To date, brain-computer interfaces (BCIs) have not achieved reliable real-time communication through auditory or tactile modalities. Such interfaces would be crucial for brain-injured patients with severe motor impairments who are also blind or deaf. This study validates the functionality of the NeuroCommTrainer, a mobile and easy-to-use multimodal BCI with flex-printed electrode strips that does not require vision and adapts to users' attentiveness levels to initiate stimulation. In a study of 20 healthy participants, we evaluated auditory and vibrotactile oddball paradigms to train the system to differentiate rare and frequent event-related potentials (ERPs). In real-time online sessions, the system detected participants' mental focus to adaptively initiate stimulation through attentiveness monitoring. The NeuroCommTrainer successfully captured auditory and tactile ERPs, achieving a classification accuracy of 75% for stimuli in the calibration session, which is not yet reflected in the online session with 34% of found targets (chance level = 16.7%). The presented early-stage prototype of the NeuroCommTrainer requires several improvements before clinical application in brain-damaged patients, which include refined algorithms to reduce classification variance across participants, and enhanced attentiveness detection specifically tuned to brain activity of the targeted patient group. The present study makes a critical step in this direction and shows that a transition into a practicable communication system for brain-damaged patients may be achievable in the future.

Brain-computer interfaces (BCIs) translate brain activity into commands, enabling applications in communication, control, and neurorehabilitation. A major challenge in noninvasive BCIs is balancing classification performance with interpretability, as many approaches prioritize accuracy while overlooking the neural mechanisms underlying their predictions. In this study, we conduct a meta-analysis of feature interpretability across widely used methods in motor imagery (MI)-based BCIs, including power spectral density, common spatial patterns (CSP), Riemannian geometry, and functional connectivity. Specifically, we explore how network topology and spatial organization contribute to MI decoding by investigating brain network lateralization. Through evaluations on multiple EEG-based BCI datasets, our results confirm the superior classification performance of CSP and Riemannian methods. However, network lateralization provides stronger neurophysiological plausibility, revealing robust lateralization patterns in sensorimotor and frontal regions contralateral to imagined movements. These findings underscore the potential of connectivity-based features as a complementary tool for enhancing interpretability, supporting the development of more transparent and clinically relevant MI-based BCIs. Impact Statement This study addresses a critical gap in motor imagery-based brain-computer interfaces (BCIs) by systematically evaluating and comparing the interpretability of widely used methods, including power spectral density, common spatial pattern, Riemannian geometry, and functional connectivity. By analyzing these approaches across wide-ranging datasets, we offer valuable insights into the underlying neural mechanisms driving their performance. Our findings contribute to enhancing the transparency and biological relevance of BCI systems, ultimately advancing the development of more clinically meaningful and neurophysiologically interpretable BCIs.

Combining functional magnetic resonance imaging (fMRI) and machine learning (ML) can be used to identify therapeutic targets and evaluate the effect of repetitive transcranial magnetic stimulation (rTMS) in neural networks in tobacco use disorder. We investigated whether large-scale network connectivity can predict the rTMS effect on smoking cessation. Smoking cue exposure task-fMRI (T-fMRI) and resting-state fMRI (Rs-fMRI) scans were acquired before and after the 10 sessions of active or sham rTMS (10 Hz, 3000 pulses per session) over the left dorsal lateral prefrontal cortex in 42 treatment-seeking smokers. Five large-scale networks (default model network, central executive network, dorsal attention network, salience network [SN], and reward network) were compared before and after 10 sessions of rTMS, as well as between active and sham rTMS conditions. We performed neural network and regression analysis on the average connectivity of large-scale networks and the effectiveness of rTMS induced by rTMS. Regression analyses indicated higher salience connectivity in T-fMRI and lower reward connectivity in Rs-fMRI, predicting a better outcome of TMS treatment for smoking cessation ( < 0.01, Bonferroni corrected). Neural Network analyses suggested that SN was the most important predictor of rTMS effectiveness in both T-fMRI (0.33 of feature importance) and Rs-fMRI (0.37 feature importance). Both T-fMRI and Rs-fMRI connectivity in SN predict a better outcome of TMS treatment for smoking cessation, but in opposite directions. The work shows that ML models can be used to target TMS treatment. Given the small sample size, all ML findings should be replicated in a larger cohort to ensure their validity.

