Front Comput Neurosci [JOURNAL]
Sun
200 papers
Information flow drives localized morphological differences across neuronal and glial cell types.
Desai-Chowdhry P, Brummer AB, Mallavarapu S … +2 more , Oakes M, Savage VM
Front Comput Neurosci · 2026 · PMID 41889601 · Full text
Neuron processes-axons and dendrites-have distinct branching patterns related to their biological function in the brain and body. Other non-neuronal cells in the nervous system, glia, also have characteristic branching m...
Neuron processes-axons and dendrites-have distinct branching patterns related to their biological function in the brain and body. Other non-neuronal cells in the nervous system, glia, also have characteristic branching morphologies. Our previous work has used biological scaling theory to connect branching patterns in neurons to biophysical function such as energy or conduction time minimization and material constrants in a compact, unifying mathematical model. Here, we use functionally relevant structural parameters related to asymmetric branching patterns extracted from our model as features in machine-learning classification methods to highlight differences between different types of neurons and glia as well as between healthy and diseased cells. Notably, we find that parameters related to information flow vary with position in the cell-that is, relative proximity of each branching junction to the soma (cell body) or synapses. We find that for some neuronal and glial cell type comparisons, such as comparisons between medium spiny neuron (MSN) dendrites, incorporating relative branching junction location significantly improves the performance of machine-learning classification methods. Our results imply that differences in information flow across cells drive specific morphological changes that correspond to localized regions of neuronal and glial cells. The promise of our methods and results lay foundation for future studies classifying neuronal and glial cells based on pathology, using our asymmetric scale factors and relative branching junction location as potential biomarkers to identify particular diseases based on both structural differences and the underlying differences in function.
Role of spinal sensorimotor circuits in triphasic muscle command: a simulation approach using goal exploration process.
Cattaert D, Guemann M, Paclet F … +4 more , Lemarchand L, Chung B, Oudeyer PY, de Rugy A
Front Comput Neurosci · 2026 · PMID 41889600 · Full text
During rapid voluntary elbow movement on horizontal plane, a stereotyped triphasic pattern is typically observed in the electromyograms (EMGs) of antagonistic muscles acting at this joint. To explain the origin of such t...
During rapid voluntary elbow movement on horizontal plane, a stereotyped triphasic pattern is typically observed in the electromyograms (EMGs) of antagonistic muscles acting at this joint. To explain the origin of such triphasic commands, two types of theories have been proposed. Peripheral theories consider that triphasic commands result from sensorimotor spinal networks, either through a combination of reflexes or through a spinal central pattern generator. Central theories consider that the triphasic command is elaborated in the brain. Although both theories were partially supported by physiological data, there is still no consensus about how exactly triphasic commands are elaborated. Moreover, capacities of simple spinal sensorimotor circuits to elaborate triphasic commands on their own have not been tested yet. In order to test this, we modelled arm musculoskeletal system operating in the absence of gravity, muscle activation dynamics, proprioceptive spindle and Golgi afferent activities and spinal sensorimotor circuits. Step commands were designed to modify the activity of spinal neurons and the strength of their synapses, either to prepare (SET) the network before movement onset, or to launch the movement (GO). Since these step commands do not contain any dynamics, changes in muscle activities responsible for arm movement rest entirely upon interactions between the spinal network and the musculoskeletal system. Critically, we selected step commands using a Goal Exploration Process inspired from baby babbling during development. In this task, the Goal Exploration Process proved very efficient at discovering step commands that enabled spinal circuits to handle a broad spectrum of functional behaviors, displayed in a behavioral space characterized by movement amplitude and maximal speed. All over the behavioral space, specific SET and GO commands elicited natural triphasic commands, thereby substantiating the inherent capacity of the spinal network in generating them.
Editorial: Unraveling information encoding and representation in memory formation and learning.
Montani F
Front Comput Neurosci · 2026 · PMID 41889599 · Full text
Abstract loading — click title to view on PubMed.
Optimized facial landmark modeling with medical aesthetic constraints by a multi-objective genetic algorithm.
