Journal of Neural Engineering

ObjectiveBiopotential measurement devices, such as EEG, ECG, and EMG recorders, are available in low-cost, open-source implementations with standard specifications. However, high-end systems remain expensive and predominantly proprietary, limiting accessibility and customization by research laboratories. In addition, neurophysiology techniques such as bioimpedance-based Fast Neural Electrical Impedance Tomography (FN-EIT) also rely on these systems for data acquisition. This work aimed to develop an open-source biopotential recording system using off-the-shelf components that achieves performance comparable to high-end devices. ApproachWe designed our system to provide simultaneous sampling over 32 channels, 24-bit resolution, 10 kHz bandwidth, 50 kHz sampling rate, and battery-powered operation while reducing cost by two orders of magnitude. System performance was evaluated comparatively against a reference device. Initial validation involved benchtop recordings in saline solution and standard non-invasive biopotential measurements (ECG and EMG). Further in-vivo validation was performed by recording evoked electrophysiological responses and FN-EIT traces from the sciatic nerve of a rat during tibial branch stimulation. Main resultsEMG recordings showed comparable RMS peak amplitudes (814{+/-}153 {micro}V vs. 897{+/-}113{micro}V, p=0.07), while ECG-derived heart rates closely matched between systems (64.8{+/-}4.0 bpm vs. 65.1{+/-}3.1bpm, p=0.54). During in-vivo recordings, compound action potentials exhibited comparable amplitudes and morphology (129{+/-}26 mV vs 128{+/-}25 mV, P=0.15). FN-EIT recordings showed strongly correlated baseline voltages (R>0.93, P=0.11), sub-microvolt noise levels (0.83{+/-}0.36{micro}V vs. 0.42{+/-}0.25{micro}V, p<0.05), and comparable impedance variations (0.006{+/-}0.002% vs 0.005{+/-}0.003, P>0.05). FN-EIT images of functional activity recorded with the novel device closely matched reference ones, exhibiting a 98.5% overlap in activated area. SignificanceThe proposed open-source device has the potential to broaden research access to customizable, high-specification data acquisition hardware and facilitate wider adoption of specialized neural recording techniques such as FN-EIT.

Objective: Adaptive neuromodulation systems and implantable brain-computer interfaces (BCIs) are promising therapies for neurological and psychiatric disorders. However, their broader translation into research and clinical practice remains limited by technological complexity, restricted access to implantable research platforms, and the lack of standardized, reproducible experimental workflows. We therefore aimed to develop and validate an open, general-purpose translational ecosystem that enables rapid development, evaluation, and dissemination of novel neuromodulation and implantable BCI paradigms. Approach: The CorTec Brain Interchange (BIC) implantable neural sensing and stimulation device was integrated with the open-source BCI2000 platform to create a modular, extensible neuromodulation ecosystem. We established a standardized battery of quantitative assessments to characterize implantable neuromodulation systems to comprehensively evaluate the CorTec BIC device through benchtop characterization, long-term preclinical in vitro and in vivo validation, and a human proof-of-concept demonstration. Results: Benchtop and saline testing provided a comprehensive technical ex vivo characterization of the BIC device, independently validating previously reported performance while extending its characterization through quantification of the recording noise floor, stimulation and acquisition latencies and impedance measurement accuracy. Long-term in vivo validation in five canines, with the longest implantation exceeding three years, demonstrated stable chronic recordings while capturing progressive channel deterioration and its underlying mechanical causes. The ecosystem enabled active functional decoding more than two years after implantation, implementation of closed-loop stimulation using arbitrary spectral biomarkers, detection and modulation of epilepsy-associated biomarkers, and brain stimulation evoked potential recordings. In addition, we translated an established one-dimensional BCI cursor control paradigm to the BIC benchtop evaluation kit and demonstrated its feasibility in a human participant. Finally, we openly provide standardized surgical, imaging, and analysis pipelines together with datasets and software to facilitate reproducible neuromodulation research. Significance: We present a versatile, open-source translational ecosystem that supports a wide range of neuromodulation and implantable BCI applications with minimal modification. This battery of quantitative assessments can be applied generally as a blueprint for systematic characterization of implantable neuromodulation systems. By combining comprehensive hardware characterization with standardized software tools and experimental workflows, this work provides both an essential reference for researchers adopting the Brain Interchange platform. The ecosystem lowers technical barriers to implantable neurotechnology research, promotes reproducibility, and provides a foundation for accelerating the development and clinical translation of next-generation adaptive neuromodulation and implantable BCI therapies for patients with neurological and psychiatric disorders.