Time-resolved functional network connectivity (trFNC) assesses the time-resolved coupling between brain regions using functional magnetic resonance imaging (fMRI) data. This study aims to compare two techniques used to estimate trFNC, to investigate their similarities and differences when applied to fMRI data. These techniques are the sliding window Pearson correlation (SWPC), an amplitude-based approach, and phase synchrony (PS), a phase-based technique. To accomplish our objective, we used resting-state fMRI data from the Human Connectome Project with 827 subjects [repetition time (TR): 0.7 sec] and the Function Biomedical Informatics Research Network with 311 subjects (TR: 2 sec), which included 151 schizophrenia (SZ) patients and 160 controls. Our simulations reveal distinct strengths in two connectivity methods: SWPC captures high-magnitude, low-frequency connectivity, whereas PS detects low-magnitude, high-frequency connectivity. Stronger correlations between SWPC and PS align with pronounced fMRI oscillations. For fMRI data, higher correlations between SWPC and PS occur with matched frequencies and smaller SWPC window sizes (∼30 sec), but larger windows (∼88 sec) sacrifice clinically relevant information. Both methods identify a SZ-associated brain network state but show different patterns: SWPC highlights low anticorrelations between visual, subcortical, auditory, and sensory-motor networks, whereas PS shows reduced positive synchronization among these networks. In sum, our findings underscore the complementary nature of SWPC and PS, elucidating their respective strengths and limitations without implying the superiority of one over the other.

Brain-computer interfaces (BCIs) can provide alternative, motor-independent means of communication for people who have lost motor function. A promising variant is the functional magnetic resonance imaging (fMRI)-based BCI, which exploits information on hemodynamic brain activity evoked by performing different mental tasks. However, due to the sluggish nature of the hemodynamic response, a current challenge is to make these BCIs as efficient and fast as possible to allow useful clinical application. Furthermore, there is yet no consensus on optimal mental-task selection for multi-voxel pattern analysis-based decoding, nor whether certain tasks generalize well across users, or if individualized task selection would yield a higher decoding accuracy. To increase BCI efficiency, we tested whether distributed patterns of 3T-fMRI brain activation evoked by two-second mental tasks could be reliably discriminated in 2- to 7-class classification. In addition, we identified optimal mental-task combinations for high-accuracy classification across all classes. Finally, we examined whether individualized task selection-based on subjects' previous decoding performance ( tasks) or their subjective preference ()-was superior to the other in a yes/no communication paradigm. The 2-class decoding resulted in a mean accuracy of 78% and 3- to 7-class accuracies were above chance level. Mental calculation and spatial navigation were most frequently associated with the highest decoding accuracy. Furthermore, subjects could encode yes/no answers using their and tasks with mean accuracies of 83% and 81%, respectively. This implies that this paradigm, using short encoding durations, is well-suited to the diversity of patients and could greatly increase BCI efficiency.

Numerous neuroimaging studies have revealed abnormal brain function in patients with herpes zoster (HZ) and postherpetic neuralgia (PHN). However, few studies have focused on the alterations of intrinsic degree centrality (DC) in the transition process from the HZ to the PHN. Resting-state functional MRI (rs-fMRI) data from 27 patients with PHN, 24 patients with HZ, and 21 healthy controls (HCs) were acquired. DC based on rs-fMRI was used to explore specific brain functional abnormalities in these participants. Compared with HCs, patients with HZ presented decreased DC values in the right superior frontal gyrus, right cingulate gyrus, bilateral inferior parietal lobule (IPL), bilateral precuneus, and right paracentral lobule. Compared with HCs, patients with PHN also exhibited decreased DC values in the bilateral IPL. However, no regions with significant DC value changes were found between the HZ and PHN groups. These results suggest that decreased DC of the IPL is associated with the underlying neural mechanisms of the HZ and PHN stages and may represent a potential biomarker or intervention target candidate that needs further longitudinal confirmation.

