Neural Noise

2026 week 22

30 min read

BCI Weekly Brief (week of 2026-05-25)

Foundation-model papers dominate this week: montage-agnostic EEG event segmentation and inner-speech decoding reviews sit alongside invasive speech-BCI articulatory generalization and an online EEG BCI for AR. Several clinical-neurophysiology and neuromodulation papers round out the feed. This batch is Clinical Neurophysiology clinical electrophysiology with no direct BCI device or company news. Highest signal: continuous resting-state EEG for coma prognosis, IFCN AI-in-practice guidance, and commercial-system TMS waveform comparisons; neuromodulation protocol debates and SEEG signal processing follow.

This week we selected 33 items from a larger pool of 40 candidates.

Adapting frozen foundation models for montage-agnostic high-resolution EEG event segmentation

Journal of Neural Engineering

Published: 2026-05-25T23:00:00+00:00

Tags: EEG, foundation-models, methods, tier-1

Frozen EEG foundation models adapted for montage-agnostic, high-resolution event segmentation address a core deployment blocker for ambulatory BCIs. Peer-reviewed JNE evidence that SSL representations transfer across hardware without full retraining.

  • A Journal of Neural Engineering study tests whether frozen EEG foundation models can be adapted for high temporal resolution event segmentation on continuous EEG.
  • The work targets ambulatory BCIs that must detect neural events outside controlled laboratories without relying on time-locked synchronization.
  • A core requirement is generalizing across diverse electrode montages from different EEG acquisition hardware.
  • The method keeps foundation-model weights frozen and adapts them for montage-agnostic, high-resolution event segmentation rather than retraining end to end.
  • Performance is evaluated on unseen montages and datasets to stress-test cross-hardware transfer.
  • Self-supervised learning representations are reported to transfer across acquisition setups without full model retraining.
  • High-resolution event segmentation on continuous streams is framed as a key deployment blocker for real-world, non-laboratory BCIs.
  • Peer-reviewed JNE evidence supports using pretrained EEG foundation models as a reusable backbone for downstream segmentation adapters.

Across-speaker articulatory reconstruction from sensorimotor cortex for generalizable brain–computer interfaces

Journal of Neural Engineering

Published: 2026-05-25T23:00:00+00:00

Tags: speech-prosthesis, ECoG, BCI, tier-1

Cross-speaker articulatory decoding from sensorimotor cortex could let speech BCIs train without paired vocalization from locked-in users. Directly targets generalization for implantable communication interfaces. Strong JNE implementation path.

  • The study is published in Journal of Neural Engineering (DOI 10.1088/1741-2552/ae4cca).
  • It reconstructs articulatory features from sensorimotor cortex to support generalizable speech brain–computer interfaces.
  • Accurate articulatory reconstruction from brain activity is proposed as a route to communication BCIs for people with vocal tract paralysis.
  • Conventional speech-BCI training has relied on paired neural–articulatory or neural–audio recordings from the same participant.
  • That paired-data requirement makes it hard to train reconstruction models in users who cannot articulate or vocalize.
  • The authors propose a novel approach for extracting articulatory information without the usual per-user paired vocalization setup.
  • Decoding is designed to work across speakers using sensorimotor cortex signals rather than speaker-specific paired calibration alone.
  • Cross-speaker articulatory decoding is aimed at implantable communication BCIs that could train without matched vocalization from locked-in users.

The current status of foundation models in decoding inner speech from non-invasive brain signals: a mini review

Frontiers in Human Neuroscience

Published: 2026-05-28T00:00:00+00:00

Tags: EEG, speech-prosthesis, methods, tier-1

Timely synthesis of FM/SSL approaches for imagined-speech BCIs via EEG and related modalities. Maps datasets, architectures, and open gaps for silent communication systems. Review-level but decision-useful for R&D prioritization.

  • A 2026 Frontiers in Human Neuroscience mini review (10.3389/fnhum.2026.1838064) surveys how foundation models decode inner speech from non-invasive brain signals, including EEG and related modalities.
  • Inner speech (IS) is imagined speech without overt articulation and is framed as a promising brain–computer interface target for restoring communication.
  • The review emphasizes BCIs aimed at people with severe speech impairments, such as locked-in syndrome.
  • Foundation models are typically trained with self-supervised learning (SSL) on large-scale datasets to learn transferable, robust representations from neural recordings.
  • The authors position FM/SSL pretraining as a way to improve generalization when building decoders from limited or heterogeneous non-invasive neural data.
  • The paper is structured as an overview of FM-based approaches for inner-speech decoding rather than a single new empirical trial.
  • It maps public datasets, model architectures, and open gaps relevant to silent, imagined-speech communication systems.
  • As a review-level synthesis, it is pitched as decision-useful for prioritizing R&D on imagined-speech BCIs.

An online brain-computer interface for detecting incongruity in augmented reality applications

Journal of Neural Engineering

Published: 2026-05-28T23:00:00+00:00

Tags: EEG, BCI, methods, tier-1

Hybrid online BCI detects when AR overlays disagree with user expectations—an underexplored UX failure mode. Demonstrates real-time EEG decoding tied to adaptive interfaces. JNE peer review supports near-term prototyping.

