Journal of neural engineering最新文献

Objective. Low back pain (LBP) is a significant public health issue. Despite current medical imaging and neurophysiological tests, up to 90% of patients lack a clear cause, leading to a diagnosis of chronic primary LBP (CPLP). Non-invasive functional near-infrared spectroscopy (fNIRS) was employed to detect spinal cord dysfunctions by recording perispinal neurovascular response (NVR).Approach. In a prospective study of 71 CPLP patients and 65 healthy age-matched volunteers, pain maps, visual analog scale (VAS), body mass index (BMI), posterior tibial nerve conduction velocity (NCV), and lumbar and cervical NVRs triggered by non-noxious electrical stimulation of this nerve were assessed.Main results. CPLP patients exhibited a 53.29% reduction in NVR amplitude at the cervical level compared to the controls, with no significant difference at the lumbar level. CPLP patients compared to controls show a rise time of 6.64% and 5.14% larger in cervical and lumbar recordings, respectively, but a duration of 10.11% and 5.32% shorter, respectively. Posterior tibial NCV was within normal clinical range in both groups. In CPLP patients, VAS scores were negatively correlated with NVR rise time, amplitude, and duration at the lumbar site, as well as with rise time and duration at the cervical site (p< 0.05). Additionally, BMI showed a negative correlation with all NVR parameters at both recording sites in CPLP patients, but not in controls (p< 0.05).Significance. This is the first report of perispinal NVR dysfunction in patients with CPLP. Its results suggest a loss of inhibitory regulation in the lumbar spinal cord in CPLP patients and demonstrate the potential of fNIRS to detect and quantify spinal cord neurovascular dysfunctions. For the first time, perispinal NVR dysfunction is reported in CPLP patients, suggesting an altered descending modulation system at the lumbar spinal cord.

Objective.In this study, we present a novel computational framework that combines the Hindmarsh-Rose (HR) neuronal model with evolutionary game theory on networks to simulate and interpret synaptic-level interactions within neuronal populations. Our approach preserves the features of the HR model-capable of generating both spiking and bursting dynamics-while integrating game-theoretic principles that govern the balance between emulative and non-emulative behaviors across neurons.Approach.Neurons were modeled as strategic agents whose interactions evolve according to game-theoretic principles, allowing us to capture emergent network dynamics beyond classical electrophysiological analyses. A key innovation of our work is the formulation of a parameter estimation method based on adaptive observers, which enables the recovery of game-theoretic parameters solely from partial state observations. The proposed framework is validated through numerical simulations, demonstrating its ability to recover hidden parameters and accurately predict system behavior under diverse conditions.Main results.By applying the devised approach to synthetic datasets mimicking real electrophysiological recordings, we highlight its applicability in distinguishing neuronal populations based on their strategic interactions. In this context, the model is shown to faithfully reproduce both spiking and bursting behaviors, capturing the diverse electrophysiological patterns observed inin vitroexperimental settings. Furthermore, we explore the potential of this model in experimental data analysis by suggesting that the estimated parameters may serve as discriminative markers for different neuronal types and structural characteristics.Significance.The integration of dynamical systems theory, game-theoretic modeling, and adaptive estimation provides a robust quantitative tool for investigating complex neuronal network dynamics. Our results quantitatively demonstrate the scalability and accuracy of the method in parameter estimation, reinforcing its value for systematic analysis of synaptic interactions and advancing our understanding of neuronal network dynamics.