Advancements in brain-computer interfaces (BCIs) have improved real-time neural signal decoding, enabling adaptive closed-loop neuromodulation. These systems dynamically adjust stimulation parameters based on neural biomarkers, enhancing treatment precision and adaptability. However, existing neuromodulation frameworks often depend on high-power computational platforms, limiting their feasibility for portable, real-time applications. We propose RONDO (Recursive Online Neural DecOding), a resource-efficient neural decoding framework that employs dynamic updating schemes in online learning with recurrent neural networks (RNNs). RONDO supports simple RNNs, long short-term memory networks, and gated recurrent units, allowing flexible adaptation to different signal type, accuracy, and real-time constraints. Experimental results show that RONDO's adaptive model updating improves neural decoding accuracy by 35% to 45% compared to offline learning. Additionally, RONDO operates within real-time constraints of neuroimaging devices without requiring cloud-based or high-performance computing. Its dynamic updating scheme ensures high accuracy with minimal updates, improving energy efficiency and robustness in resource-limited settings. RONDO presents a scalable, adaptive, and energy-efficient solution for real-time closed-loop neuromodulation, eliminating reliance on cloud computing. Its flexibility makes it a promising tool for clinical and research applications, advancing personalized neurostimulation and adaptive BCIs.

Tinnitus is a neurological condition characterized by the perception of ringing or other phantom sounds in the absence of external auditory stimuli. It affects an estimated 10%-15% of adults worldwide and can significantly affect sleep and mood. Neuroimaging techniques, particularly functional Magnetic Resonance Imaging (fMRI), have been widely used to investigate the auditory system and brain networks in tinnitus. Resting-state fMRI (rs-fMRI), a noninvasive approach, is particularly effective in examining spontaneous neural activity and functional connectivity (FC) across brain regions. This study investigated alterations in FC in individuals with chronic, non-bothersome tinnitus due to acoustic trauma using both static FC (sFC) and dynamic FC (dFC) analyses. A constrained independent component analysis was applied to identify five resting-state networks across the 23 regions of interest. sFC analysis revealed increased connectivity between the posterior cingulate cortex (a key region in the default mode network) and left angular gyrus (in the executive control network) in the tinnitus group. The dFC analysis showed that patients with tinnitus spent significantly more time in a weakly connected state, whereas healthy controls predominantly occupied a more segregated and strongly connected state. Findings suggest reduced network differentiation and altered temporal stability in individuals with non-bothersome tinnitus, potentially influenced by hearing loss. These alterations in both static and dynamic FC patterns provide insights into the neural underpinnings of tinnitus and its interaction with large-scale brain networks.

To explore brain function differences between patients with residual dizziness (RD) caused by benign paroxysmal positional vertigo (BPPV) and persistent postural-perceptual dizziness (PPPD) with resting-state functional magnetic resonance imaging. Using the Data Processing and Analysis for Brain Imaging software to analyze differences in the amplitude of low-frequency fluctuations (ALFF) and functional connectivity (FC) among RD, PPPD, and healthy controls groups. Then constructed a brain network and compared FC within the network. Further evaluated the correlation between abnormal brain regions and clinical characteristics. (1) Analysis of clinical characteristics: dizziness handicap inventory (DHI) scores differed between RD and PPPD groups. (2) Comparison of ALFF: RD group exhibited increased ALFF values in the right postcentral gyrus, right superior occipital gyrus, and right angular gyrus compared with the PPPD group. (3) Comparison of FC: the PPPD group exhibited weakened FC between the right cerebellum 8 region and right cerebellum crus1 region compared with the RD group. (4) Brain network analysis: Compared with the RD group, the PPPD group exhibited significantly reduced FC between the left supramarginal gyrus and the right angular gyrus. (5) Correlation analysis: The DHI scale scores of PPPD group were positively correlated with ALFF values of the right angular gyrus and negatively correlated with FC values between the right cerebellum 8 region and right cerebellum crus1 region. Significant differences in brain functional activity were observed between RD and PPPD patients, which reveals that there are differences between RD and PPPD patients regarding neural mechanisms in the process of pathogenesis.