Ye Y, Yan G, Wen D … +1 more , Tan M
Front Comput Neurosci · 2026 · PMID 41821527 · Full text
"Facial Beauty" is not an absolute physical attribute but a subjective social and cultural construct. Facial beauty assessment is an interdisciplinary field that integrates computer vision and medical aesthetics (MAs) to...
"Facial Beauty" is not an absolute physical attribute but a subjective social and cultural construct. Facial beauty assessment is an interdisciplinary field that integrates computer vision and medical aesthetics (MAs) to quantify personal judgment regarding facial attractiveness. In this study, the beauty assessment we adopted was based on the scores given by plastic surgeons; this method is more professional and is supported by a theoretical basis. We derived a set of MA features that encompass global traits, local details, and curvature aspects from established aesthetic principles. Incorporating these features enhances predictive accuracy in facial beauty. Furthermore, we propose a feature selection algorithm with aesthetic-driven initialization embedded in a multi-objective evolutionary framework. Additionally, we introduce an MA facial landmark model that provides explicit annotation of bilateral zygomatic, orbital, and nasal points for precise attractiveness scoring. Experimental results on the South China University of Technology-Facial Beauty Perception (SCUT-FBP) and SCUT-FBP5500 datasets and the Chicago Face Dataset demonstrate superior performance (Pearson's correlation coefficient = 0.8216, mean absolute error = 0.2638, and root mean square error = 0.3743) over state-of-the-art methods, validating its clinical relevance. This study provides a practical tool for beauty evaluation, where the selected features align with professional judgments, enabling transparent and explainable outcomes in both clinical and cosmetic applications.
Population-level neural rejuvenation dynamics in addiction: a computational framework for understanding developmental plasticity reactivation.
Borjkhani M, Borjkhani H, Sharif MA
Front Comput Neurosci · 2026 · PMID 41810357 · Full text
BACKGROUND: The neural rejuvenation hypothesis proposes that drugs of abuse reactivate developmental plasticity mechanisms to create abnormally persistent addiction memories. While individual molecular components have be...
BACKGROUND: The neural rejuvenation hypothesis proposes that drugs of abuse reactivate developmental plasticity mechanisms to create abnormally persistent addiction memories. While individual molecular components have been characterized experimentally, the population-level dynamics and their collective contribution to addiction pathophysiology remain poorly understood. OBJECTIVES: To develop a computational framework tracking theoretical synaptic population dynamics during simulated drug exposure and withdrawal, and to demonstrate how coordinated population-level transitions could account for key experimental observations in addiction neuroscience. METHODS: We constructed a mathematical model tracking four theoretical synaptic populations (adult, juvenile, silent, and matured synapses) using differential equations. The model incorporates two distinct processes: (1) rejuvenation of existing synapses through receptor composition switching, and (2) generation of silent synapses during drug exposure. Critically, the total synapse population is dynamic, increasing during drug exposure due to synaptogenesis and decreasing during withdrawal due to pruning. State transitions are explicitly phase-gated: silent synapse generation occurs only during exposure, while maturation and pruning occur predominantly during withdrawal. Rate constants were derived from experimental time scales reported in the literature, with explicit biological time mapping (1 time unit = 2 h). Simulations involved five intermittent exposures followed by extended withdrawal, with comprehensive parameter sensitivity analysis to assess model robustness across ±50% parameter variations. Initial conditions were fixed to represent the experimentally motivated baseline (adult synapses only); alternative initial states were also tested and did not change qualitative conclusions. RESULTS: The model demonstrated coordinated synaptic population transformations that qualitatively paralleled experimental observations. In simulation, results revealed distinct phases of neural rejuvenation with characteristic population dynamics: adult-to-juvenile conversion during exposure (reaching ~500 juvenile synapses in the model), silent synapse generation (~400 synapses), and progressive maturation during withdrawal (~300 matured synapses). The modeled total synapse population increased dynamically from baseline (1,000) to ~1,400 during exposure due to synaptogenesis, then decreased to ~1,300 during withdrawal due to pruning. NMDA receptor composition shifted from 80% GluN2A to 80% GluN2B during simulated exposure. Memory strength increased continuously through biphasic mechanisms: during exposure, memory formation was driven by enhanced plasticity capacity; during withdrawal, memory strengthening was driven by the maturation flux (the rate of CP-AMPAR recruitment into silent synapses), with saturation preventing unbounded growth. Parameter sensitivity analysis demonstrated robust qualitative behavior across ±50% parameter variations. Comparative simulations with natural rewards (modeled with = 0) showed minimal rejuvenation effects and attenuated incubation, consistent with experimental observations of drug specificity. CONCLUSION: This computational framework demonstrates how neural rejuvenation might operate as a population-level phenomenon, with sequential recruitment of different plasticity mechanisms creating robust addiction-related memories. The model generates testable hypotheses and provides a foundation for understanding potential therapeutic intervention windows targeting different phases of rejuvenation.
