ObjectiveThis study aimed to explore the potential of EEG in monitoring the extent of white matter microstructural damage in patients with neuromyelitis optica spectrum disorder (pwNMOSD) for the first time.MethodsThirty-two pwNMOSD and 20 healthy controls were recruited and all received DTI scan, while the pwNMOSD underwent also resting state EEG (rs-EEG). DTI indices were compared between two groups to identify impaired WM tracts in pwNMOSD. Correlations between the 240 rs-EEG indices (including spectral and functional connectivity indices in five frequency bands of six brain regions) and the fractional anisotropy (FA) of the impaired WM fiber tract were calculated to identify the rs-EEG biomarkers of WM microstructural damage. The relationships of the identified rs-EEG biomarkers with disease characteristics and cognitive function were further analyzed.ResultsSeventeen of the 20 main WM tracts were found to have microstructural damage in pwNMOSD. The functional connectivity indices were significantly and positively correlated with the FA of WM tracts in pwNMOSD, especially the coherence and phase locking value strengths in the theta and gamma frequency bands (r=0.40-0.61). Different patterns between the rs-EEG indices in different frequency bands and the integrity of WM tracts were revealed. These identified rs-EEG indices were significantly related to patients’ disease characteristics (number of attacks (r=-0.44), expanded disability status scale score (r=-0.41--0.47), serum anti-glial fibrillary acidic protein level (r=-0.43)) and cognitive functions (symbol digit modality test score, r=0.46).

The clinical diagnosis of perinatal depression (PD) presents considerable challenge, as it is much harder to identify than non-perinatal depression. Psychologically, common emotional fluctuations during pregnancy are easily confounded with depressive symptoms, leading to missed and incorrect diagnoses. Neurologically, pregnancy-induced alterations in brain activity could obscure neuroimaging features specific to PD. Therefore, this study introduced an innovative approach that combined brain network analysis with Common Orthogonal Basis Extraction (COBE) to identify a PD-Specific Network Pattern from resting-state brain networks, as well as validating its efficacy in diagnosis and assessment. Resting-state electroencephalography (EEG) data were collected from 21 patients with PD and 20 healthy pregnant (HP) individuals, from which functional brain networks were constructed. An optimized COBE method was then employed to extract Exclusive Network Pattern for each group, as well as Common Network Pattern shared by all participants (PD + HP). This process enabled the identification of the PD-Specific Network Pattern that most consistent with neural mechanisms of PD. Based on the PD-Specific Network Pattern, PD-Specific Features were derived and applied to train support vector machine and multiple linear regression models, which respectively performed individual-level classification and assessment. This study effectively addressed the limitation of traditional neuroimaging techniques in the diagnosis of PD, providing a new avenue for objective screening and dynamic monitoring of perinatal depression.

Abstract INTRODUCTION Alzheimer’s disease (AD) manifests a specific spatial progression pattern, but its propagation mechanisms remain unclear. METHODS We employed nine brain connectomes spanning multiple biological levels to investigate the mechanisms underlying cortical atrophy propagation in AD. Individual gray matter atrophy maps were quantified using normative modeling and were then mapped onto the connectomes by assessing the relationship between regional atrophy and the atrophy of neighboring regions defined by each connectome. RESULTS Cross-sectionally, node-neighbor relationship was weak in the preclinical stage, suggesting limited influence of connectome architecture. Longitudinally, atrophy became progressively more aligned with the neurotransmitter receptor similarity connectome in individuals with MCI converting to AD dementia and dementia patients. DISCUSSION Our findings described a stage-dependent shift in cortical atrophy propagation, with neurotransmitter receptor similarity playing an increasing role as AD progresses.

Abstract Adaptive emotion regulation is essential for mental health. Reappraisal and acceptance are effective yet cognitively distinct emotion regulation strategies. A key unanswered question is whether their neural implementations are supported by common overarching or distinct neurofunctional processes, especially under dynamic, naturalistic conditions that mirror real-life scenarios. Here, we combined naturalistic fMRI with multivariate predictive modeling to develop neurofunctional signatures that accurately and comprehensively characterize negative affect and its regulation via acceptance and reappraisal in dynamic, immersive contexts (n = 59). These signatures demonstrated process-specificity and generalizability across cohorts, cultures, and modalities (n = 33, 358, 45, and 33, respectively). Emotion regulation strategies were encoded in distributed, distinguishable neural representations, with shared contributions from the default mode network and strategy-specific contributions from the amygdala, somatomotor and attention (acceptance), and the frontoparietal control (reappraisal) networks. The neuromarkers precisely identified strategy-specific ER impairments in male cannabis users (nhealthy_controls = 48, ncannabis_users = 49), underscoring their potential clinical translational relevance. Collectively, these findings demonstrate shared and distinct neural signatures of reappraisal and acceptance, highlight the critical role of whole-brain integration in emotion regulation, and provide comprehensive, clinically relevant brain models of emotion regulation and dysregulation in naturalistic contexts.