Deep brain stimulation (DBS) modeling relies heavily on biophysical neuron models to estimate neural activation thresholds and predict stimulation spread. In this study, we systematically compared a widely adopted axon model, the McIntyre-Richardson-Grill (MRG) model (Model I), with a more detailed biophysical model, the Cohen model (Model II), to assess how structural and electrophysiological differences affect predicted DBS outcomes. Electric field distributions generated by 2202 DBS lead were applied to the neuron models as extracellular input stimuli. Both models were simulated under biphasic pulse stimulation across varying axon-electrode distances, pulse widths, and stimulation frequencies. Activation distances ranged from approximately 2 to 10 mm depending on stimulation parameters and contact location. At 2 mA, Model I achieved an activation distance of 6 mm, whereas Model II reached 10 mm, indicating greater excitability. Across matched fiber tracts, threshold differences ranged from -1.40 mA to 0.27 mA, with Model II requiring lower thresholds in 97.7% of cases. Both models showed a strong inverse relationship between pulse width and activation threshold. However, frequency responses differed: Model II exhibited increasing thresholds at higher frequencies, while Model I showed a slight decrease. Machine learning regressors trained on distance, pulse width, and frequency achieved high predictive accuracy, with Gradient Boosting performing best. Model II demonstrated superior prediction metrics (R^2 = 0.986; RMSE = 0.045 mA; MAE = 0.034 mA) compared to Model I (R^2 = 0.977; RMSE = 0.089 mA; MAE = 0.068 mA). Overall, both models reliably estimate DBS-induced activation, but structural differences significantly affect excitability and frequency-dependent behavior. With appropriate awareness of their respective strengths and limitations, either model can be used to derive activation distances for estimating electric field isolevels and the volume of tissue activated in patient-specific DBS simulations.

Surface electromyography (sEMG) is a robust non-invasive modality for human-machine interaction, yet its application remains largely limited to coarse motor tasks such as grasping or rotation. The decoding of fine motor skills, specifically handwriting, remains a challenging problem with potential relevance for prosthetic control and natural communication interfaces. In this work, we explore a Transformer-based alternative to classical signal-processing pipelines that treats multi-channel sEMG signals as complex time series. We introduce DualMyo, a specialized model integrating Patch Embeddings and Rotary Positional Embeddings (RoPE) to capture the intricate spatio-temporal dynamics of myoelectric activity. Our experimental results show strong intra-session performance. Furthermore, we address the inherent challenges of signal drift and sensor displacement in cross-session applications. We show that a lightweight fine-tuning strategy of 10 epochs enables DualMyo to effectively adapt to session variability, achieving approximately 91\% accuracy with two examples per digit. These findings provide a promising step toward adaptive sEMG-based handwriting interfaces, although further validation is required for real-time and multi-subject deployment and neuromuscular control.

Beta oscillations (13-30 Hz) are a prominent sensorimotor neural rhythm and an important biomarker in neurorehabilitation. These oscillations propagate along the corticospinal pathway and are expressed in the discharge patterns of spinal motor neurons, enabling the assessment of corticomuscular coupling. Moreover, peripheral beta band oscillations have recently emerged as a potential control signal for motor augmentation interfaces. However, it remains unclear whether peripheral beta activity simply reflects cortical oscillations or is partly shaped by peripheral mechanisms, and to what extent it can be voluntarily controlled. To address these questions, we developed a 10-day neurofeedback protocol in which participants learned to up-regulate peripheral beta band activity. Subjects were trained to exploit movement cancellation, a behaviour naturally associated with increased cortical and muscle beta band activity, as a two-state strategy to voluntarily modulate peripheral beta band power. Each session included a guided familiarization phase based on a GO/NO-GO task, in which participants familiarized with movement cancellation through guided visual cues, followed by an asynchronous control phase in which they self-initiated the same strategy without external guidance to increase peripheral beta band activity in a target window. Participants progressively improved their ability to voluntarily modulate peripheral beta band activity. Peripheral beta band power during movement cancellation increased significantly across training days in both the familiarization and asynchronous control phases. Intramuscular coherence in the beta band also increased, indicating enhanced common synaptic input to the motor neuron pool in this band. In contrast, cortical beta power and corticomuscular coherence remained unchanged. Together, these findings demonstrate that peripheral beta band activity is a dynamic neural feature that can be voluntarily shaped through training, supporting its potential as a non-invasive control signal for future neurorehabilitation and motor augmentation technologies.