  • Augmented reality can overlay digital information about physical objects in their real-world context.
  • That information may conflict with user expectations because of factual errors in the data or cognitive biases.
  • When physical objects and digital overlays disagree, user experience suffers and trust in the AR system can erode.
  • The authors propose detecting these physical–digital inconsistencies with a hybrid brain–computer interface.
  • The system is built for online (real-time) operation rather than offline post-hoc analysis.
  • It uses EEG decoding linked to adaptive AR interfaces to flag incongruity as users interact.
  • The work appears in the peer-reviewed Journal of Neural Engineering.
  • Mismatch between AR overlays and user expectations is an underexplored failure mode in AR UX design.

Feasibility of a hybrid SSVEP-motor imagery BCI with robotic feedback for upper limb motor rehabilitation in stroke patients

J. Neuroscience Methods

Tags: EEG, BCI, clinical, tier-1

Combines SSVEP and motor-imagery decoding with robotic feedback in stroke rehab—a concrete closed-loop rehabilitation BCI design. Feasibility framing signals translational intent. Worth tracking for hybrid paradigm adoption.

  • Researchers led by Xiaohu Pan report a feasibility study of a hybrid brain–computer interface that combines steady-state visual evoked potential (SSVEP) and motor-imagery decoding for upper-limb motor rehabilitation after stroke.
  • The system pairs EEG-based intent decoding with robotic feedback in a closed-loop rehabilitation design aimed at stroke patients.
  • The work is scheduled for publication in August 2026 in the Journal of Neuroscience Methods, Volume 432.
  • Authors include Xiaohu Pan, Ruoqing Zhang, Xiaoqian Xia, Hongyan Cui, Lixu Liu, and Xiaogang Chen.
  • The study targets upper-limb motor recovery rather than communication or device control alone.
  • SSVEP and motor imagery are integrated as complementary non-invasive EEG control signals within one rehabilitation BCI.
  • Robotic feedback is used to translate decoded brain signals into physical movement assistance during rehab.
  • The paper is framed around feasibility, signaling an early translational step toward clinical stroke-rehabilitation BCIs.

Interpretable side-aware kinematic-sEMG gait-state representations relevant to adaptive neurorobotic assistance after stroke: a public-dataset study

Frontiers in Neurorobotics

Published: 2026-05-25T00:00:00+00:00

Tags: neurorobotics, sEMG, methods, tier-2

Transparent side-aware kinematic-sEMG gait states aim to replace opaque embeddings in post-stroke exoskeleton control. Public-dataset secondary analysis lowers replication barrier. Relevant to adaptive neurorobotics stack design.

  • A Frontiers in Neurorobotics study proposes interpretable, side-aware gait states that combine kinematics and surface EMG for adaptive lower-limb assistance after stroke.
  • The authors argue that post-stroke exoskeleton control needs gait-state features that preserve locomotor structure instead of reducing walking to a single asymmetry score or an opaque latent embedding.
  • Because post-stroke gait is multimodal and side dependent, they motivate transparent per-side representations over modality-isolated summaries for future adaptive-assistance design.
  • The work is a secondary analysis of a public multimodal gait dataset, chosen to lower the replication barrier for neurorobotics research.
  • The dataset excerpt reported in the available text includes 138 able-bodied adults plus additional participants whose full cohort count was truncated in the source.
  • Kinematic and sEMG signals are fused into gait-state representations intended as readable control inputs rather than black-box embeddings.
  • The study frames these representations as relevant to the design of adaptive neurorobotic stacks for stroke rehabilitation.

Emotion recognition based on the temporal patterns of electroencephalogram signals and electrodermal response signals using the TRANSFORMER network

Frontiers in Neuroscience

Published: 2026-05-29T00:00:00+00:00

Tags: EEG, methods, tier-2

Transformer multimodal fusion of EEG and EDA for subject-independent emotion decoding. Methods paper with implications for affective BCIs and neuroadaptive interfaces. Moderate confidence pending cross-lab validation.

  • Published in Frontiers in Neuroscience, the study targets emotion recognition from physiological signals using a Transformer network applied to temporal EEG and electrodermal response (EDA) patterns.
  • The authors propose a Transformer-based multimodal framework for four-class discrete emotion recognition, with neutral among the labeled categories in the study design.
  • The model jointly analyzes central nervous system activity (EEG) and autonomic arousal (EDA) rather than relying on either modality alone.
  • Existing emotion-recognition methods are limited in modeling long-range temporal dependencies across physiological time series.
  • The framework is designed to represent central–autonomic coupling—the coordinated relationship between brain signals and peripheral autonomic responses during emotion.
  • The work evaluates decoding under a subject-independent protocol, testing generalization to people not included in model training.
  • Prior approaches still need better generalization when applied across unseen individuals in subject-independent settings.
  • Physiological emotion recognition is framed as relevant to affective neuroscience and human-centered artificial intelligence.
  • As a methods paper, the approach is positioned for downstream use in affective brain–computer interfaces and neuroadaptive systems that respond to user emotional state.
  • Replication and performance claims will need confirmation in independent laboratories before broad adoption.

Resting-state EEG for continuous prognostic monitoring and prediction of coma recovery after acute brain injury

Clinical Neurophysiology

Tags: EEG, clinical, methods, tier-1

Resting-state EEG for continuous coma-recovery prognosis targets ICU neural monitoring pipelines adjacent to wearable and bedside BCIs. ML on longitudinal EEG time series with acute-injury cohorts signals near-term deployment path. Peer-reviewed Clin Neurophysiol.