Towards precision medicine in Tourette syndrome: a perspective on AI-driven predictive modelling and personalised care.
Zhao C, Li R, Hua L … +3 more , Li H, Zhang M, Wang B
Front Comput Neurosci · 2026 · PMID 41800283 · Full text
Tourette Syndrome (TS) is a complex neurodevelopmental disorder characterised by motor and vocal tics that significantly impair quality of life. Conventional diagnostic and therapeutic methods face challenges due to subj...
Tourette Syndrome (TS) is a complex neurodevelopmental disorder characterised by motor and vocal tics that significantly impair quality of life. Conventional diagnostic and therapeutic methods face challenges due to subjectivity, lack of personalisation, and difficulties in prognostic prediction. Artificial Intelligence (AI) offers novel solutions, advancing TS management towards precision medicine. This article presents a conceptual framework for AI-driven technologies in TS, advocating for a paradigm shift from empirical treatment to precision medicine. We discuss key components including predictive model construction, personalised diagnosis, treatment strategies, and intelligent monitoring. Research indicates that the core value of AI in TS precision medicine lies in its predictiveness, individualisation, and intelligence. Predictive models using multimodal data enable early identification and prognostic assessment. Furthermore, personalised approaches tailor diagnosis and treatment to individual patient characteristics, thereby improving outcomes. Intelligent systems enable automated monitoring and real-time adjustments, optimising clinical workflows. Substantial clinical evidence demonstrates that AI-driven precision medicine improves diagnostic accuracy, optimises treatments, and enhances patient prognosis. Despite this potential, challenges remain in data quality, algorithm interpretability, and clinical translation. Future efforts should focus on enhancing interdisciplinary collaboration, promoting standardisation, and facilitating clinical adoption to deliver more precise, effective, and accessible care for TS patients.
Extracellular stimulation and ephaptic coupling of neurons in a fully coupled finite element-based Extracellular-Membrane-Intracellular (EMI) model.
Jæger KH, Tveito A
Front Comput Neurosci · 2026 · PMID 41766775 · Full text
INTRODUCTION: The extracellular potential surrounding neurons is of great importance: it is measured to interpret neural activity, it underpins ephaptic coupling between neighboring cells, and it forms the basis for exte...
INTRODUCTION: The extracellular potential surrounding neurons is of great importance: it is measured to interpret neural activity, it underpins ephaptic coupling between neighboring cells, and it forms the basis for external stimulation of neural tissue. These phenomena have been studied for decades, both experimentally and computationally. In computational models, variants of the classical cable equation for membrane dynamics and an electrostatic equation for the extracellular field are the most common approaches. Such formulations however, typically decouple the governing equations and therefore neglect the bidirectional coupling between the extracellular (E) space, the cell membrane (M), and the intracellular (I) space. METHODS: We use a finite element-based Extracellular-Membrane-Intracellular (EMI) approach that solves a fully coupled system to study extracellular stimulation and ephaptic coupling in detailed models of cerebellar Purkinje neurons and neocortical layer 5 pyramidal neurons. We vary the distance to the stimulation source, the amplitude, and the frequency of an external current, and we simulate two-cell configurations to assess ephaptic spike-timing effects, synchronization, and the possibility of direct ephaptic action potential triggering. RESULTS: We find that weak sinusoidal stimulation induces subthreshold membrane oscillations that follow the stimulus frequency, and that constant or sinusoidal extracellular stimulation modulate spike rates and spike timing in a manner that depends on stimulation strength and distance. In two-cell simulations, we find that Purkinje neurons synchronize ephaptically in a distance-and extracellular-conductivity-dependent manner, and that pyramidal neuron spike timing is altered by a neighboring firing cell. Direct ephaptic triggering requires markedly reduced extracellular conductivity relative to bulk values. DISCUSSION: The results provide quantitative insight into extracellular field-mediated neural coupling and how externally applied fields, such as those used in deep brain stimulation, interact with single-neuron biophysics. The results support the view that ephaptic interactions between neurons are more plausibly expressed as spike-timing modulation and synchronization than as direct excitatory triggering under physiological conditions.