Brain-computer interface (BCI) research relies on multistage computational pipelines, but progress has been slowed by fragmented data formats, heterogeneous decoder implementations and hardware-specific deployment toolchains. Here, we introduce BCIJelly, a unified ecosystem that standardizes 18 BCI datasets into AI-ready inputs and integrates 15 benchmark decoders, 80 reusable modules, automated architecture search (AAS) and hardware-aware neuromorphic deployment. Our AAS constructs task-specific decoders without manual design and extends into a large language model (LLM)-driven closed-loop mode supporting single-task, multitask and cross-species decoder design. A single-command pipeline compiles trained decoders for neuromorphic hardware, reducing power consumption by 30 to 50 times while preserving decoding performance. An interactive visualization software enables code-free exploration of neural recordings and decoding outputs. BCIJelly is validated across five BCI paradigms (motor, visual, speech, emotion and auditory) in humans, macaques and mice, providing an extensible ecosystem connecting data standardization, decoder development, systematic evaluation and hardware-aware deployment for BCI research.

Surface electromyography (sEMG) offers a silent and wearable input modality, but its practical use is limited by variability across users and recording sessions. This study presents a compact CNN- Transformer model for decoding isolated handwritten digits from eight-channel sEMG signals. The model combines trainable signal preprocessing, convolutional feature extraction, and Transformerbased temporal modeling. It was evaluated on ten recordings from five participants using recordingseen classification, leave-one-recording-out (LORO) generalization, and few-shot adaptation. The model achieved a mean macro F1 score of 0.924 {+/-} 0.059 in the recording-seen setting and 0.619 {+/-} 0.252 under zero-shot LORO evaluation. Adaptation using two labeled trials per digit increased macro F1 to 0.828 {+/-} 0.112, while ten trials per digit achieved 0.925 {+/-} 0.053. The proposed architecture also outperformed classical and neural baselines in the controlled LORO benchmark. These results indicate that compact CNN-Transformer models, combined with lightweight target-recording calibration, provide a promising basis for adaptive sEMG-based input systems.

Generalizable movement decoding remains a central challenge for invasive brain--computer interfaces (BCIs), as decoders trained under limited calibration conditions often fail to generalize to unseen movement speeds, limbs, and subjects. Existing decoding methods are typically trained on paired data collected under restricted conditions. How to incorporate behavioral structure from unpaired data for robust out-of-distribution (OOD) decoding therefore remains unresolved. To address this, we propose LAND (Latent Aligned Neural-behavioral Dynamics), a framework that aligns latent neural and behavioral dynamics through flow matching. By learning a neural-to-behavioral transport map and using behavioral-dynamics priors from unpaired data to encourage structured neural manifolds, LAND regularizes representation geometry to promote cross-domain generalization. We evaluate LAND on synthetic neural data, epidural BCI recordings from a tetraplegia participant, and multi-electrode array (MEA) recordings from nonhuman primates (NHPs). Across these settings, LAND improves zero-shot generalization to OOD movement speeds and yields speed-modulated manifolds. With limited target-domain fine-tuning, it further improves transfer across limbs and subjects. These results support flow-based neural--behavioral alignment with unpaired kinematic priors as an approach for learning transferable neural representations and robust movement decoding across behavioral and recording domains.

Pathological tremor is a neurological condition that impairs fine motor tasks, affecting 1% of the general population and 4% of the elderly. Tremors arise when muscles micro-oscillations synchronize and phase lock, typically within a 4-12 Hz frequency range. Administering beta-blockers can reduce tremor severity, but doses are hard to personalize, with heavy doses of propranolol correlating with low blood pressure, dizziness, and nausea. In this project, we aimed to model tremor and create a closed-loop control framework to suppress tremor amplitude while minimizing pharmacological dependence. Because side effects constrain the use of pharmacological suppression alone, we investigated noninvasive neuromodulation. We used vibrotactile stimulation (VTS) to disrupt pathological tremor synchronization and reduce oscillatory amplitude. We hypothesized that tremor suppression involving VTS followed a nonmonotonic relationship, tested by determining whether maximum relief requires an adaptable framework. The procedure consisted of constructing a propranolol-reduction simulation by implementing a Hill curve, where we calculated and utilized tremor reduction, heart rate (HR) drop, and blood pressure (BP) drop. We then built a device to capture tremor-related data and create vibration using two linear resonant actuator (LRA) coin motors. We connected it to a microcontroller, where we determined optimal vibration frequencies through a feedback loop. Across 50 trials, VTS alone reduced tremor amplitude by an average of 37.3%, reducing the propranolol dose needed to reach 50% total tremor reduction by 71.9%, lowering the modeled blood pressure drop from 38.1 to 18.9 mmHg. This device demonstrates proof-of-concept for a nonmonotonic tremor-vibration relationship to reduce dependency on propranolol in the treatment of pathological tremor. These propranolol dose-reduction estimates are derived from computational simulation and have not been clinically validated; they are not intended as a recommendation to alter prescribed medication.