  • Clinical Neurophysiology (August 2026, Volume 188) reports on resting-state EEG for continuous prognostic monitoring and prediction of coma recovery after acute brain injury.
  • Authors include Morteza Zabihi, Sophie E. Ack, Daniel B. Rubin, Emily J. Gilmore, Valdery Moura Junior, Quanzheng Li, Michael J. Young, Brian L. Edlow, Yelena G. Bodien, and Eric S. Rosenthal.
  • The study focuses on resting-state EEG as a tool for ongoing prognosis rather than one-off assessments in critically ill patients with acute brain injury.
  • Machine learning is applied to longitudinal resting-state EEG time series drawn from acute brain injury cohorts.
  • The approach is aimed at ICU neural monitoring pipelines, including wearable and bedside brain–computer interface setups.
  • Continuous EEG-based coma-recovery prediction is positioned as a near-term path toward deployment in critical care monitoring.
  • The work addresses prognostic monitoring for patients in coma after acute brain injury, a setting where recovery trajectories are difficult to forecast with standard bedside measures alone.

A non-invasive neuromodulation tuned to regional neuronal activity

Clinical Neurophysiology

Tags: neuromodulation, clinical, tier-2

Activity-tuned non-invasive neuromodulation could improve closed-loop stimulation targeting beyond open-loop protocols. Clinical Neurophysiology venue signals electrophysiology grounding. Watch for parameter generalization data.

  • The paper presents a non-invasive neuromodulation approach tuned to regional neuronal activity.
  • It was published in Clinical Neurophysiology, Volume 188, in August 2026.
  • Authors are Franca Tecchio, Giovanni Pellegrino, and Luca Paulon.
  • Activity-tuned non-invasive stimulation is framed as a path toward closed-loop targeting beyond conventional open-loop protocols.
  • The Clinical Neurophysiology venue points to an electrophysiology-grounded study context.
  • Readers should watch for whether stimulation parameters generalize beyond the conditions tested in the paper.

Differential modulation of sensorimotor beta oscillations by beta tACS alone and combined repetitive paired-pulse TMS with tACS

NeuroImage

Tags: neuromodulation, EEG, tier-2

Compares beta-band tACS alone vs combined rTMS-tACS effects on sensorimotor beta oscillations—relevant to oscillation-informed BCI calibration and neuromodulation stacks. EEG outcome anchors relevance despite neuroimaging venue.

  • Published in NeuroImage, Volume 335, on 15 July 2026.
  • Authors Hisato Nakazono, Akinori Takeda, Emi Yamada, Tsubasa Mitsutake, Takanori Taniguchi, Daiki Matsuda, and Katsuya Ogata compared two neuromodulation protocols targeting sensorimotor beta oscillations.
  • One condition applied beta-band transcranial alternating current stimulation (tACS) alone.
  • The other combined repetitive paired-pulse transcranial magnetic stimulation (rTMS) with beta-band tACS.
  • The study focused on how each protocol differentially modulated sensorimotor beta oscillations.
  • EEG was the outcome measure, anchoring the work to oscillation-based brain–computer interface calibration.
  • The design tests whether stacking rTMS with tACS changes beta dynamics beyond tACS by itself.
  • The article appears in a neuroimaging journal but reports electrophysiological endpoints rather than structural or functional imaging alone.

Translating artificial intelligence into clinical practice: statements of the International Federation of Clinical Neurophysiology

Clinical Neurophysiology

Tags: clinical-neurophysiology, methods, tier-1

IFCN consensus on deploying AI in clinical neurophysiology sets validation and workflow expectations for EEG-based decoders and monitoring tools. Useful regulatory and R&D framing for BCI teams shipping ML on neural time series. Authoritative specialty-body guidance.

  • The International Federation of Clinical Neurophysiology published consensus statements on translating artificial intelligence into routine clinical neurophysiology practice.
  • The article appeared online in Clinical Neurophysiology on 14 May 2026.
  • Antonio Suppa and Alessandro Zampogna are listed as the authors.
  • The IFCN statements define validation requirements that AI tools must meet before entering electrophysiology workflows.
  • They set workflow expectations for integrating AI into EEG-based decoders and neural monitoring applications.
  • The guidance is intended as federation-level specialty-body consensus rather than single-site or vendor-specific advice.
  • Teams developing machine learning on neural time series can use the document to frame regulatory submissions and R&D planning.

Distinct electrophysiological profiles of bacterial mechanosensitive channels for sonogenetic actuator selection

Journal of Neural Engineering

Published: 2026-05-28T23:00:00+00:00

Tags: neuromodulation, methods, tier-2

In vivo benchmarking of bacterial mechanosensitive channels under matched ultrasound stimulation informs sonogenetic actuator selection. Exploratory rat data but directly supports next-gen cell-targeted neuromodulation interfaces.

  • Sonogenetics pairs ultrasound stimulation with genetically encoded mechanosensitive ion channels to enable cell-targeted neuromodulation.
  • Actuator choice in sonogenetics remains largely empirical because in vivo electrophysiological response signatures are rarely compared under matched stimulation conditions.
  • Researchers conducted an exploratory in vivo benchmark of three bacterial mechanosensitive channels—MscL-G22S, MscL-G22N, and MscS—under matched transcranial ultrasound in anesthetized rats.
  • The channels were tested during transcranial ultrasound stimulation in anesthetized rat primary visual cortex.
  • The study title frames the main finding as distinct electrophysiological profiles across the three bacterial mechanosensitive channels.
  • Matching ultrasound parameters across conditions lets in vivo electrophysiology guide sonogenetic actuator selection rather than ad hoc channel choice.
  • The work was published in the Journal of Neural Engineering.
  • Although exploratory, the rat in vivo data is intended to support next-generation cell-targeted neuromodulation interfaces.