Autism spectrum disorder (ASD) is a neurodevelopmental condition that exhibits a wide range of clinical heterogeneity. This study aimed to explore the heterogeneity of ASD based on deviations in brain functional networks. Resting-state functional magnetic resonance imaging data from the Autism Brain Imaging Data Exchange database were analyzed in 105 children with ASD and 102 demographically matched typical controls (TC) children. Heterogeneity through discriminative analysis (HYDRA) was utilized to identify subtypes of ASD based on the degree centrality (DC) maps. Voxel-wise group comparisons were then performed between ASD subtypes and the TC group. The relationship between the altered DC and the symptom severity was finally analyzed for ASD subtypes using the multivariate support vector regression approach. HYDRA identified three subtypes of ASD. Distinct DC alteration patterns were observed in brain regions including the fusiform gyrus, insula, and inferior frontal gyrus in ASD subtypes. Moreover, the altered DC values for ASD subtype 1 and subtype 3 can predict the restricted and repetitive behavior and social communication impairments in ASD, respectively. Our findings demonstrated the heterogeneity of brain functional networks in ASD and provided a promising way to explain the high heterogeneity of clinical symptoms and outcomes.

Executive functions (EF) are cognitive processes supporting language and reading. Children with dyslexia show reading difficulties primarily due to phonological processing, with additional reported deficits in EF. This study aimed to determine the differences in EF involvement during written (reading) versus oral language (narrative) comprehension in children with dyslexia versus typical readers neurobiologically and behaviorally. Reading, language, and EF behavioral measures and functional MRI data were collected from 55 typical readers (TR) and 65 English-speaking children with dyslexia ages 8-12 years during reading and narrative comprehension tasks. Differences within and between functional connectivity of EF and attention networks were calculated and then compared between groups and tasks using Fisher Z-transformation. Children with dyslexia showed higher functional connectivity values in EF and attention networks in both reading and narrative comprehension tasks, whereas TR showed higher functional connectivity in narrative versus reading comprehension. Within groups, analysis showed higher functional connectivity within dorsal attention functional brain network (DAN) and between DAN-fronto-parietal (FP), cingulo-opercular (CO)-FP, and ventral attention functional brain network (VAN)-DAN, in the reading versus narrative comprehension task in the dyslexia group. TR showed higher functional connectivity within VAN, and between VAN-FP in the narrative compared to the reading comprehension tasks. Children with dyslexia seem to greatly utilize EF and attention-related networks in narrative and reading comprehension tasks and demonstrate a greater network integration for the written versus oral comprehension task. TR, however, utilize these networks only during oral comprehension, which may point to a greater reliance on memory and processing effort in the absence of written information.

Emerging video services (EVS) offer users various multimedia presentations, and satisfaction assessment is crucial for enhancing their user experience and competitiveness. However, existing research methods are unable to provide a quantitative satisfaction assessment. Electroencephalogram (EEG), as a popular signal source in brain-computer interface (BCI), with the advantage of being difficult to disguise and containing rich brain activity information, has gained increasing attention from researchers. This article aims to investigate the advantages of employing EEG for modeling satisfaction in EVS. Unlike the subjective metrics assessment in traditional video services, generating satisfaction in EVS involves a range of cognitive functions, including cognitive load, emotion, and audiovisual perception, which are difficult to characterize using a single feature. The representation of brain states for complex cognitive functions has been a major challenge for EEG modeling approaches. To address this challenge, we propose an EEG-based EVS satisfaction assessment BCI by raising a Point-to-Global graph representation learning strategy (P2G) that efficiently identifies satisfaction level through a parallel coding module and a graph-based brain region perception module. P2G captures satisfaction-sensitive graph representations in EEG samples based on coding and integrating point features and the global topography. We validate the effectiveness of introducing a P2G learning strategy in EVS satisfaction modeling using a self-constructed dataset and a relevant public dataset, and our method outperforms existing methods. Additionally, we provide a detailed visual analysis to unveil neural markers associated with EVS satisfaction, thereby laying a scientific foundation for the optimization and development of video services.