Mechanistic explanation of neuroplasticity using equivalent circuits.
Nilsson MNP
Front Comput Neurosci · 2026 · PMID 41766774 · Full text
INTRODUCTION: This paper presents a comprehensive mechanistic model of a neuron with plasticity that explains how information input as time-varying signals is processed and stored. Additionally, the model addresses two l...
INTRODUCTION: This paper presents a comprehensive mechanistic model of a neuron with plasticity that explains how information input as time-varying signals is processed and stored. Additionally, the model addresses two long-standing, specific biological challenges: Integrating Hebbian and homeostatic plasticity, and identifying a concise synaptic learning rule. METHOD: A biologically accurate small-signal equivalent-circuit model is derived through a one-to-one mapping from established ion-channel properties. The often-overlooked dynamics of the synaptic cleft is essential in this process. Analysis of the model reveals a simple and succinct learning rule, indicating that the neuron functions as an internal-feedback adaptive filter, a common concept in signal processing. RESULTS: Simulations confirm the model's functionality, stability, and convergence, demonstrating that even a single neuron without external feedback can act as a potent signal processor. The model replicates several key characteristics typical of biological neurons, which are seldom captured in other neuron models. It can encode time-varying functions, learn without risking instability, and bootstrap from a state where all synaptic weights are zero. DISCUSSION: This paper explores the function of neurons with a focus on biological accuracy, not computational efficiency. Unlike neuromorphic models, it does not aim to design devices. The electronic circuit analogy aids understanding by leveraging decades of electronics expertise but is not intended for physical implementation. This interdisciplinary work spans a broad range of subjects within the realm of neurobiophysics, including neurobiology, electronics, and signal processing.
Cross-subject mapping of neural activity with restricted Boltzmann machines.
Yang H, Angjelichinoski M, Wu S … +3 more , Putney J, Sponberg S, Tarokh V
Front Comput Neurosci · 2026 · PMID 41766773 · Full text
Subject-to-subject variability is a common challenge in generalizing neural data models across subjects, discriminating subject-specific and inter-subject features in large neural datasets, and engineering neural interfa...
Subject-to-subject variability is a common challenge in generalizing neural data models across subjects, discriminating subject-specific and inter-subject features in large neural datasets, and engineering neural interfaces with subject-specific tuning. While many methods exist that map one subject to another, it remains challenging to combine many subjects in a computationally efficient manner, especially with highly non-linear features such as populations of spiking neurons or motor units. Consider subjects with trained neural decoders as sources and those without as targets. Our objective is to transfer data from one or more target subjects to the domain of the source subjects to directly apply the source neural decoder such that no target decoder needs to be trained. We propose to use the Restricted Boltzmann Machine (RBM) with Gaussian inputs and Bernoulli hidden units; once trained over the entire feature set of subjects, the RBM allows the mapping of target features on source feature spaces using Gibbs sampling. We also consider a novel computationally efficient training technique for RBMs based on the Fisher divergence, which allows closed-form gradients of the RBM to be computed. We apply our methods to decode turning behaviors from neuromuscular recordings of spike trains from the ten muscles that primarily control wing motion in an agile flying hawk moth, . The dataset consists of this comprehensive motor program recorded from nine subjects, each driven by six discrete visual stimuli. The evaluations show that the target features can be decoded using the source classifier to classify the visual stimuli with an accuracy of up to 95% when mapped using an RBM trained by Fisher divergence, suggesting that RBMs for multi-cross-subject mapping applications are effective and efficient.
Girls just wanna have funds: a new Transparent Reporting Scale for evaluating grant data reporting from funding agencies.