The comparison of biphasic and monophasic waveforms through commercially available systems

Clinical Neurophysiology

Tags: neuromodulation, methods, tier-1

Head-to-head comparison of biphasic vs monophasic TMS waveforms on commercial stimulators directly affects neuromodulation protocol reproducibility. Hardware-specific dose differences matter for closed-loop tACS/tRNS stacks. Empirical Clin Neurophysiol data.

  • Clinical Neurophysiology published this study online on 25 May 2026.
  • Authors are K.R. Ramdeo, S.D. Foglia, M.A. Sader, F.C. Adams, R.S. Rehsi, C.C. Drapeau, and A.J. Nelson.
  • The paper compares biphasic and monophasic TMS waveforms delivered through commercially available stimulator systems.
  • It is a head-to-head hardware comparison rather than a single-device or purely theoretical waveform analysis.
  • Waveform choice on commercial TMS units can change effective dosing, which affects whether stimulation protocols reproduce across labs and devices.
  • Hardware-specific dose differences between biphasic and monophasic outputs matter when translating parameters between stimulators or study sites.
  • The findings are relevant to closed-loop neuromodulation setups that stack transcranial alternating or random-noise stimulation with other stimulation modalities.
  • The study provides empirical Clinical Neurophysiology data on waveform effects in clinically used TMS hardware.

Transcranial ultrasound stimulation for neuromodulation in movement disorders

Clinical Neurophysiology

Tags: neuromodulation, clinical, tier-2

Clinical-neurophysiology review of transcranial ultrasound for movement-disorder neuromodulation. Useful for mapping TUS evidence vs tDCS/rTMS in motor populations. Review tier limits immediate implementation signal.

  • Published August 2026 in Clinical Neurophysiology, Volume 188.
  • Review title: Transcranial ultrasound stimulation for neuromodulation in movement disorders.
  • Authors: Carly Pellow, Camila Aquino, Fady Girgis, Zelma Kiss, Davide Martino, Samuel Pichardo, and G. Bruce Pike.
  • Synthesizes transcranial ultrasound stimulation (TUS) evidence for neuromodulation in movement-disorder populations.
  • Frames TUS within clinical neurophysiology as a non-invasive stimulation modality for motor disorders.
  • Useful for comparing TUS evidence against tDCS and rTMS in motor populations.
  • Published in Clinical Neurophysiology, a venue focused on electrophysiology and clinical neuromodulation.
  • Article type is a narrative review rather than a primary clinical trial.
  • Review-tier evidence limits immediate clinical implementation signal from this paper alone.
  • Available at ScienceDirect (PII S1388245726004141).

Confounding factors in the comparison of iTBS protocols − Comments on Ramdeo et al.: “Waveform matters”

Clinical Neurophysiology

Tags: neuromodulation, methods, tier-1

Debate on iTBS protocol confounds clarifies which waveform and coil variables actually drive outcomes—critical for reproducible transcranial stimulation in adaptive interfaces. Paulus co-authorship adds methodological weight. Informs near-term protocol design.

  • Tao Han, Yehong Chen, Walter Paulus, and Ken Möhwald published a Clinical Neurophysiology commentary online on 19 May 2026.
  • The article is titled “Confounding factors in the comparison of iTBS protocols” and comments on Ramdeo et al.’s “Waveform matters.”
  • The piece focuses on confounding factors that can distort comparisons between intermittent theta-burst stimulation (iTBS) protocols.
  • The authors weigh which waveform parameters versus coil variables actually drive stimulation outcomes.
  • Resolving those confounds is presented as important for reproducible transcranial stimulation research and practice.
  • Walter Paulus’s co-authorship is noted as adding methodological weight to the protocol-comparison debate.
  • The discussion is framed as relevant to near-term iTBS protocol design, including work on adaptive brain–computer interfaces.

Propagation concordance index reflects spatial proximity to the spike generator by linking interictal spikes with stimulation-induced neural propagation

Clinical Neurophysiology

Tags: iEEG, clinical, tier-2

Propagation concordance index links interictal spikes to stimulation-evoked propagation for localizing epileptic generators. Method may transfer to intracranial BCI electrode planning and closed-loop stimulation targeting.

  • Researchers introduce a propagation concordance index (PCI) that links interictal spikes with stimulation-induced neural propagation.
  • PCI is designed to reflect spatial proximity to the epileptic spike generator.
  • The method compares spontaneous interictal spike activity with neural spread evoked by stimulation to help localize seizure foci.
  • The study is published in Clinical Neurophysiology, Volume 189, with a September 2026 publication date.
  • Authors include Takumi Mitsuhashi, Hiroharu Suzuki, Kazuki Nishioka, Kazuki Nomura, Tetsuya Ueda, Yuki Takaki, Madoka Nakajima, Hidenori Sugano, Yasushi Iimura, and Akihide Kondo.
  • The approach may inform planning of intracranial BCI electrode placements.
  • Potential applications also include closed-loop stimulation targeting for epilepsy.

Correction of connectivity induced by autocorrelation and filtering in resting state functional near-infrared spectroscopy data

J. Neuroscience Methods

Tags: fNIRS, methods, tier-2

Corrects spurious fNIRS connectivity from autocorrelation and filtering—important preprocessing for wearable BCIs using hemodynamic signals. Methods-focused but directly affects decoder validity in fNIRS pipelines.

