuro-inspired computational model suited to time-series data, so that confidence is updated continuously as a decision forms. In a simulated decision task, confidence activity rose as stimuli became stronger and easier to distinguish. In EEG-based motor imagery classification, confidence activity increased with decision accuracy and distinguished correct from incorrect decisions, especially in better-performing participants. The study is a preprint and has not been peer reviewed.

Can people with poor eyesight still steer a brain-computer interface (BCI) just by shifting their attention? Researchers had healthy volunteers covertly attend to colored targets in their peripheral vision under simulated blurred and monocular vision. The impairments changed the brain's electrical responses to the stimuli and shrank some of them, but the BCI decoded the users' choices just as accurately. The results suggest that as long as a target's color can be recognized, a gaze-independent, non-invasive BCI may stay usable for people with visual impairments. The study is a preprint and has not been peer reviewed.

Two participants typed Vietnamese at 4.95 and 3.45 characters per minute using motor imagery EEG in a pilot test of BCI-vSpeller, a speller built around a custom virtual keyboard of more than 160 selectable keys covering characters, tone marks and special symbols. Researchers from VNU University of Engineering and Technology in Hanoi, reported the findings in Technology and Disability. BCI spellers let users enter text by controlling an interface with mental signals, an approach aimed at people with severe speech and mobility impairments.

Speech imagery (SI) brain-computer interface results may be hard to reproduce, a University of Essex team reports. When the researchers re-ran published decoding pipelines, classification accuracy averaged 11.25 percentage points below the original reports, with gaps ranging from 2% to 39%; in a replication analysis, only 36% of SI participants cleared the threshold for statistical significance, against 91% in motor imagery (MI) datasets. Published in the Journal of Neural Engineering on September 4, 2026, the study is the first to assess both the reproducibility and the replicability of SI decoding. Every SI study evaluated was missing methodological details, and the authors conclude that SI's feasibility as a practical BCI paradigm may have been overestimated, a warning for clinical applications and follow-up research that rely on it.

Endovascular EEG recorded approximately 3.7 times the power of concurrent scalp EEG in 5 patients undergoing an intracarotid amobarbital injection, known as the Wada test, and the nearest endovascular-scalp electrode pairs showed consistently higher coupling in every participant, a mean difference of 4.9 percentage points ranging from 1.8% to 7.6% across individuals and most pronounced at separations under 30 mm. Endovascular EEG has emerged as a brain monitoring technique that balances signal fidelity against invasiveness, the authors write, matching subdural recordings in bandwidth and signal-to-noise ratio in animal studies, but its signal properties have been sparsely quantified in people. All signals were preprocessed with artifact rejection and independent component analysis, then assessed with power spectral density, imaginary coherence, phase-locking value and amplitude envelope correlation.

A preprint posted to arXiv on September 3, 2026, sets out the 2026 PNPL competition, built on an extended LibriBrain100 dataset that adds 32 subjects at about 40 minutes each and roughly 80 more hours of within-subject data. The competition runs two tracks: a Deep track for within-subject word classification at scale, and a Broad track for cross-subject generalization that steps subject-specific fine-tuning data down from about 40 minutes to 20 and then 10, a clinically feasible range. Winning 2025 submissions reached F1-macro scores of 95.6% on speech detection and 73.6% on phoneme classification.

A wireless implantable neural interface, the Brain Interchange (BIC), paired with artifact-removal algorithms captured about 82% of the high-frequency oscillations (HFOs) that a clinical-grade amplifier detected across 24-hour intracranial EEG recordings in 10 patients with drug-resistant epilepsy, according to a preprint. Using those HFOs, the system localized the seizure onset zone with 84% sensitivity and 90% specificity, comparable to clinical equipment. The authors present the work as a translational framework for chronic tracking of epileptogenic networks and for future biomarker-guided adaptive neuromodulation.

Researchers reanalyzed 3,373 spelling selections from 47 patients with ALS and found that adjusting how much a P300 brain-computer interface speller trusts its language model on each selection did not beat a tuned fixed weight. P300 spellers choose letters by detecting the P300 response in EEG and often use a language model to predict the next character. Across 22 language models of up to 46.7B parameters, none showed adaptive fusion outperforming the fixed weight under fair comparison. The fixed-fusion output probability also flagged erroneous selections better than purpose-built controllers, with 2.8- to 3.8-fold error enrichment. The study is a preprint and has not been peer reviewed.

Researchers at the Technical Institute of Physics and Chemistry of the Chinese Academy of Sciences and collaborating institutions have developed a neural electrode interface that, they report, significantly extended electrode operational lifetime and improved signal fidelity over six months of in vivo implantation. The material, a benzyloxycarbonyl-substituted poly(ornithine-alt-glycine) coating abbreviated OGCbz, resists biofouling and immunogenic rejection without degrading electrical performance, targeting what the authors call the critical obstacle to electrodes that combine long-term recording with tunable biofunctional control: an immune-mediated foreign body response in which glial scar encapsulates implants like cement. Substituting the Cbz group with other functional moieties preserves those antifouling and biocompatibility properties while adding new ones, and as a proof of concept the peptide sequence IKVAV and the antibody cetuximab gave the interface neuron affinity and tumor cell proliferation inhibition respectively; the six-month result came from the IKVAV-functionalized version.

Spatial-covariance EEG features can flag whether an operator's decision will be correct before the response is committed, and weighting group votes by that signal raised accuracy on contested trials from 57% to 88% as team size grew from 2 to 16, according to a preprint. Twenty-three participants ran a virtual reality target-detection task under high and low cognitive workload, and the gain appeared only in the high-workload condition; under low workload the weighting hurt performance. EEG-based decision-reliability signals are therefore workload-conditional rather than a general-purpose team augmentation tool, the authors write; the preprint has not been peer reviewed.

A preprint from Minyi Wang, Zhenqin Wu and Rihui Li reports that aligning EEG signals to an intermediate CLIP layer rather than the final one raised 200-way image retrieval on the THINGS-EEG dataset to 78.1% mean Top-1 accuracy, and to 86.4% with CSLS. Layer-wise contrastive learning selects that layer, which the authors call the Neural Visibility Optimal Layer (NVOL); the same representation then drives generation, with a conditional diffusion prior reconstructing subject-specific NVOL features and mapping them into CLIP space for Stable Diffusion XL, which the authors say beats single-stage final-layer diffusion on semantic and structural metrics. The paper was posted to arXiv on September 2, 2026, and has not been peer reviewed.

Researchers have proposed Open-Vocabulary Mutual Information (OVMI), an information-theoretic metric that scores speech brain-computer interfaces on a common scale, and used it to show that accuracy figures computed only over a system's supported vocabulary can overstate how much of a user's intended speech actually gets through. Speech BCIs translate neural activity into language and offer a path to restoring communication for people with paralysis, but systems differ in datasets, recording methods and vocabularies, leaving their reported scores hard to compare. Choosing a vocabulary that maximizes OVMI yielded up to 16.3% relative accuracy improvement across three speech domains; the preprint has not been peer reviewed.

Researchers at Guangzhou Maritime University, Guangdong University of Technology and Guangzhou University in southern China combined a genetic algorithm with Bayesian linear discriminant analysis to pick a fixed, strongly generalizable three-channel subset from a conventional eight-channel P300 BCI, avoiding per-user recalibration. In offline and online virtual reality experiments with 48 healthy subjects, the three-channel system matched the eight-channel system on accuracy and information transfer rate, while equipment preparation fell from 30 minutes to 3, a 90% cut. NASA-TLX scores showed significantly lower mental and physical workload. The study was published in Biomimetics on September 1, 2026.