Clarke N, Licata AE, Imarraine S … +7 more , Dao T, Sperandio G, Pinho AL, Borghesani V, Mengotti P, Liuzzi AG, Pischedda D
Front Comput Neurosci · 2026 · PMID 41766772 · Full text
INTRODUCTION: Despite the increasing representation of women in scientific fields, disparities in research funding allocation remain. This inequity deprives talented women researchers of necessary resources, limiting the...
INTRODUCTION: Despite the increasing representation of women in scientific fields, disparities in research funding allocation remain. This inequity deprives talented women researchers of necessary resources, limiting the diversity of perspectives and ideas, and contributes to the "scissor-shaped curve" seen in neuroscience, where women leave before obtaining senior positions. Data transparency and comprehensive reporting of information on grant winners and applicants, as well as reporting of gender and other intersecting demographics and key metrics, are crucial to effectively evaluate funding equity. However, there is a lack of guidelines on which data funders should report. In this study, we aimed to investigate the transparency of neuroscience funders across Europe, focusing on the European Union, Schengen area, and the United Kingdom. METHODS: To this end, we developed a Transparent Reporting Scale (TRS), composed of 15 items crucial to facilitate transparent and meaningful reporting, and searched for public data from funders in order to apply the scale and evaluate their transparency in data reporting. Across 32 countries and the European Union as a whole, we identified 39 funders, with 90% sharing publicly available data on funding results. RESULTS: Using the TRS, five funders received a "gold" rating, eighteen a "silver" one, and thirteen a "bronze" rating. Scale scores were significantly correlated with the Gender Equality Index [ = 0.64, 95% CI (0.33, 0.83), = 0.001] and gross domestic product of the countries where funders are based [ = 0.51, 95% CI (0.20, 0.74), = 0.003], suggesting that collection and/or publication of funding data may reflect overall commitments to gender equity, and be limited due to resources. Data from only 29% of funders could be disaggregated for the neuroscience category specifically, indicating the difficulty in evaluating equity in our field. DISCUSSION: We collated all available data into an Open Science Framework repository to enable data sharing and further analyses. The TRS can support funders in adopting transparent, standardized reporting practices in order to support evidence-based progress toward gender equity.
Editorial: Neuromorphic and deep learning paradigms for neural data interpretation and computational neuroscience.
Zou C, Yuan R, Wen J
Front Comput Neurosci · 2026 · PMID 41766771 · Full text
Abstract loading — click title to view on PubMed.
Tension shapes memory: computational insights into neural plasticity.
Lee KY, Saif MTA
Front Comput Neurosci · 2026 · PMID 41756310 · Full text
Mechanical forces have recently emerged as critical modulators of neural communication, yet their role in high-level cognitive functions remains poorly understood. Here, we present a biologically inspired spiking neural...
Mechanical forces have recently emerged as critical modulators of neural communication, yet their role in high-level cognitive functions remains poorly understood. Here, we present a biologically inspired spiking neural network model that integrates mechanical tension, vesicle dynamics, and spike-timing-dependent plasticity to examine how tension influences learning, memory, and cognitive operations such as pattern completion, projection, and association. We find that increased tension enhances synaptic efficiency by accelerating vesicle clustering and recovery, resulting in a 67% improvement in memory recall speed and a 17% increase in inter-regional synchrony during projection relative to relaxed states. Conversely, a 20% reduction in tension leads to a 31% decline in memory association performance, highlighting the tension-sensitive accessibility of stored information. The model further reveals that an appropriate balance of inhibition is essential for these tension-driven effects: networks with 20% inhibitory neurons achieve optimal spatial precision in memory encoding and recall, whereas insufficient inhibition allows tension-amplified excitation to spread uncontrollably and degrade recall fidelity. Together, these findings position mechanical tension as a functional neuromodulator and suggest new directions for neuromorphic design and energy-efficient, living computing platforms.
Shifts in brain dynamics and drivers of consciousness state transitions.
Bodenheimer J, Bogdan P, Pequito S … +1 more , Ashourvan A
Front Comput Neurosci · 2026 · PMID 41743844 · Full text
Understanding the neural mechanisms underlying the transitions between different states of consciousness is a fundamental challenge in neuroscience. Thus, we investigate the underlying drivers of changes during the resti...