  • Published in Journal of Neuroscience Methods, Volume 433, with a September 2026 publication date.
  • Authors are Mengmeng Wang, Ishara Paranawithana, Darren Mao, Yan T. Wong, Colette M. McKay, Leigh A. Johnston, and Catherine E. Davey.
  • The paper focuses on resting-state functional near-infrared spectroscopy (fNIRS) functional connectivity.
  • It addresses connectivity estimates that can be distorted by signal autocorrelation and standard filtering steps in fNIRS preprocessing.
  • The work presents methods to correct those artifacts so reported brain-network links are not spuriously inflated.
  • Autocorrelation and filtering are identified as key sources of artificial connectivity in resting-state fNIRS analyses.
  • The contribution is methods-oriented, aimed at improving the validity of hemodynamic connectivity metrics used in neuroimaging pipelines.
  • Findings are relevant to teams building wearable brain–computer interfaces that rely on fNIRS hemodynamic signals for decoding or network analysis.

‘No-No’ head movement as a true epileptic phenomenon – A case series with SEEG and signal processing evaluation

Clinical Neurophysiology

Tags: iEEG, clinical, tier-2

SEEG case series with explicit signal-processing evaluation advances intracranial electrophysiology methods shared with implantable recording platforms. Indirect BCI relevance via iEEG decoding and artifact handling. Clinical n limits generalization.

  • Published in Clinical Neurophysiology, Volume 188 (August 2026), the paper by Futoon Alotaibi, Ayman A. AbdelHamid, and colleagues reports a case series on rhythmic side-to-side “No-No” head movement as a genuine epileptic phenomenon.
  • The study design is a clinical case series that uses stereoelectroencephalography (SEEG) to evaluate whether No-No head movements are ictal rather than non-epileptic or artifactual.
  • Each case was assessed with explicit signal-processing analysis to relate head-movement events to concurrent intracranial electrophysiology.
  • SEEG provides the intracranial recording ground truth needed to distinguish epileptic head movement from movement-related recording artifacts.
  • The signal-processing evaluation is framed as a methods advance for intracranial electrophysiology workflows used with implantable recording platforms.
  • Results have indirect relevance to invasive EEG (iEEG) decoding and to handling movement-related artifacts in chronic implant recordings.
  • Because this is a small clinical case series, the reported findings have limited generalizability beyond the included patients.
  • The author team includes Jackie Y. Ying, Ibrahim Althubaiti, and Sasha Dionisio alongside the Saudi epilepsy-surgery collaborators listed in the publication.

Neural correlates and network dynamics of uncommon semiologies in the cingulo-insulo-opercular network explored with Stereo-EEG

Clinical Neurophysiology

Tags: iEEG, clinical, tier-2

Stereo-EEG network analysis of rare seizure semiologies advances intracranial mapping methods shared with implantable recording platforms. Indirect BCI relevance via sEEG signal-processing and network decoding.

  • AbdelHamid et al. published in Clinical Neurophysiology, Volume 189 (September 2026), on neural correlates and network dynamics of uncommon seizure semiologies using Stereo-EEG.
  • The study focuses on the cingulo-insulo-opercular network as the anatomical system of interest.
  • Stereo-EEG network analysis is used to characterize rare or atypical seizure semiologies.
  • Authors include Ayman A. AbdelHamid, Mohammed AlShammri, Mohammad Alkouli, Faisal Alotaibi, Ibrahim Althubaiti, Jackie Y. Ying, and Sasha Dionisio.
  • The methods extend intracranial mapping approaches used with implantable recording platforms.
  • sEEG signal-processing and network decoding in the work have indirect relevance to brain-computer interface research.
  • The paper links uncommon semiology phenotypes to network-level dynamics rather than isolated focal activity.

Pallidal physiology and predictors of successful deep brain stimulation in jerky dystonia, dystonia with tremor, and their combination

Clinical Neurophysiology

Tags: neuromodulation, clinical, tier-2

Pallidal electrophysiology predicting DBS success couples invasive recording with closed-loop stimulation targeting—parallel skill set to adaptive neuroprosthetics. Dystonia cohort specificity lowers immediate BCI transfer. Solid invasive-interface evidence.

  • Published August 2026 in Clinical Neurophysiology, Volume 188.
  • Authors include Indiko Dzhalagoniya, Ulia Semenova, Anna Gamaleya, Alexey Tomskiy, Hyder A. Jinnah, Alexey Sedov, and Aasef G. Shaikh.
  • The study links pallidal electrophysiology to predictors of successful deep brain stimulation in movement-disorder dystonia.
  • Three phenotypes are compared: jerky dystonia, dystonia with tremor, and patients with both features combined.
  • The work uses invasive pallidal recording alongside stimulation targeting rather than non-invasive signals alone.
  • Electrophysiological markers from the globus pallidus are evaluated as preoperative or intraoperative guides to DBS benefit.
  • Success is framed as meaningful clinical improvement after pallidal DBS, stratified by dystonia subtype.
  • Results are specific to dystonia cohorts, so generalization to other brain–computer interface populations is limited.

Electrophysiological biomarkers of concussion in adolescents measured using magnetoencephalography

Clinical Neurophysiology

Tags: MEG, clinical, tier-2

MEG-derived electrophysiological biomarkers for adolescent concussion extend non-invasive neural time-series phenotyping. Peripheral to core BCI but relevant to high-density electrophysiology decoding methods.

