中 / EN

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

Latest

September 2026
09-24 Papers Foundations & Methods

Researchers Find Compressed EEG Decoders Are No Harder to Fool

Researchers tested whether shrinking brain-computer interface (BCI) decoders through pruning or low-precision quantization, common steps for running them on low-power devices, changes how they hold up against adversarial attacks, in which imperceptible perturbations to EEG signals trick a model into misclassifying. Compressing a model while preserving accuracy did not make it more robust: at a perturbation strength of ε=0.005, EEGNet accuracy fell to 22% to 24% in its original, 50% pruned and two INT8 quantized versions alike. Attacks transferred markedly less well between pruned and original models, while quantization left transfer almost unchanged. The study is a preprint and has not been peer reviewed.
Why it matters As BCIs move to portable hardware, model compression is close to unavoidable, and this work warns that unchanged accuracy is not a proxy for security: robustness, attack transfer and deployment efficiency need separate evaluation. The conclusions rest on motor imagery data from 1 public dataset with 9 subjects.
View details →
09-23 Papers Foundations & Methods

Researchers' TCLA Method Keeps Implanted BCI Decoders Stable Across Sessions

Canyang Zhao and four co-authors posted a preprint on arXiv on September 23, 2026, proposing Task-Conditioned Latent Alignment (TCLA) to tackle a long-standing problem with implanted brain-computer interfaces: the neurons an electrode records change from day to day, so a decoder trained yesterday may fail today. TCLA combines neural signal reconstruction with continuous behavioral supervision to learn a low-dimensional, behavior-related shared space from one recording, then holds that space fixed and aligns each new session into it separately by task. Across seven monkey datasets, long-term cross-session tests averaged R² 0.476±0.014 with a 6.8% decoding failure rate; 1,356 within-subject session pairs averaged R² 0.371±0.009; across different animals, R² averaged 0.218±0.004 with a 12.9% failure rate. Source code is public, and the work has not been peer reviewed.
Why it matters Frequent recalibration is one of the barriers keeping implanted BCIs out of daily use; the failure rate nearly doubles when TCLA is applied across individuals (2,134 cross-subject session pairs), and every result comes from monkeys, with no validation yet in human implant recipients.
View details →
09-23 Papers Foundations & Methods

Spanish Team Finds Immersive VR Strengthens Brain Motor Response in Motor Imagery

Spanish researchers asked 15 healthy participants to imagine raising an arm while watching a humanoid robot do it, viewed in immersive virtual reality, on 2D video or live in person, and recorded their brain activity with EEG. Immersive VR produced the strongest activation of the sensorimotor cortex, measured as desynchronization, a weakening of EEG rhythms that indicates a more active region. The approach of engaging motor areas by watching and imagining, without moving, underlies rehabilitation for people with limited or lost mobility, and the researchers say immersive VR could help optimize such programs and brain-computer interface applications. The study included only healthy people and has not been tested in patients.
Why it matters Motor imagery is one of the most widely used training paradigms in non-invasive BCIs and stroke rehabilitation, yet what to show patients had lacked a direct EEG comparison; this within-subject test points to immersive VR, though a sample this small leaves its value for patients unproven.
View details →
09-23 · University of Essex United Kingdom Papers Movement & Control

Lanzhou Jiaotong University Graph Model Beats EEGNet at Decoding VR Motor Imagery

Researchers at Lanzhou Jiaotong University proposed a hierarchical graph learning framework to decode motor imagery EEG recorded while people watch a movement in virtual reality and imagine performing it. The method splits scalp EEG into rhythmic components, selects key features, represents each trial's brain-region connectivity as a graph and classifies it with a hierarchical graph convolutional network. On the team's self-collected VR motor imagery dataset, it outperformed common baselines including EEGNet in three-class decoding. The team also found that observing plus imagining in VR strengthened connectivity among sensorimotor regions, engaging a wider brain network. The abstract does not give specific accuracy figures or the number of participants.
Why it matters VR is often used to give visual guidance in rehabilitation, but EEG in dynamic scenes is noisy and hard to decode; this work offers both a better decoder and a neural-level explanation of how VR enhances sensorimotor connectivity, though testing on a single in-house dataset limits how far the results generalize.
View details →
09-23 Papers Rehabilitation & Assistance

Review of 32 Studies Finds EEG-Based BCIs Improve Motor Function After Stroke

A systematic review of 32 studies asks whether stroke survivors can retrain paralyzed limbs by imagining movement. Most approaches had patients imagine an action, read the signal from scalp EEG, and used it to drive virtual reality, functional electrical stimulation or robotic exoskeletons, closing the feedback loop. Training ranged from small pilots to as many as 80 sessions, averaging about 30. The authors conclude such methods can significantly improve clinical outcomes in stroke survivors, though the benefit depends on the specific modality and stroke characteristics, and there is still no agreed treatment dose.
Why it matters By putting training parameters side by side, the review exposes how narrow the evidence base is: most studies used 20- or 32-channel EEG and enrolled mainly patients with chronic ischemic stroke, which is why the authors explicitly call for standardized protocols and for including underrepresented patient groups.
View details →
09-23 Clinical Launch & Registration