Understanding the neural mechanisms underlying the transitions between different states of consciousness is a fundamental challenge in neuroscience. Thus, we investigate the underlying drivers of changes during the resting-state dynamics of the human brain, as captured by functional magnetic resonance imaging (fMRI) across varying levels of consciousness (awake, light sedation, deep sedation, and recovery). We deploy a model-based approach relying on linear time-invariant (LTI) dynamical systems under unknown inputs (UI). Our findings reveal distinct changes in the spectral profile of brain dynamics-particularly regarding the stability and frequency of the system's oscillatory modes during transitions between consciousness states. These models further enable us to identify external drivers influencing large-scale brain activity during naturalistic auditory stimulation. Our findings suggest that these identified inputs delineate how stimulus-induced co-activity propagation differs across consciousness states. Notably, our approach showcases the effectiveness of LTI models under UI in capturing large-scale brain dynamic changes and drivers in complex paradigms, such as naturalistic stimulation, which are not conducive to conventional general linear model analysis. Importantly, our findings shed light on how brain-wide dynamics and drivers evolve as the brain transitions toward conscious states, holding promise for developing more accurate biomarkers of consciousness recovery in disorders of consciousness.
Synergy mediates long-range correlations in the visual cortex near criticality.
Rajpal H, Stefens C, Saeedian M … +6 more , Canzano JS, Kareithi MG, Barahona M, Smith SL, Schultz SR, Jensen HJ
Front Comput Neurosci · 2026 · PMID 41728358 · Full text
Long-range correlations are a key signature of systems operating near criticality, indicating spatially-extended interactions across large distances. These extended dependencies underlie other emergent properties of crit...
Long-range correlations are a key signature of systems operating near criticality, indicating spatially-extended interactions across large distances. These extended dependencies underlie other emergent properties of critical dynamics, such as high susceptibility and multi-scale coordination. In the brain, along with other signatures of criticality, long-range correlations have been observed across various spatial scales, suggesting that the brain may operate near a critical point to optimize information processing and adaptability. However, the mechanisms underlying these long-range correlations remain poorly understood. Here, we investigate the role of synergistic interactions in mediating long-range correlations in the visual cortex of awake mice. We leverage recent advances in mesoscale two-photon calcium imaging to analyse the activity of thousands of neurons across a wide field of view, allowing us to confirm the presence of long-range correlations at the level of neuronal populations. By applying the Partial Information Decomposition (PID) framework, we decompose the correlations into synergistic and redundant information interactions. Our results reveal that the increase in long-range correlations during visual stimulation is accompanied by a significant increase in synergistic rather than redundant interactions among neurons. Furthermore, we analyse a combined network formed by the union of synergistic and redundant interaction networks, and find that both types of interactions complement each other to facilitate efficient information processing across long distances. This complementarity is further enhanced during the visual stimulation. These findings provide new insights into the computational mechanisms that give rise to long-range correlations in neural systems and highlight the importance of considering different types of information interactions in understanding correlations in the brain.
Relative timing and coupling of neural population bursts in large-scale recordings from multiple neuron populations.
Olarinre M, Siegle JH, Kass RE
Front Comput Neurosci · 2025 · PMID 41728266 · Full text
INTRODUCTION: The onset of a sensory stimulus elicits transient bursts of activity in neural populations, which are presumed to convey information about the stimulus to downstream populations. Although the timing at whic...
INTRODUCTION: The onset of a sensory stimulus elicits transient bursts of activity in neural populations, which are presumed to convey information about the stimulus to downstream populations. Although the timing at which these synchronized bursts reach their peak is highly variable across stimulus presentations, the relative timing of bursts across interconnected brain regions may be less variable, particularly for regions that are strongly functionally coupled. METHODS: We developed a simple analytical framework that provides accurate trial-by-trial estimates of population burst times and of the correlations in the timing of evoked population bursts across areas. The method was evaluated using simulated data and compared to a recently published alternative model. We then applied the approach to large-scale Neuropixels recordings from six cortical visual areas and one visual thalamic nucleus in thirteen mice presented with drifting grating stimuli. RESULTS: Our method performed well on simulated data and was 85-90% faster than the alternative model while being substantially easier to apply. Applied to real data, the approach enabled identification of mouse-to-mouse variation in both peak times and region-to-region functional coupling for the first two population bursts following stimulus onset. The observed timing relationships were consistent with known anatomy and physiology. DISCUSSION: Examining sequences of activity across areas revealed that some timing relationships were preserved across all mice, while others varied across individuals. These findings demonstrate that the general approach can produce sensitive, trial-resolved analyses of timing relationships across neural populations and can capture both shared and individual-specific patterns of population burst propagation.