EEG microstates are a distinct number of quasi-stable spatial distributions of brain activity. Microstate trajectories are strongly suspected to reflect the underlying neural mechanisms during information processing and are therefore also called the "building blocks" of human thought. In this study, we examined, if EEG microstate sequences can serve as potential triggers for a Brain-Computer Interface (BCI). To this end, a semi-supervised deep learning model architecture consisting of an LSTM-based autoencoder and a dense neural network was utilized to classify between left- and right-hand motor imagery EEG data, with the resulting classification output serving as the BCI trigger. On the one hand, this was done in a 2-step approach, in which the autoencoder and classifer have been trained separately. On the other hand, an end-to-end approach was employed, where training was performed by combining reconstruction and classification losses. Results show that the proposed model architecture was able to extract relevant features from microstate sequences and exploit them for within subjects and sessions classification. Applying transfer learning to session-to-session or across-subject transfer resulted in peak classification accuracies around 89%. We also investigated to what extent transfer learning has to be applied to reach considerable classification accuracies serving as the calibration time representative. We found that on average around 400s are needed for BCI calibration when emplyoing our approach to reach 80% classification accuracy. The present study signifies that the investigation of EEG microstate trajectories can be a promising approach for extracting BCI triggers, as it reduces the dimensionality of multi-channel recorded EEG signals to a distinct number of brain states over time. Deep learning methods, especially transfer learning, applied to EEG microstate trajectories seem promising regarding user-convenient and calibration-free BCIs in real-world applications.

Implantable brain-computer interfaces (iBCIs) with single-neuron resolution are showing great promise for restoring mobility and communication in individuals with spinal cord injury or motor neuron disease. Stroke is the most common cause of acquired brain injury and a major contributor to long-term disability, making chronic stroke a highly relevant indication for iBCIs. However, whether stable intracortical recordings can be obtained from the structurally lesioned human brain is unknown. We report recordings from four 64-channel microelectrode arrays implanted in a participant with chronic aphasia after a large left-hemispheric stroke. The arrays targeted right-hemispheric frontoparietal regions homotopic to the damaged left-hemispheric language network. Across 111 sessions spanning 1,240 days, unit yield and signal quality remained stable. Waveform-based tracking reliably identified individual units across sessions, including across extended recording gaps. Short- and long-term unit stability was comparable to previous reports from iBCI participants without structural brain lesions, and tracked units showed consistent spiking properties across sessions. Our findings provide the first evidence that single-neuron recordings can remain stable over the long term in the stroke-lesioned human brain. They establish the feasibility of chronic, cellular-resolution iBCIs after stroke and support the development of neurorestorative applications for deficits caused by structural brain injury.

Drug-resistant epilepsy affects one-third of patients with persistent seizures despite optimal therapy. Intersectional short-pulse (ISP) stimulation is a novel transcranial electrical stimulation technique designed to deliver temporally precise, spatially targeted modulation of pathological brain activity. Here, we combined computational modeling with measurements in a rat epilepsy model and in patients with epilepsy to map the relationship between stimulation phase and seizure attenuation. In silico simulations of epileptiform networks showed that ISP stimulation significantly shortened seizure duration, with efficacy strongly depending on the phase of delivery. Phase-targeted stimulation during the rising phase and around the peaks (~45-90{degrees}) of the seizure oscillations led to the greatest reduction in seizure length. In rodents, ISP decreased seizure duration by 42.4% and shortened generalized seizure segments by 58.3%. In humans, stimulation reduced seizure length by 60.9% compared to control seizures. Phase dependence was evident across models and species, with a prominent efficacy window in the rising-to-peak portion of the ictal oscillation and model-specific secondary windows. These findings show that phase-targeted ISP can substantially shorten seizures and support phase-resolved stimulation as a precision-neuromodulation approach for epilepsy.