Chronic social isolation (SI) beginning in adolescence can lead to serious mental health problems and social skill deficits, potentially linked to altered development and function of the prefrontal cortex (PFC), a brain region frequently implicated in neuropsychiatric disorders. Oxytocin (OXT), a neuropeptide renowned for its prosocial effects, holds significant potential as an intervention for neuropsychiatric disorders. However, the efficacy of OXT in ameliorating mental disorders induced by adolescent-onset chronic SI remains uncertain. In this work, four-week-old C57BL/6 J mice were subjected to three months of SI, and subsequent alterations in their emotional and social behaviors were assessed. Thereafter, OXT was administered intranasally to SI mice to evaluate the effects of the intervention. The results show that exposure to SI leads to anxiety- and depression-like behaviors, deficits in social novelty recognition, and long-term impairments in social memory. OXT intervention effectively reversed the damage caused by SI, including improvements in behavioral deficits, increased expression of MAP-2 and PSD-95 in PFC, downregulation of abnormally elevated OXT receptor levels, reduction of neuroinflammation, and modulation of gut microbiota homeostasis. Our study confirms the therapeutic effects of OXT in reversing isolation-induced neuropsychiatric disorders and elucidates its potential regulatory mechanisms, offering important implications for clinical interventions.

Accumulating evidence indicates that repetitive transcranial magnetic stimulation (rTMS) outcomes are state-dependent, with ongoing emotional states influencing neuromodulatory effects. However, how emotional context during stimulation influences subsequent brain functional state organization remains unclear. A total of 99 healthy participants were recruited in this study and allocated to one of three groups: active rTMS with sad film viewing (sad group), active rTMS with neutral film viewing (neutral group), or sham rTMS with sad film viewing (sham group). We combined rTMS during emotional film viewing with co-activation pattern (CAP) analysis to investigate the potential state-dependent effects of rTMS on brain dynamics. The fraction of time, resilience, and transition probabilities of CAPs were calculated to characterize brain dynamics. Four recurring CAPs were identified in this study. Relative to the sham group, participants following stimulation during the sad condition exhibited increased engagement of CAP 2, characterized by deactivations in the frontoparietal network and co-activations of the visual and somatomotor networks, along with decreased engagement of CAP 4, characterized by deactivations in the default mode network and co-activations of the somatomotor and salience networks. In contrast, the neutral condition exhibited attenuated effects. Furthermore, the change in transition probability from CAP 2 to CAP 3 was significantly greater in the sad condition than in the neutral condition. Together, these findings indicate that the effects of rTMS on brain dynamics may be modulated by the emotional context during stimulation. This work offers novel insights into the state-dependent modulation by rTMS and may help inform the optimization of therapeutic outcomes for rTMS interventions.

Background The renin-angiotensin system (RAS) has been increasingly recognized as potent modulator of cognitive and affective functions, with angiotensin II type 1 receptor (AT1R) antagonists emerging as repurposing candidate for anxiety and stress-related disorders. However, it remains unclear whether transient AT1R blockade modulates emotional attentional control and whether these effects are sex-dependent. Methods We conducted a preregistered, randomized, double-blind, placebo-controlled pharmacological eye-tracking study in 79 healthy adults (males and females) and determined effects of transient AT1R blockade via losartan (50 mg) on emotional attention control using a validated anti-saccade paradigm with social (emotional faces) and non-social stimuli. Treatment effects on state anxiety and oculomotor responses were characterized using traditional metrics and a novel trial-history informed dynamic control framework. Results Losartan reduced state anxiety irrespective of sex but induced sexually dimorphic effects on attentional control. In females, losartan enhanced performance by reducing endpoint error without altering latency. Conversely, in males, losartan increased endpoint error and prolonged latency of the first correct saccade. Trial-history analyses revealed losartan reduced error probabilities following errors and repeat trials in both sexes. Yet, following correct trials, females receiving losartan maintained lower error probabilities, while males exhibited higher errors, potentially reflecting failure to disengage from effortful control. Conclusions The RAS modulates anxiety and attentional control, the latter sex-dependently. AT1R blockade reconfigures attentional processing and adaptive control, suggesting sex-specific therapeutic potential in disorders characterized by excessive anxiety and attentional dysregulation. Clinical trials registration ClinicalTrials.gov; https://clinicaltrials.gov/; NCT06329050.