Explainable AI uncovers novel EEG microstate candidate neurophysiological markers for autism spectrum disorder.
Kuriyakose D, M G
Front Comput Neurosci · 2026 · PMID 41717390 · Full text
BACKGROUND: Autism spectrum disorder (ASD) is a complex neurodevelopmental condition characterized by atypical brain connectivity and impaired cognitive flexibility. Electroencephalography (EEG) based microstate analysis...
BACKGROUND: Autism spectrum disorder (ASD) is a complex neurodevelopmental condition characterized by atypical brain connectivity and impaired cognitive flexibility. Electroencephalography (EEG) based microstate analysis provides insight into the rapid temporal dynamics of brain networks, offering potential biomarkers for ASD. OBJECTIVE: This study proposes an interpretable classification framework for ASD diagnosis using multidomain microstate-informed features derived from EEG, integrating temporal, spectral, complexity-based, and higher-order metrics to comprehensively characterize brain dynamics. METHODS: Resting state EEG data from 56 participants (28 with ASD and 28 neurotypical controls; age range: 18-68 years) from the publicly available Sheffield dataset were preprocessed and segmented into microstates using a data-driven clustering approach. From these microstate sequences, we extracted a rich set of features across four domains: (i) temporal, (ii) spectral, (iii) temporal complexity, and (iv) higher-order metrics. Multiple classifiers were evaluated using 10-fold cross-validation, with hyperparameter tuning via a randomized search. RESULTS: Among all classifiers, XGBoost achieved the highest performance, with an accuracy of 80.87% when utilizing the complete multidomain feature set, significantly outperforming single domain models. Explainable AI analysis using SHapley Additive exPlanations (SHAP) identified the top 20 discriminative features, including fractional occupancy derivative for microstate 3, delta-band power in states 1 and 3, and mean inter-transition interval. Retraining XGBoost on these SHAP-selected features yielded 80.34% accuracy, confirming their robustness as potential biomarkers. Statistical validation via Mann-Whitney -tests and effect size measures further established their significance. CONCLUSION: The findings from the study demonstrated that microstate-informed features capturing temporal instability, transition unpredictability, and spectral alterations serve as clinically relevant and interpretable candidate neurophysiological markers of ASD, offering translational potential for objective diagnosis, treatment monitoring, and personalized interventions.
Editorial: The convergence of AI, LLMs, and industry 4.0: enhancing BCI, HMI, and neuroscience research.
Asgher U
Front Comput Neurosci · 2026 · PMID 41710300 · Full text
Abstract loading — click title to view on PubMed.
Metaheuristic-driven dual-layer model for classifying Alzheimer's disease stages.
Anicin L, Andjelic S, Markovic Blagojevic M … +5 more , Bulaja D, Zivkovic M, Zivkovic T, Antonijevic M, Bacanin N
Front Comput Neurosci · 2026 · PMID 41710299 · Full text
INTRODUCTION: Accurate determination of the progression phase of Alzheimer's disease (AD) is crucial for timely clinical decision-making, improved patient management, and personalized therapeutic interventions. However,...