BackgroundHearing aids and cochlear implants (CIs) are the primary interventions for sensorineural hearing loss, restoring auditory function through amplification and intracochlear electrical stimulation, respectively. For those with residual low-frequency hearing, the combined electric-acoustic stimulation (EAS) has demonstrated superior speech perception, particularly in noisy environments, compared to either modality. However, CI surgery carries inherent risks, including postoperative hearing loss, which undermines EAS benefits and limits future rehabilitation options. To overcome these limitations, we propose a non-invasive alternative: extracochlear electric and acoustic stimulation (EEAS), delivering electrical stimulation via transcutaneous electrodes without surgery. Here, we present a first systematic investigation of non-invasive extracochlear electrical stimulation using ear canal electrode montages, evaluating its feasibility, perceptual effects, and key parameters across diverse hearing statuses. MethodsWe conducted a controlled, within-subject study with 15 participants: 5 with normal hearing (NH), 5 with high-frequency hearing loss (HI), and 5 with severe-to-profound deafness (PL). We used charge-balanced sinusoidal stimuli (125-4000 Hz) applied via an ear canal electrode and four return electrode montages, including contralateral ear canal, contralateral mastoid, ipsilateral mastoid, and forehead electrodes. Participants rated auditory sensations, including loudness, sound quality, and lateralization, as well as side effects on separate 0-10 scales, with current intensity increased up to 2 mA/cm{superscript 2}. Thresholds and perceptual responses were analyzed across frequencies, electrode configurations, and hearing groups. ResultsReliable auditory percepts were elicited across all groups. NH participants reported pure-tone sensations, whereas HI and PL participants perceived broadband, noise-like sounds. Loudness decreased with increasing frequency, particularly for HI and PL, with minimal responses in the high-frequency range. The current threshold increased with stimulation frequency, whereas the threshold expressed as charge per phase remained constant, suggesting that charge per phase primarily determines neural activation, whereas current amplitude is more closely associated with the intensity of auditory and side effect perception. Contralateral montages produced significantly higher loudness ratings than ipsilateral or forehead configurations. The forehead montage was poorly tolerated, leading to early termination due to discomforting side effects. Sound lateralization was predominantly central or bilateral with contralateral setups, while ipsilateral and forehead configurations yielded ipsilateral perceptions. ConclusionsNon-invasive extracochlear electrical stimulation via ear canal electrodes is feasible and perceptually effective across a spectrum of hearing statuses. Perceptive outcomes are strongly influenced by electrode montage and residual hearing, with evidence of electrophonic excitation in NH individuals and electroneural activation in HI and PL participants. Contralateral mastoid electrode configurations offer the optimal balance of perceptual strength, tolerability, and spatial localization. These findings establish a critical foundation for the development of EEAS devices, demonstrating that non-invasive electrical stimulation can generate meaningful auditory percepts, paving the way for safe, accessible, and integrated hearing rehabilitation solutions. This work informs future EEAS developments and advances the path toward clinically viable, non-invasive cochlear stimulation.

AbstractComputational models of temporal interference stimulation (TIS) commonly report a single electric-field estimate for a given anatomy and electrode montage. Because non-deterministic tetrahedral mesh generation does not produce a unique discretisation of a fixed tissue-label image, a single mesh realisation may introduce numerical variability. We quantified variation across independent mesh realisations and contrasted it with repeated downstream simulation execution on a single selected mesh. Ten head models were evaluated for stimulation of the left hippocampus and right primary motor cortex (M1). For every model and target, we generated 40 independent meshes and performed one complete simulation on each. Separately, we selected the mesh whose parcel-level field estimate was closest to the median and repeated downstream operations 40 times while holding that geometry fixed, yielding 1,600 TIS simulations in total. The primary outcome was the spatial median of the TIS envelope field within a spherical target region. Across independently remeshed runs, within-participant coefficients of variation were 1.81-3.65% for the hippocampus and 1.62-2.79% for M1. Repeated execution on a fixed mesh reduced run-to-run standard deviation by more than 99%, demonstrating that workflow variability is driven almost entirely by non-deterministic mesh generation rather than solver instability, numerical rounding, or post-processing. Single-run mesh realisations preserved overall cohort ordering (median Kendalls{tau} of 0.867 for the hippocampus and 0.911 for M1) but frequently inverted the rank order of participant pairs with similar predicted fields. Furthermore, a bootstrap analysis demonstrated that averaging five to ten independent remesh runs effectively suppressed this stochastic noise. These results quantify single-workflow repeatability rather than absolute error. Stochastic mesh variation should therefore be controlled or mitigated through multi-run averaging whenever experimental conclusions depend on subtle field differences or fixed neuromodulation thresholds.