Objective.Most existing seizure prediction approaches rely on cohort-based models or assume a single model suffices per patient, overlooking clinical and electrophysiological variability across seizures. This study aims to overcome these limitations by introducing a subject-specific seizure prediction framework that models intra-subject heterogeneity by identifying and clustering seizure-specific preictal patterns usinglong-termintracranial EEG (iEEG) recordings collected over a one to two week duration.Approach: Absolute, relative, and ratio power spectral density features are extracted from twelve frequency bands, and the minimum uncertainty and sample elimination algorithm is used for unsupervised feature selection on a per-seizure basis. Weighted aggregation is then applied to form seizure-specific feature sets. Seizures are grouped into clusters based on feature similarity, and separate classifiers are trained for each cluster. Model predictions are combined using a grid optimizedk-of-Nvoting strategy. Evaluation is conducted on long-term iEEG recordings from ten patients using cross-validation across seizure-containing sessions.Main results.When clustering is applied, the mean sensitivity across subjects is improved from 89.17% to 98.54%, while the mean FPR is reduced from 1.15/day to 0.62/day. Additionally, the median number of features required per subject decreased from 22 to 14, reflecting a 36.4% reduction in model complexity. Finally, in 72.5% of subject-folds, the number of algorithm-identified clusters equaled or exceeded the clinically annotated seizure types, with a linear trend indicating latent electrophysiological variability beyond clinical labels.Significance.These findings highlight the value of modeling seizure diversity within individuals and support the development of more personalized and interpretable seizure forecasting systems.

Objective.Utah arrays are widely used in both humans and non-human primates (NHPs) for intracortical brain-computer interfaces, primarily for detecting electrical signals from cortical tissue to decode motor commands. Recently, these arrays have also been applied to deliver electrical stimulation aimed at restoring sensory functions. A key challenge limiting their longevity is the micromotion between the array and cortical tissue, which may induce mechanical strain in surrounding tissue and contribute to performance decline. This strain, due to mechanical mismatch, can exacerbate glial scarring around the implant, reducing the efficacy of Utah arrays in recording neuronal activity and delivering electrical stimulation.Approach.To investigate this, we employed a finite element model to predict tissue strains resulting from micromotion.Main results.Our findings indicated that strain profiles around edge and corner electrodes were greater than those around interior shanks, affecting both maximum and average strains within 50µm of the electrode tip. We then correlated these predicted tissue strains within-vivoelectrode performance metrics. We found negative correlations between 1 kHz impedance and tissue strains in human motor arrays and NHP area V4 arrays at 1 month, 1 year, and 2 years post-implantation. In human motor arrays, the peak-to-peak waveform voltage and signal-to-noise ratio (SNR) of spontaneous activity were also negatively correlated with strain. Conversely, we observed a positive correlation between the evoked SNR of multi-unit activity and strain in NHP area V4 arrays.Significance.This study establishes a spatial dependence of electrode performance in Utah arrays that correlates with tissue strain.

Objective.Passive brain-computer interface (BCI) based on electroencephalography (EEG) has gained traction as reliable method for monitoring human vigilance in attention-demanding critical contexts. Unfortunately, the lack of extensive public datasets compromises artificial intelligence (AI) research. Proposing a solution to this issue, we augmented two EEG datasets using generative adversarial networks (GANs). Furthermore, we defined a quality-assessment pipeline to overcome the absence of a univocal method to test synthetic data.Approach.Using GAN, we augmented a publicly resting-state EEG dataset sustained attention to response task and a custom one simulating activity during repetitive tasks. After extracting relevant time-variant rhythms via the continuous wavelet transform, we quantitatively compared synthetic data with the real one using L2 distance and cross-correlation function. To evaluate the impact of data augmentation, we trained six forecasting models, three on the original and three on the augmented datasets, over the whole, half and a quarter of total available data, and compared improvements in MAE and symmetric mean absolute percentage error (SMAPE). To study the forecaster's embeddings, we computed a metric inspired by the Fréchet inception distance (FID) between latent values of real and synthetic data. Finally, to offer a baseline comparison, we extended the performance and embeddings analysis to data generated by a simple linear interpolation method.Main results.The integration of GAN-produced synthetic data improved signal prediction, as evidenced by a 29.0%, 46.4%, 37.4% reduction in mean absolute error (MAE) for splits of the resting-state dataset, and an average MAE reduction of 15.4%, 21.2% for 100% and 50% splits, and a ∼2.5% increase for the 25% split. Conversely, training on interpolated data manifests worse performance and denotes extremely small FID distances w.r.t real signals, a sign of overspecialization.Significance.This study contributes a reproducible and complete framework for EEG signal generation and evaluation, addressing one of the main barriers to scalable AI application in BCI.