Spiking neural networks (SNNs) are promising for neuromorphic computing, but high-performing models still rely on dense multilayer architectures with substantial communication and state-storage costs. Inspired by autapses, we propose time-delayed autapse SNN (TDA-SNN), a framework that reconstructs SNNs with a single leaky integrate-and-fire neuron and a prototype-learning-based training strategy. By reorganizing internal temporal states, TDA-SNN can realize reservoir, multilayer perceptron, and convolution-like spiking architectures within a unified framework. Experiments on sequential, event-based, and image benchmarks show competitive performance in reservoir and MLP settings, while convolutional results reveal a clear space–time trade-off. Compared with standard SNNs, TDA-SNN greatly reduces neuron count and state memory while increasing per-neuron information capacity, at the cost of additional temporal latency in extreme single-neuron settings. These findings highlight the potential of temporally multiplexed single-neuron models as compact computational units for brain-inspired computing.

Healthy aging involves complex neural reconfigurations across both structural and functional domains. While resting-state functional magnetic resonance imaging (rs-fMRI) has linked static functional connectivity alterations to aging, the whole-brain dynamics of functional activity and their covariance with structural changes remain poorly characterized. To address this gap, we integrated three data-driven approaches to profile functional dynamics in the aging brain and decode their association with structural atrophy. Using rs-fMRI data from 252 participants—145 young adults (22.7±3.4 years) and 107 older adults (68.7±6.5 years)—we made several key observations. First, normalized Shannon entropy revealed a significant reduction in spatiotemporal complexity among older individuals. Second, phase synchronization analysis of BOLD signals indicated enhanced global integration and metastability in older adults, particularly within the dorsal attention (DAN), ventral attention (VAN), and frontoparietal networks (FPN). Third, temporal asymmetry analysis demonstrated increased nonreversibility and a heightened functional hierarchy in the aging brain, again most evident in the FPN. Morphometric analyses confirmed widespread structural atrophy in older participants. Crucially, partial least squares (PLS) analysis uncovered significant covariance between morphometric patterns and dynamic functional metrics, underscoring a tight structure-dynamics coupling in aging. Furthermore, structural atrophy correlated significantly with variations in micro-architecture maps. Finally, we evaluated the behavioral relevance of these dynamics through correlations with cognitive performance. Our findings offer an integrative, multiscale perspective on neural decline in aging, emphasizing the interplay between dynamic functional reorganization and structural atrophy.

The incidence of neuropsychiatric disorders such as Autism Spectrum Disorder (ASD), Attention Deficit Hyperactivity Disorder (ADHD), and Major Depressive Disorder (MDD) continues to rise. Deep learning-based computer-aided diagnosis (CAD) has emerged as a promising approach to alleviate the increasing burden on neuroimaging-based clinical resources. However, neuroimaging modalities such as functional magnetic resonance imaging (fMRI) involve complex spatiotemporal characteristics, making their representations susceptible to various types of noise and interference, which in turn hampers the effectiveness of CAD. To address this challenge, we propose a pseudo-label consistency-driven framework for functional connectivity (FC) reconstruction and discriminative modeling (PL-FCDM), aiming to enhance both the representational quality and discriminative power of FC features. Specifically, two complementary pseudo-labeling models are developed to independently capture discriminative features from the temporal domain (time series) and spatial domain (dynamic functional connectivity), enabling pseudo label prediction from distinct modalities. Then a consistency-based filtering strategy is applied to construct high-confidence reconstructed functional connectivity. These graphs are subsequently fed into a classification model comprising a Feature Optimization Autoencoder and a Depthwise Separable Convolutional Neural Network for efficient identification of neuropsychiatric disorders. Extensive experiments conducted on four publicly available multi-site datasets—ABIDE I, ABIDE II, ADHD-200, and REST-meta-MDD demonstrate that the proposed method achieves classification accuracies of 76.14%, 74.37%, 72.89%, and 71.15%, respectively. These results consistently outperform several state-of-the-art approaches, validating the effectiveness and robustness of the proposed framework in feature refinement and multi-disorder recognition.

Overarching conceptualizations propose a critical role of the default mode network (DMN) in self-referential mental time travel, particularly in autobiographical memory retrieval and episodic future thinking, and internal (intrinsic) emotion generation and regulation. However, these conceptualizations have not been directly evaluated. Against this background, the present fMRI study aimed to identify both shared and distinct neural systems underlying autobiographical episodic processing across different temporal contexts - specifically, episodic memory retrieval (EMR) and episodic future thinking (EFT) - and to examine how these systems interact with affective experiences, including valence and arousal. Our findings demonstrated the central role of the DMN - encompassing the medial prefrontal cortex (mPFC), posterior cingulate cortex (PCC), and medial temporal lobe (MTL) - in both EMR and EFT. Importantly, we identified a functional dissociation along both valence and temporal dimensions: the ventromedial prefrontal cortex (vmPFC) was more strongly associated with positive experiences and simulations, whereas the dorsomedial prefrontal cortex (dmPFC) was consistently engaged during the processing of negative affect across past and future contexts. Moreover, representational similarity and parametric analyses indicated that the hippocampus supports differential processing of valence and arousal across temporal domains. Together, these findings provide empirical evidence for the involvement of cortical midline core DMN systems in autobiographical processing across time and suggest overlapping and distinct systems for the integration of emotional experiences across mental time travel. ### Competing Interest Statement The authors have declared no competing interest. National Natural Science Foundation of China, 82271583 Ministry of Science and Technology of China, 2022ZD0208500 Hong Kong University Grants Council, 17615525 University of Hong Kong seed funding and start-up schemes, 2407102536