Objective: Low-intensity transcranial magnetic stimulation (LI-TMS) is being investigated as a gel-free alternative to transcranial electrical stimulation (tES), but existing systems remain almost exclusively single-channel and cannot electronically steer the induced electric field. We present the design, modeling, and experimental measurement of a wearable whole-head, multichannel, steerable LI-TMS array. Methods: The system comprises a 102-channel conformal coil array with independently controlled drivers capable of arbitrary waveform synthesis, together with a boundary element fast multipole method (BEM-FMM) framework that computes the coil currents required to produce prescribed cortical field patterns. A 12-channel prototype was characterized by coil-current, electric-field, and thermal measurements. Results: The prototype produced a peak primary electric field of approximately 1 V/m measured in air 4 cm from the inner helmet surface. Whole-array modeling attained cortical fields of up to 1.5 V/m, reproduced the field distribution of a clinically validated low-intensity stimulator to within 3%-5%, and demonstrated focal targeting of the dorsolateral prefrontal cortex, simultaneous delivery of electric field to the default mode network nodes, and synthesis of electric fields following the traveling alpha wave. Conclusion: Electronically steerable, whole-head LI-TMS is feasible using accessible microprocessor-controlled power electronics. Significance: The array reaches the cortical field regime of tES without scalp contact or the associated shunting of current through the scalp, offering a route to testing network-level, phaselocked weak-field neuromodulation.

BackgroundThe integration of virtual reality (VR) and non-invasive brain stimulation (NIBS), particularly transcranial magnetic stimulation (TMS), represents a promising approach for closed-loop neuromodulation. Yet the concurrent application remains limited, partly due to insufficient characterization of hardware compatibility of head-mounted displays with standard TMS coil placement protocols. ObjectiveTo systematically quantify the coil-to-scalp distance constraints imposed by VR headsets across cortical targets and coil orientations and to determine feasible intensity compensation ranges based on stimulator output parameters. MethodsNeuronavigated coil positioning was performed on five anatomically realistic 3D-printed head models across 26 scalp positions in eight coil orientations based on the 10-10 EEG system and dorsolateral prefrontal cortex (DLPFC) using two VR headsets of notably different form factors (Meta Quest 2 and Bigscreen Beyond). The deviations of coil positions from intended targets were registered and quantified as coil-to-scalp distance displacement. Individual electric field (E-field) simulations were conducted in SimNIBS at the F3 position across 4-40 mm coil-to-scalp distance to characterize field decay and assess the limits of intensity compensation. ResultsBoth in the directed DLPFC targeting and in systematic scalp positions evaluation, the Meta Quest 2 headset substantially increased coil-to-scalp distance over prefrontal regions, exceeding the compensable range across all metrics. The Bigscreen Beyond headset produced significantly smaller coil-to-scalp distance displacement in prefrontal regions, remaining within feasible E-field intensity compensation limits. Single-pulse and iTBS protocols did not induce functional interference with the hardware under realistic targeting conditions. ConclusionVR headset geometry is the primary determinant of concurrent VR-TMS feasibility. The findings define practical quantitative hardware design requirements and boundaries for future integrated VR-TMS systems and provide a practical framework for optimizing existing VR-TMS protocols.

People with lower limb amputation lack somatosensory feedback from their prosthesis, and this loss contributes to functional deficits, including balance and gait impairments. Recent advances in neuroprostheses have demonstrated that electrical stimulation of sensory nerves in the residual limb and spinal cord can restore lost sensations in the lower limb. To maximize the efficacy of these somatosensory neuroprostheses, the restored sensations should be intuitive, seamlessly integrating into the sensorimotor network. However, it is challenging to quantify the intuitiveness of these evoked sensations. Recent studies have proposed using crossmodal congruency effect (CCE) tasks for this purpose in people with upper-limb amputation. The current study tests the feasibility of the CCE task for assessing the intuitiveness of sensory feedback in the lower limb. We hypothesized that CCE score would reliably differentiate between a more natural (pneumatic) sensation and a less natural (electric) sensation at two locations: the knee and the foot. Across fifteen able-bodied individuals, we observed that the CCE task differentiates sensory modalities at the knee, but not at the foot. Identification of external factors affecting the CCE is needed before it can be implemented to measure intuitiveness of sensory feedback in lower-limb amputees.