Ultrasound is a promising new approach for noninvasive brain stimulation. Low-intensity focused ultrasound (LIFU) allows targeting the deep brain with high spatial and temporal resolution. For clinical use, ultrasound systems must fulfill specific requirements. Three-dimensional (3D) steering and focusing either requires mechanical displacement of (focused) transducers or multielement arrays and corresponding multichannel electronics. Since the waveform has an impact of the induced neurostimulation effect, electronics need sufficient flexibility for generating arbitrary temporal signal patterns. For compensation of skull aberration artifacts, elements must be excited with defined phase resulting of phase aberration correction (PAC) algorithms. Finally, for being clinically usable, systems must be combined with planning hardware and software. A versatile system for 3D steered LIFU based on two-dimensional matrix arrays was designed, fabricated, and characterized in terms of focusing, steering, and output of temporal patterns. Our PAC algorithm was validated on an skull. The system was tested for compliance with defined medical device standard by accredited laboratories, and an initial Magnetic resonance imaging (MRI) phantom study was performed. Our system allows 3D beam steering and focusing with lateral focus sizes down to 4 mm, which is less than the size of a human gyrus, such that detailed targeting is possible. Arbitrary temporal signal patterns (different wave forms, pulse length, duty cycle, and ramping) were generated. Different software interfaces allow patient-specific planning with a Magnetic resonance Tomograph (MR)- or neuronavigation-based workflow, in which a custom-developed PAC algorithm allows compensation of the skull bone. The absence of transducer susceptibility artifacts was shown in the MRI phantom study, and the acoustic focus was localized using magnetic resonance acoustic radiation force imaging. Our new versatile ultrasound neuromodulation platform represents a compromise between conformal helmet-like systems and single element transducer setups. It is flexible in terms of spatiotemporal stimulation patterns and can be accommodated to different workflows. Impact Statement Progress in the field of ultrasound neurostimulation is depending on the availability of suitable hardware fulfilling a range of practical, technical, safety, and regulatory requirements. Systems must fit in established clinical workflows (e.g., usable with MR and/or neuronavigation systems), allow accessing deep brain regions, and generate defined spatiotemporal ultrasound patterns. Furthermore, basic regulatory constraints (e.g., IEC 60601-1) must be fulfilled. Our new low-intensity focused ultrasound (LIFU) system addresses these requirements and is flexible enough for use in a research environment. It was developed for facilitating the clinical transfer of LIFU and helping to gain a better understanding of underlying effects in ultrasound neurostimulation.

Paroxysmal depolarization shifts (PDSs), correlated with interictal epileptiform discharges, involve significant membrane potential changes and action potentials. While synchronicity is crucial in paroxysmal activity, the precise function of PDSs and their propagation mechanisms, especially non-synaptic pathways like ephaptic coupling, remains poorly understood. This study investigates the role of ephaptic coupling in PDS propagation in hippocampal cultures, focusing on voltage-gated calcium channel (VGCC) subtypes. PDSs were induced in hippocampal neurone-glial cultures using bicuculline. The outside-out patch-clamp technique was used to record PDS activity at varying distances from the neuronal network. The effects of L-type (nifedipine) and T-type (ML-218) VGCC inhibitors on PDS amplitude and frequency were assessed. Membrane capacitance and resistance were monitored to verify the outside-out configuration. PDSs could be recorded up to 16 µm from the network, with amplitude decreasing exponentially with distance. PDS frequency remained constant. Blocking L-type VGCCs completely abolished PDS activity at a distance, while T-type VGCC inhibition significantly reduced PDS amplitude. The transition from whole-cell to outside-out configuration was confirmed by a significant decrease in membrane capacitance. The findings suggest that ephaptic coupling contributes to PDS propagation , with L-type VGCCs playing a critical role in field-mediated signal transmission. Constant PDS frequency with varying amplitude at a distance highlights a potential synchronization mechanism during epileptiform activity. Further research should investigate the interplay between ion channels and the extracellular environment during ephaptic coupling, paving the way for brain stimulation-based therapies. Research demonstrates that ephaptic coupling can propagate PDSs in hippocampal neurone-glial cultures, highlighting a promising mechanism for understanding epileptiform foci. This finding is critical for comprehending how these foci form and expand, and it also opens avenues for developing brain stimulation-based therapies.