INTRODUCTION: Accurate determination of the progression phase of Alzheimer's disease (AD) is crucial for timely clinical decision-making, improved patient management, and personalized therapeutic interventions. However, reliably distinguishing between multiple disease stages using neuroimaging data remains a challenging task. METHODS: This study proposes an advanced machine learning framework for multi-stage AD classification using magnetic resonance imaging (MRI) data. The architecture follows a two-tier design. In the first stage, convolutional neural networks (CNNs) are employed to extract deep and discriminative feature representations from MRI images. In the second stage, these features are classified using ensemble learning models, specifically XGBoost and LightGBM. Metaheuristic optimization strategies are applied to further enhance model performance. The proposed framework was evaluated using a publicly available Alzheimer's disease dataset under three different experimental configurations. RESULTS: Experimental results demonstrate that the proposed approach effectively addresses the multi-class classification problem across different AD progression stages. The optimized models achieved a maximum classification accuracy of 89.55%, indicating robust predictive performance and strong generalization capability. DISCUSSION: To improve transparency and clinical relevance, explainable artificial intelligence (XAI) techniques were incorporated to interpret model predictions and highlight feature importance. The results provide meaningful insights into neuroimaging biomarkers associated with AD progression and support the development of more interpretable and trustworthy diagnostic systems. Overall, the proposed framework contributes to improved data-driven decision support and offers a promising direction for future Alzheimer's disease diagnosis and staging research.
Cross-population amplitude coupling in high-dimensional oscillatory neural time series.
Bong H, Ventura V, Yttri EA … +2 more , Smith MA, Kass RE
Front Comput Neurosci · 2026 · PMID 41710298 · Full text
Neural oscillations have long been considered important markers of interaction across brain regions, yet identifying coordinated oscillatory activity from high-dimensional multiple-electrode recordings remains challengin...
Neural oscillations have long been considered important markers of interaction across brain regions, yet identifying coordinated oscillatory activity from high-dimensional multiple-electrode recordings remains challenging. We sought to quantify time-varying covariation of oscillatory amplitudes across two brain regions, during a memory task, based on local field potentials recorded from 96 electrodes in each region. We extended Canonical Correlation Analysis (CCA) to multiple time series through the cross-correlation of latent time series. This, however, introduces a large number of possible lead-lag cross-correlations across the two regions. To manage that high dimensionality, we developed rigorous statistical procedures aimed at finding a small number of dominant lead-lag effects. The method correctly identified ground truth structure in realistic simulation-based settings. When we used it to analyze local field potentials recorded from the prefrontal cortex and visual area V4, we obtained highly plausible results. The new statistical methodology could also be applied to other slowly varying high-dimensional time series.
EPIC-NET: EEG-based epilepsy classification and brain localization using Optuna wave-gated recurrent unit network.
Manjupriya R, Leema AA
Front Comput Neurosci · 2025 · PMID 41710068 · Full text
INTRODUCTION: Epilepsy is a chronic neurological disorder characterized by abnormal brain activity, often diagnosed through visual analysis of electroencephalography (EEG) signals. However, the existing works focused onl...
INTRODUCTION: Epilepsy is a chronic neurological disorder characterized by abnormal brain activity, often diagnosed through visual analysis of electroencephalography (EEG) signals. However, the existing works focused only on general epilepsy and failed to focus on location-based wave detection. METHODS: In this work, a novel deep learning-based EPIC-NET is proposed for epilepsy classification and brain localization using EEG signal. The EEG signals are fed into ResGoogleNet to extract both temporal and spatial features such as frequency variations, waveform morphology, and amplitude changes for epilepsy detection and localization of the affected brain regions. Stochastic Variance Reduced Gradient Langevin Dynamics based Honey Badger (SVGL-HBO) algorithm is utilized for feature selection effectively reducing dimensionality and retaining the most relevant features for detection. Based on the selected features, a fully connected layer classifies the normal and epilepsy. The Seizure Activity Index of epilepsy is classified into Low, Medium, and High using a Bell Elliptic Fuzzy Logic System (BE-FLS) guided by predefined fuzzy rules. The Optuna Wave-Gated Recurrent Unit (OW-GRU) combines GRU with wavelet processing to extract both temporal and frequency-domain features from EEG signals. Optuna is used for automatic hyperparameter tuning, which improves GRU performance, reduces overfitting, and enables accurate localization of epilepsy within specific brain lobes. RESULTS: The proposed EPIC-NET achieves the classification accuracy (CA) of 98.80% and Matthews Correlation Coefficient (MCC) of 97.43%. DISCUSSION: The EPIC-NET model improves the overall accuracy by 5.92, 10.02, and 0.59% better than RNN, SVM and CNN, respectively.
About
- Frequency
- Sun
- Papers found
- 200
- RSS feed
- Subscribe