Background and Objective Epileptic seizure prediction is a critical challenge requiring the discrimination of subtle preictal physiological changes from interictal brain activity. While deep learning has shown promise in this domain, existing models often face limitations due to small EEG datasets, high computational costs for training from scratch, and a lack of patient-independent generalizability. In this paper, we present a novel framework for EEG-based seizure prediction that leverages pre-trained Vision Transformers (ViTs) through custom architectural modifications and optimized re-training strategies. Methods Our primary contributions include: [bullet]CVIT-ESP: A family of vision transformer architectures that replaces standard patch embedding layers with custom N-dimensional CNN stages to refine EEG representations. [bullet] ESPFormer: A lightweight, custom-designed transformer specifically engineered to mitigate overfitting on limited-scale EEG datasets. We identified optimal fine-tuning combinations for transformer blocks by devising a heuristic search-space reduction strategy, significantly reducing the training complexity. We validated our methods using the patient-independent MLSPred-Bench, involving 12 diverse benchmarks with varying seizure prediction horizons. Results Results demonstrate a clear progression in performance: while prior ResNet and vanilla Transformer models achieved an AUC-ROC of 69.0%, our CVIT-ESP architectures achieved the highest performance with a maximum average AUC of 76.4%. Conclusions These findings suggest that adapting pre-trained ViTs with domain-specific CNN front-ends and strategic fine-tuning offers a robust, generalizable, and resource-efficient path forward for clinical seizure prediction systems. Our code is available at: https://github.com/pcdslab/CVitEsp and https://github.com/pcdslab/ESPFormer

Large language models are integrated into brain-computer interfaces for communication, but accuracy does not show whether an emitted character depended on neural evidence or on the language-model prior. We retrospectively re-decoded 3,373 P300-speller selections from 47 people with amyotrophic lateral sclerosis, reconstructing a neural posterior and combining it with 25 language priors ranging from 5-grams to 46.7-billion-parameter models. Under held-out, per-source calibration and equal fusion weighting, the prior accounted for a participant-weighted mean 8.6% of posterior displacement (median across selections, 2.9%); the corresponding neural contribution fraction was 0.914 (95% CI, 0.896-0.934). In 4.4% of selections (95% CI, 3.5-5.3), the fused system emitted the intended character although neural evidence alone would not have selected it. Results were similar across 21 neural language models. Accuracy alone does not reveal how strongly a fused brain-computer interface decision depends on its language prior.

Accurate phase tracking of deep-brain activity is critical for effective closed-loop and phase-locked neuromodulation therapies. However, direct access to deep neural phase through intracranial recordings remains clinically restrictive due to the invasiveness. Here we validate and clinically benchmark the Gabor-Nelson (GN) dipole estimation method for reconstructing deep-brain oscillatory phase from non-invasive scalp EEG. GN is a geometry-based, imaging-independent approach that offers computationally efficient dipole reconstruction and has rarely been applied to source-level phase estimation in human neuroscience. We compared GN with an established MRI-informed Inverse Solution (IS) method using a three-stage reconstruction pipeline consisting of dipole modeling, dimensionality reduction, and frequency-dependent phase-delay correction. Validation is performed using (i) cadaveric recordings, where known ground-truth seizure waveforms were replayed through implanted deep electrodes, and (ii) simultaneous scalp EEG and SEEG recordings in human patients, where pseudo-ground truth was approximated via the intracranial contacts. GN achieved phase accuracy and signal fidelity comparable to IS across both datasets despite requiring no anatomical imaging. In cadaver recordings, phase-corrected reconstruction correlations exceeded r > 0.91 and {Delta}{Phi} < 9{degrees} in mean phase error. In patient SEEG data, GN reached up to r {approx} 0.80 with phase offsets suitable for neuromodulatory timing. GN offers a viable, low-barrier, imaging-independent alternative to traditional inverse modeling for non-invasive seizure phase tracking. This framework opens pathways for scalable, phase-locked and closed-loop stimulation therapies in epilepsy and potentially other network-based brain disorders.