  • A Clinical Neurophysiology study published in August 2026 (Volume 188) examines electrophysiological biomarkers of concussion in adolescents using magnetoencephalography (MEG).
  • Authors include Nikou Kelardashti, Rouzbeh Zamyadi, Jaehyun Sur, Phillip Johnston, Pavreet Gill, Alexa Irvin, Andrea Hickling, Adrienne Davis, George M. Ibrahim, Anne L. Wheeler, Shannon E. Scratch, and Benjamin T. Dunkley.
  • The study uses MEG to measure electrophysiological biomarkers associated with concussion in an adolescent population.
  • The work targets non-invasive neural time-series phenotyping via MEG-derived electrophysiological signals rather than relying on structural imaging alone.
  • Findings are framed as MEG-based electrophysiological biomarkers that may support concussion assessment and decoding in teenagers.
  • The paper appears in Clinical Neurophysiology and is accessible at the ScienceDirect link provided in the source citation.

Cortical changes in amyotrophic lateral sclerosis: comparing biomarkers of glymphatic flow and cortical excitability

Clinical Neurophysiology

Tags: clinical, neuroprosthetics, tier-2

ALS cortical excitability biomarkers matter for implantable motor/speech BCI candidacy and longitudinal interface stability. Glymphatic comparison is peripheral to decoding but frames patient stratification. Relevant to BrainGate-class populations.

  • Clinical Neurophysiology published “Cortical changes in amyotrophic lateral sclerosis: comparing biomarkers of glymphatic flow and cortical excitability” online on 26 May 2026.
  • The paper compares glymphatic-flow biomarkers with cortical excitability biomarkers in the context of cortical changes in amyotrophic lateral sclerosis (ALS).
  • Authors are Nathan Pavey, Sicong Tu, Sheryl Foster, Nimeshan Geevasinga, Aicee Calma, Yukiko Tsuji, Matthew C. Kiernan, Steve Vucic, and Parvathi Menon.
  • The article is indexed as PII S1388245726004499 on ScienceDirect.
  • ALS cortical excitability biomarkers are positioned as relevant to candidacy for implantable motor and speech brain–computer interfaces and to longitudinal interface stability.
  • Glymphatic-flow comparison is framed as secondary to decoding but potentially useful for ALS patient stratification.
  • The ALS population context overlaps invasive motor/speech BCI cohorts such as those studied in BrainGate-class trials.

Multimodal imaging-based targeting approach for network-level brain stimulation

Frontiers in Neuroscience

Published: 2026-05-29T00:00:00+00:00

Tags: neuromodulation, methods, tier-2

Multimodal fMRI framework for tDCS target selection improves network-level stimulation planning. fMRI-heavy and per calibration, but informs closed-loop neuromodulation interface design.

  • Neural network effects of transcranial direct current stimulation (tDCS) remain poorly understood.
  • The authors introduce a prospective, empirically informed multimodal functional MRI (fMRI) framework to guide tDCS target selection.
  • The framework supports hypothesis-based analysis planning for future focal tDCS-fMRI studies.
  • The approach is framed for network-level brain stimulation targeting using multimodal imaging.
  • They illustrate the method with data from 37 healthy participants pooled across two tDCS-fMRI studies.
  • The illustrative sample included 19 females with a mean age of 25.8 years (SD 5.9).
  • The work was published in Frontiers in Neuroscience (2026; article 10.3389/fnins.2026.1803897).

Effects of midazolam and propofol on EEG detection of epileptiform activity in patients with altered consciousness

Clinical Neurophysiology

Tags: EEG, clinical, tier-2

Quantifies how sedatives alter EEG epileptiform detection in altered-consciousness patients—relevant to ICU neural monitoring and any BCI operating under anesthesia or sedation. Clinical EEG focus.

  • Published in Clinical Neurophysiology (September 2026, Volume 189), the study compares how midazolam and propofol affect EEG detection of epileptiform activity in patients with altered consciousness.
  • Authors are Milène Guinchard, Andrea O. Rossetti, Kaspar Schindler, Stephan Rüegg, Vincent Alvarez, and Jan Novy.
  • The work focuses on two widely used sedatives—midazolam and propofol—and their impact on whether epileptiform patterns are seen on EEG.
  • Participants are patients with altered consciousness, a group common in ICU and acute neurology settings where continuous EEG is used.
  • The central question is how sedation choice changes the reliability of EEG-based identification of epileptiform activity, not just raw signal appearance.
  • Findings matter for ICU neural monitoring pipelines that must interpret EEG under routine sedative use.
  • The results also bear on brain–computer interface and other neural decoding systems that may run while patients are sedated or anesthetized.
  • Clinical EEG practice in altered-consciousness care depends on knowing sedative-specific effects on spike-and-wave or other epileptiform signatures.

Beyond motor cortex: A novel relationship between associative parietal cortex and ipsilateral silent period

Clinical Neurophysiology

Tags: clinical-neurophysiology, methods, tier-2

Maps parietal–motor cortical interactions via ipsilateral silent period—useful context for motor BCI channel selection beyond M1. TMS neurophysiology focus without decoding or device data. Moderate translational signal.