Registry Marks BCI Stroke Study as Completed

A study investigating brain-computer interface training combined with a lower-limb exoskeleton for walking difficulties after stroke is listed as completed. It compares conventional rehabilitation alone, rehabilitation plus exoskeleton training, and rehabilitation plus combined BCI and exoskeleton training. Each group has a target of 20 participants, not a confirmed actual enrollment figure. The primary assessment concerns lower-limb motor function; other measures cover EEG, balance, and walking ability. First enrollment was dated 2024-03-22, before registration on 2026-09-23, making registration non-prospective. The registry contains no results, so it does not establish whether combined training outperforms the other interventions.
Why it matters Eligibility requires a first stroke, disease duration of less than 6 months, and a Functional Ambulation Category score of 0–2. These population limits will matter when interpreting any subsequent results, rather than generalizing them to everyone with walking difficulties after stroke.
View details →
09-23 Papers Foundations & Methods

Purdue Finds More Simultaneous Stimulation Speeds Sensory Learning in Eight Mice

Purdue University researchers compared sparse and redundant intracortical microstimulation encoding in eight mice. Stimulating more electrodes simultaneously accelerated learning, shortened navigation trials and improved success rates and path efficiency when stimulation was combined with vision. The findings were released as a preprint.
Why it matters Comparing learning and navigation after changing the number of active electrodes tests how sensory encoding affects the training burden.
View details →
09-22 Papers Foundations & Methods

Paper outlines clinical evidence framework for BCI

As brain-computer interfaces move toward clinical use, regulators, clinicians and developers need agreed rules on what counts as sufficient evidence. A new paper addresses how a clinical evidence framework for BCI can be built and applied. Only bibliographic metadata is publicly available so far, so the framework's components and intended uses cannot yet be verified.
Why it matters How much evidence a BCI needs before it is considered clinically proven remains unsettled, and this paper tackles that question directly. Only the title and authors are public so far; the framework's tiers and metrics cannot be assessed until the full text is available.
View details →
09-22 Papers Foundations & Methods

New Ranking Method Moves Motor Imagery BCI's 'Golden Subject' From S5 to S1

In cross-subject transfer learning for motor imagery BCIs, picking the right 'golden subject', the person whose EEG best serves as a training template for others, directly affects decoding accuracy for other users. An earlier study using the fuzzy decision method FDOSM named S5, but that method did not separate weighting from ranking. Researchers at Imam Ja'afar Al-Sadiq University in Iraq built a unified 9 x 18 golden-subject decision matrix covering nine subjects on 18 training- and evaluation-phase metrics, including accuracy, recall, F1 score, log loss and specificity. Using FWZIC for weighting and VIKOR for compromise ranking, they found the best candidate shifted to S1 (Q = 0.0000), which ranked first in three of four weighting scenarios in a sensitivity analysis. The work is an early methodological exploration involving only nine subjects and has not been validated on larger datasets.
Why it matters Cross-subject transfer in motor imagery BCIs leans heavily on the choice of golden subject, yet there has been no transparent framework that handles weighting and ranking separately. Switching methods reversed the result from S5 to S1, showing how much the decision method shapes the conclusion and suggesting earlier single-method findings need cross-checking. With only nine subjects, larger datasets are needed.
View details →
09-22 Industry News Therapy & Modulation

Epilepsy Neuromodulation Moves Toward Closed-Loop Sensing and Less Invasive Implants

About one-third of people with epilepsy keep having seizures despite medication, which remains first-line treatment. Over two decades, neuromodulation has grown from vagus nerve stimulation alone into several technical routes, and the FDA has approved three implantable therapies: LivaNova's VNS (1997), NeuroPace's RNS System (2013) and Medtronic's deep brain stimulation of the anterior thalamic nucleus (2018). Only RNS reads brain signals, continuously monitoring intracranial EEG and stimulating when it detects abnormal patterns. Newer devices are trying to add sensing and closed-loop algorithms while avoiding craniotomy, placing electrodes under the scalp or outside the skull. Most systems from companies including PRECISIS, Adraxe, Synergia Medical and iVEAcare remain in early studies with small published samples, and PRECISIS's planned pivotal trial of about 200 patients has yet to report.
Why it matters Among approved epilepsy implants, only RNS actually reads brain activity; DBS runs on programmed settings and VNS AutoStim is triggered by heart rate. Newcomers are pushing sensing and closed-loop control into less invasive hardware, but Adraxe and iVEAcare are still in early studies and PRECISIS's pivotal trial of about 200 patients has yet to report.
View details →

© 2026 BCIwiki.com Digest Topics Tips Subscribe Revisions About