Objective.Investigation into complex neural circuits necessitates interfaces capable of high channel count recording and stimulation. However, existing commercial neural headstages often have limited scalability, restrictive proprietary designs, and constrained bidirectional capabilities, which worsens accessibility challenges and compels researchers to reinvent tools rather than build on a shared foundation.Approach.Here, we present two open-source, 128 channel headstages-Iris 128B and Iris 128S-designed for integration with microelectrode arrays. The Iris 128B enables fully bidirectional interfacing, with stimulation or recording across all 128 electrode channels, while the Iris 128S provides recording on 128 channels and stimulation on 16 simultaneous channels, which can be assigned to any 16 of the 32 available stimulation channels. Both designs use Intan Technologies' RHS and RHD series integrated circuits for amplification, filtering, digitization and stimulation, and are available on GitHub.Main results.The headstages were validated through benchtop impedance, noise, and frequency response measurements, as well as acutein vivorecordings in an anesthetized rat. Results demonstrate low noise levels and reliable signal acquisition across all channels.Significance.By releasing fully documented printed circuit board designs for headstages, this work aims to take a step towards broader adoption of bidirectional recording and stimulation systems while increasing channel counts. Future iterations will focus on miniaturization and wireless integration to improve usability in chronic and freely moving small animal experiments.

Objective. Sources of epileptic traveling waves offer critical insights into seizure onset zone (SOZ) localization, making them invaluable for preoperative assessment in patients with epilepsy. However, the absence of tailored source-tracing methods and the inherent instability of epileptiform activity make it difficult to achieve reliable source identification for SOZ localization. This study aimed to analyze the propagation pattern during seizure events and develop a framework to trace the sources of epileptic expanding traveling waves (ETWs).Approach. 101 seizure events were recorded from five 4-Aminopyridine-induced acute cortical rat epilepsy models. In each seizure event, epileptiform activities were classified into two categories according to their time-frequency diagrams (multiband and non-multiband epileptiform activities). The center of the SOZ was regarded as the recording site with the largest amplitude of epileptiform activities. Using the spatial-phase-based analysis, we analyzed the propagation pattern during the seizure event and extracted the ETWs. The sources of ETWs were traced by the intersection of spatial-phase-gradient.Main results. The ETW proportion of multiband epileptiform activities was 62.7%±8.3%, significantly higher than those in non-multiband epileptiform activities (53.8%±9.0%). ETWs with stable propagation patterns gave rise to a concentrated source tracing outcome. The single-band signal (component of the multiband activities) had a more stable ETW propagation pattern than both the multiband and non-multiband activities. The source tracing results of the single-band signals clustered around the SOZ center and remained stable even when the SOZ center was out of coverage (removing half of the recording sites, among which the SOZ center was included).Significance. The proposed framework enables ETW extraction from epileptiform activities and can trace ETW sources even when the sources are out of coverage. Therefore, the proposed framework may prove clinically valuable in cases with sparse intracranial recordings, addressing the limitation of traditional SOZ localization methods.