  • Published in Clinical Neurophysiology and available online on 30 May 2026.
  • Authored by Costanza Iester, Alice Bellosta, Ludovico Pedullà, Sabrina Brigadoi, Monica Biggio, Ambra Bisio, Simone Cutini, Elena Monteleone, Laura Bonzino, and Marco Bove.
  • Reports a novel relationship between associative parietal cortex and the ipsilateral silent period, extending motor-cortex–centered TMS work beyond primary motor cortex (M1).
  • Uses TMS neurophysiology to characterize how associative parietal cortex relates to ipsilateral silent period measures of cortical inhibition.
  • Maps parietal–motor cortical interactions through ipsilateral silent period rather than relying on M1-only stimulation paradigms.
  • Focuses on cortical physiology and stimulation methodology, with no neural decoding experiments or BCI device data.
  • Suggests parietal cortex may matter for motor interface planning because it couples to the same-side silent period pathway used to probe motor inhibition.
  • Offers moderate translational relevance for teams evaluating non-M1 cortical targets or channels when designing motor BCIs.

Modulating cortical inhibition in Functional Gait Disorder − Neurophysiological evidence from low-frequency rTMS

Clinical Neurophysiology

Tags: neuromodulation, clinical, tier-2

Low-frequency rTMS modulates cortical inhibition in functional gait disorder with paired neurophysiology readouts. Tangential to BCI but documents stimulation-response biomarkers usable in closed-loop rehab systems.

  • Published in August 2026 in Clinical Neurophysiology, Volume 188.
  • The study examines modulating cortical inhibition in functional gait disorder using low-frequency repetitive transcranial magnetic stimulation (rTMS).
  • The title frames the contribution as neurophysiological evidence linking low-frequency rTMS to cortical inhibition changes in functional gait disorder.
  • Authors include Sattwika Banerjee, Supriyo Choudhury, Dipanwita Santra, Praveen Kumar, Soumen Karmakar, Soumava Mukherjee, Jacky Ganguly, Purba Basu, Mark R Baker, Stuart N Baker, and Hrishikesh Kumar.
  • Low-frequency rTMS is reported to modulate cortical inhibition in patients with functional gait disorder.
  • Stimulation was paired with neurophysiology readouts to characterize the response.
  • The work documents stimulation–response patterns that can serve as biomarkers.
  • Those biomarkers may be applicable to closed-loop rehabilitation system design.

Enhancing the understanding and clinical utility of P300 in psychotic disorders: the contribution of a bimodal oddball paradigm

Clinical Neurophysiology

Tags: EEG, clinical, tier-2

Bimodal oddball P300 paradigms refine ERP features underlying P300-speller BCIs, but psychiatry application dominates. Methods may transfer to clinical BCI calibration. per psychiatry-only calibration.

  • Published in Clinical Neurophysiology, Volume 188 (August 2026), the paper asks how a bimodal oddball paradigm can improve understanding and clinical use of the P300 event-related potential in psychotic disorders.
  • The author team comprises Hendrik Kajosch, Marie Mahoux, Florence Hanard, Clémence Dousset, Paul Deltenre, Matthieu Hein, Anais Ingels, Charles Kornreich, Olivier Le Bon, Gwenolé Loas, Oliver Pogarell, Geerke Steegen, Seline Van den Ameele, and Salvatore Campanella.
  • The study centers on P300 measurement in psychotic disorder populations rather than on assistive or communication brain-computer interfaces.
  • Its core methodological contribution is a bimodal oddball paradigm—an oddball task design that incorporates two sensory modalities to elicit and characterize P300 responses.
  • Bimodal oddball P300 paradigms can sharpen event-related potential features that also support P300-speller BCI decoding pipelines.
  • Although the work is dominated by psychiatric application, the oddball and ERP methods may still inform clinical BCI calibration around P300 responses.
  • Newsletter triage treated the item as secondary BCI relevance because calibration insights are plausible but the evidence base is psychiatry-specific.

Gap junctional coupling of molecular layer interneurons enables transient NMDA driven synchronization

bioRxiv Neuroscience

Published: 2026-05-29T00:00:00+00:00

Tags: computational-neuroscience, methods, tier-3

Mechanistic cerebellar interneuron synchronization via gap junctions—computational neuroscience with indirect relevance to neural population decoding models. Preprint status lowers implementation confidence.

  • Molecular layer interneurons (MLIs) modulate cerebellar cortex output by inhibiting Purkinje cells.
  • MLIs inhibit one another synaptically and are also electrically coupled through gap junctions.
  • Synchronization among MLIs has been observed, but how gap junctional coupling shapes MLI network activity is still poorly understood.
  • MLI gap junctions are dendro-dendritic and propagate signals between coupled interneurons.
  • The authors report that gap junctional coupling among MLIs enables transient, NMDA-driven synchronization.
  • The work is a bioRxiv Neuroscience preprint titled “Gap junctional coupling of molecular layer interneurons enables transient NMDA driven synchronization.”
  • The study focuses on mechanistic cerebellar interneuron synchronization rather than direct brain–computer interface applications.

Clinical Neurophysiology

Tags: EEG, methods, tier-2

Multimodal EEG/ERP plus MRI ML for MCI detection uses neural time-series features but targets diagnostic neurology not interfaces. ERP pipeline knowledge is transferable fMRI weighting reduces BCI relevance. Validation scope unclear.