As the cerebellum has reciprocal communications with the frontal cortex, this retrospective cohort study examined the effects of dual-site repetitive transcranial magnetic stimulation (ds-rTMS: dorsolateral prefrontal cortex [DLPFC] + cerebellum) in disorders of consciousness (DoC). Single-center study in the Department of Rehabilitation of Jinhua Hospital of TCM Affiliated to Zhejiang University of Traditional Chinese Medicine. Twenty-nine patients with DoC. Systematic review of clinical records comparing ds-TMS (DLPFC + cerebellum) with conventional single-site DLPFC-rTMS. Coma Recovery Scale-Revised (CRS-R) scores, mismatch negativity (MMN) latency, P300 latency, Judson grade, and Hall grade. ds-TMS was associated with larger gains in consciousness (CRS-R scores) compared with DLPFC-rTMS in a retrospective cohort. Both interventions had comparable improvement in cognitive and somatosensory outcomes (MMN, P300, and Judson/Hall grades). Higher CRS-R scores correlated with shorter MMN latency and better Hall grades. ds-TMS treatment may represent an effective therapeutic approach for DoC, with potential effects on consciousness recovery.

Here we aim to search for stable intra- and inter-band cross-correlations during the peri-ictal transition of focal onset seizures. Furthermore, we search for dynamic features by analyzing relative eigenvalues of the cross-correlation matrix. In this study, we analyze 50 extracranial electroencephalographic recordings from 24 patients with different types of focal epilepsy, separating the data into different frequency bands. Thereby we construct a multiband cross-correlation matrix, evaluate stability of the correlation structures and the time evolution of relative eigenvalues using a running window approach. We find a consistent, pronounced average cross-correlation pattern that is independent of the physiological state, is subject-independent, and is highly similar across different frequency bands. In contrast, dynamic features of brain activity are encoded in deviations from this baseline pattern, expressed by relative eigenvalues along the whole spectrum. We associate the stable background pattern as the dynamics upon (or close to) the attractor dynamics, necessary to maintain the brain in an efficient operational mode. Transient dynamical features are expressed by temporal deviations from this pattern. Our results are congruent with the hypothesis that the brain is a complex system operating close to a critical point of a phase transition.

Chemotherapy-related cognitive impairment (CRCI), commonly known as "chemobrain," frequently occurs during breast cancer treatment and has been linked to altered brain function. This resting-state functional magnetic resonance imaging study examined chemotherapy-related changes in functional brain activity, network connectivity, and associations with cognitive outcomes. Twenty-eight patients with breast cancer were assessed prechemotherapy (BB) and postchemotherapy (BBF), alongside 27 healthy controls of comparable age at baseline (BH) and follow-up (BHF). Mean fractional amplitude of low-frequency fluctuations (mfALFF) and mean regional homogeneity (mReHo) quantified functional brain activity. Graph theoretical analysis (GTA) assessed network topology; network-based statistics (NBS) evaluated interregional connectivity. Cognitive performance was evaluated through standardized assessments. Postchemotherapy patients exhibited reduced anxiety and lower FACT-Cog scores. Voxel-wise analyses showed increased mfALFF in frontal regions and mReHo in superior temporal and inferior frontal gyri, alongside decreases in postcentral, lingual, and parahippocampal areas. Healthy controls showed increased activity in medial frontal and cingulate regions, with reductions in the temporal lobe and putamen. GTA revealed higher global efficiency and reduced modularity, path length, and network complexity in the BBF group compared with BHF. NBS showed weaker structural connectivity in motor and occipital regions prechemotherapy and decreased parietal and insular connectivity postchemotherapy. Multiple regression showed brain-behavior correlations: declines in FACT-Cog, Digit Symbol Substitution, and mood scores were linked to altered activity in frontal, parietal, cingulate, and occipital areas, while positive correlations suggested compensatory activation. Chemotherapy was associated with longitudinal alterations in brain activity, network organization, and connectivity in breast cancer survivors. Brain-behavior associations suggest disrupted neural networks may underlie CRCI.