Objective.Electroencephalography (EEG) records the spontaneous electrical activity in the brain. Despite the growing application of deep learning in EEG decoding, traditional methods still rely heavily on supervised learning, which is often limited by task specificity and dataset dependency, restricting model performance and generalization. Inspired by the success of large language models, EEG foundation models (EEG FMs) are attracting increasing attention as a unified paradigm for EEG decoding. In this study, we review a selection of representative studies on EEG FMs, aiming to extract trends and provide recommendations for future research.Approach.We provide a comprehensive analysis of recent advances in EEG FMs, with a focus on downstream tasks, benchmark datasets, model architectures, and pre-training techniques. We analyze and synthesize core FMs components, and systematically compare their performances and generalizabilities.Main results.Our review reveals that EEG FMs are pre-trained on large-scale datasets, typically involving several hundred subjects. The number of subjects can reach up to 14 987, with a maximum total duration of 27 062 h. Current EEG FMs most adopt mask-based reconstruction pre-training strategy and employ efficient transformer-based architectures. Our comparative analysis shows that EEG FMs demonstrate significant potential in advancing EEG decoding tasks, particularly in seizure detection. However, their performance in complex scenarios such as motor imagery decoding remains limited.Significance.This review summarizes the existing approaches and performance outcomes of EEG FM, offers valuable insights into their current limitations and delineates prospective avenues for future research.

The treatment of glioblastoma (GBM) presents significant challenges, with median survival rates remaining low despite standard-of-care therapies. A novel approach, cytostatic hypothermia (CH), is under development against GBM; it is a window of temperature (typically 20 °C-25 °C) which halts tumor growthin vivo.Objective.This feasibility study expands upon the findings through the computational evaluation of a fully implantable system. Our simulations evaluate a thermoelectric cooler with a microwire array (NeuraTEC) and a novel ambient recirculating core (ARC) to achieve uniform cooling of a region in the brain without overheating local skin temperature.Approach.Finite-element modeling was employed to simulate coupled bioheat transfer and laminar non-isothermal fluid flow dynamics.Main results.Our results indicate that NeuraTEC can attain local tissue temperatures within a cytostatic range while minimizing thermal gradients. The use of multiple narrow, thermally conductive wires enhances cooling uniformity with minimal tissue displacement. The ARC provides a unique form of heat management that enables full implantability and hence portability. This work suggests it can facilitate the transfer of heat from a brain region to the skin. Future work will focus on device prototyping and validation throughin vitroandin vivostudies in large animal models.Significance.These simulations suggest that the proposed intracranial cooling system could make CH a practicable approach against GBM. Furthermore, this approach to internal heat management may also open new avenues for treating neurological conditions through local and chronic hypothermia, extending beyond the short-duration (acute) cooling methods currently tested.

Objective.In brain-computer interface (BCI) applications, signals are conventionally acquired from the cerebrum, and only a subset of the complex interactions that occur in several areas of the brain are collected. One area that has not been investigated for BCI application is the cerebellum, despite its involvement in movement and executive function. The present study aimed to determine the features of movement-related cerebellar electrocerebellography (ECeG) that are most useful for decoding, and how performance compares with conventional electroencephalography (EEG) recordings from the cerebrum.Approach.ECeG and EEG data were collected from six healthy adults to identify useful movement-related features from both cerebrum and cerebellum. Electromyography was used to capture the movements from the muscles. Decoding was conducted in binary movement vs. rest and movement vs. movement systems using support vector machines. Decoding performance was compared between cerebral, cerebellar, a combination of both, and temporal groups. Re-referencing techniques were applied to compensate for possible common reference artefacts or volume conduction effects.Main results. Movement-related features were decoded from over the cerebellum and the cerebrum. Classification accuracies were similar in both the cerebrum and cerebellum, when classifying movement vs. rest (cerebrum: 0.78 ± 0.02, cerebellum: 0.70 ± 0.01) and movement vs. movement states (cerebrum: 0.76 ± 0.02, cerebellum: 0.71 ± 0.02). The delta band (1-3 Hz) was the most useful feature for decoding.Significance.This study demonstrated, for the first time, that ECeG is a feasible source of movement related signals for implementing a BCI. The present study also demonstrated that the ECeG closely resembled the EEG signals and represents an alternate approach for BCI where the signal from the cerebrum is unreliable either due to disease or injury.