  • Researchers led by Young Wook Song, Young Ho Park, Seunghu Kim, Euijin Kim, Sungkean Kim, and Kun Ho Lee developed a machine learning approach to detect mild cognitive impairment (MCI).
  • The method combines EEG recordings, event-related potentials (ERPs), and MRI biomarkers into a single multimodal diagnostic pipeline.
  • The study was published in Clinical Neurophysiology, Volume 188, in August 2026.
  • Neural time-series features extracted from EEG and ERP data are central inputs to the classification model.
  • Structural and functional MRI markers are integrated alongside electrophysiological signals rather than used alone.
  • The work is framed for clinical diagnostic neurology in MCI, not for brain–computer interface or real-time control applications.
  • ERP preprocessing and feature-extraction steps in the pipeline may transfer to other EEG-based signal workflows even outside MCI screening.
  • How broadly the model generalizes across cohorts and sites is not clearly established from the available summary.

Cortico-cortical paired associative stimulation increases SMA-M1 facilitation in tremor-dominant Parkinson’s disease

Clinical Neurophysiology

Tags: neuromodulation, clinical, tier-3

Cortico-cortical PAS boosts SMA-M1 facilitation in tremor-dominant Parkinson's—documents plasticity protocols that could inform adaptive stimulation interfaces. Peripheral to BCI decoding pipelines.

  • Cortico-cortical paired associative stimulation increased supplementary motor area–primary motor cortex (SMA–M1) facilitation in tremor-dominant Parkinson’s disease.
  • The study was published in Clinical Neurophysiology, Volume 188, in August 2026.
  • Authors are Jane Tan, Brittany K. Rurak, Rick C. Helmich, Julian P. Rodrigues, Brian D. Power, Peter Drummond, Hakuei Fujiyama, and Ann-Maree Vallence.
  • The protocol used cortico-cortical PAS, pairing stimulation across cortical sites rather than linking peripheral and cortical inputs.
  • Participants were characterized as having tremor-dominant Parkinson’s disease.
  • The main reported outcome was increased SMA–M1 facilitation, indicating strengthened connectivity between those motor regions.
  • The findings describe a plasticity-oriented brain stimulation approach in a Parkinson’s subtype where tremor is the dominant motor feature.
  • The results may help guide timing-based cortical stimulation protocols for adaptive neuromodulation interfaces.

qEEG-guided rTMS for auditory hallucination: a case study of oscillatory-informed neuromodulation

Clinical Neurophysiology

Tags: EEG, neuromodulation, tier-3

Single-case qEEG-guided rTMS for auditory hallucinations illustrates oscillation-informed closed-loop stimulation workflow. Limited n but demonstrates EEG-to-stimulation pipeline pattern.

  • A single-case report in Clinical Neurophysiology describes qEEG-guided repetitive TMS (rTMS) for auditory hallucinations.
  • Authors Harin Oh and Jaewon Lee report oscillatory-informed neuromodulation in a case-study format.
  • The paper is scheduled for August 2026 in Clinical Neurophysiology, Volume 188.
  • Quantitative EEG (qEEG) is used to guide rTMS rather than applying fixed, non-personalized stimulation parameters.
  • The workflow links EEG oscillatory features to stimulation choices in an EEG-to-stimulation pipeline.
  • The approach is framed as oscillation-informed, aligning neuromodulation with patient-specific brain rhythms.
  • Evidence is limited to one patient, so findings illustrate feasibility and workflow rather than population-level efficacy.
  • For BCI and neuromodulation readers, the case shows how closed-loop, brain-state–informed stimulation can be staged from EEG readouts.

Machine learning classification of patients after suicide attempts using demographic data, EEG connectivity and heart rate variability

Clinical Neurophysiology

Tags: EEG, clinical, tier-3

EEG connectivity plus HRV ML for post-suicide-attempt classification is psychiatric phenotyping, not BCI control. Connectivity features echo decoder inputs but lack interface or neuromodulation path. Low near-term execution signal.

  • Researchers used machine learning to classify patients after suicide attempts, combining demographic data with EEG connectivity and heart rate variability (HRV).
  • The study was published in Clinical Neurophysiology, Volume 188, with a publication date of August 2026.
  • Authors are Matej Pribis, Anna Bankwitz, Dávid M. Gyurkó, and Sebastian Olbrich.
  • EEG connectivity features were used alongside HRV as neural and autonomic inputs for patient classification.
  • The work is framed as psychiatric phenotyping rather than brain–computer interface control or neuromodulation.
  • Connectivity features overlap with signal types used in neural decoders but here serve post–suicide-attempt patient stratification, not device control.
  • The study offers limited near-term relevance for BCI execution or therapeutic interface development.

EEG connectivity changes in early response to antidepressant treatment

Clinical Neurophysiology

Tags: EEG, clinical, tier-3

Antidepressant-response EEG connectivity shifts are pharmacology monitoring, not neural interface work. Connectivity methods overlap BCI pipelines but clinical psychiatry framing limits decision utility for BCI teams.

  • The paper is titled “EEG connectivity changes in early response to antidepressant treatment” and was published in Clinical Neurophysiology, Volume 188, in August 2026.
  • Authors are Aditi Kathpalia, Ioannis Vlachos, Jaroslav Hlinka, Martin Brunovský, Martin Bareš, and Milan Paluš.
  • The study centers on EEG connectivity changes during the early phase of antidepressant treatment response.
  • Antidepressant-response EEG connectivity shifts are framed as pharmacology monitoring rather than neural-interface development.
  • Connectivity analysis methods used in this work overlap techniques common in brain–computer interface pipelines.
  • The clinical psychiatry framing limits how directly the findings inform decision-making for BCI teams.

previous << weekly >> next

How this week was triaged

Graph View

Backlinks