Brain-computer interface (BCI) establishes a bidirectional pathway between the brain and external devices. Its applications fall into two main categories: utilizing the brain as a controller (e.g., for prosthetics) or as a modulation target (e.g., for cognitive regulation). Progress in BCI is constrained by two core bottlenecks: in brain control, limited understanding of neural coding mechanisms restricts improvements in the accuracy and robustness of encoding/decoding algorithms; in brain regulation, one-size-fits-all regulatory strategies struggle to address significant individual variability, resulting in heterogeneous therapeutic responses. Inspired by neuroscience advances, this perspective proposes a new biological brain – digital twin brain based BCI (BDBCI) paradigm. Here, the biological brain acts as an empirical anchor and ultimate validation platform, while a high-fidelity digital twin brain (DTB) serves as a theoretical inference engine and virtual testbed. Specifically, experimental induction is applied to the biological brain to distill preliminary conclusions, such as brain-behavior mappings and brain-stimulation causal relationships, which are then used to construct and calibrate the DTB model. Subsequently, on the DTB platform, large-scale model deduction is conducted to validate and deepen these preliminary insights mechanistically, thereby optimizing control/regulation parameters or informing the parameter ranges for the next round of experimental induction and model deduction. Through this BDBCI paradigm, we aim to advance BCI research from empirical trial-and-error toward a new era of model-driven, predictable, and explainable precision science.

Visual perceptual deficits are increasingly acknowledged as a core yet underexplored feature of Major Depressive Disorder (MDD), involving impairments in contrast sensitivity, contextual modulation, and social-emotional perception. However, the underlying circuit-, neuron-, and network-level mechanisms through which depressive states influence early visual processing remain poorly understood. In this study, mice subjected to chronic restraint stress (CRS) underwent single-unit recordings in layer 2/3 (L2/3) of the primary visual cortex (V1) under light anesthesia to examine the effects of chronic stress on visually evoked neuronal responses and local network dynamics. V1 neurons in CRS-exposed mice exhibited broadened orientation tuning bandwidths, diminished surround suppression, and impaired center-surround discontinuity discrimination. At the network level, visually evoked oscillatory power in the θ, low γand high γ frequency bands was significantly attenuated in CRS mice, accompanied by a marked reduction in visual evoked potential (VEP) amplitude. These findings show that chronic stress impairs both the tuning of V1 neurons to visual stimuli and the local neural rhythms underlying early visual processing. The results uncover a multi-level physiological basis for depression-related visual deficits and provide a preclinical framework that could help develop objective sensory biomarkers for MDD.

This manuscript describes the characteristics of a dataset with EEG recordings of 173 participants: 87 confirmed COVID-19 cases (positive Reverse Transcription Polymerase Chain Reaction RT-PCR) and 86 healthy close contacts (negative PCR). EEG was acquired using a 10–20 standard montage. Resting EEG was recorded for 8 minutes in all participants with eyes closed. Afterwards, it was also recorded 2 minutes of alternating closed and open eyes, followed by 2 minutes of recovery. All participants underwent physical, neurological, and clinical evaluations, including a retrospective neurological survey and the Schedules for Clinical Assessment in Neuropsychiatry (SCAN) version 2.1.

BACKGROUND:Anhedonia is a core but underrecognized symptom of major depressive disorder (MDD). Whether remission defined by Hamilton Depression Rating Scale (HAMD) and Hamilton Anxiety Rating Scale (HAMA) reflects improvement in anhedonia capacity measured by the Snaith-Hamilton Pleasure Scale (SHAPS) remains unclear. Evidence regarding repetitive transcranial magnetic stimulation (rTMS) efficacy for anhedonia is inconsistent. In this study, we examined SHAPS-stratified neurophysiological phenotypes using resting-state electroencephalography (EEG) spectra and functional connectivity to evaluate differential responses to intermittent theta burst stimulation (iTBS). METHODS:Fifty MDD patients with anhedonia were classified into high-anhedonia (HA) (n = 25) and low-anhedonia (LA) (n = 25) groups based on SHAPS score, alongside 25 healthy control (HC) participants. All patients received a 2-week course of iTBS over the left dorsolateral prefrontal cortex. Resting-state EEG and clinical assessments were acquired at baseline and posttreatment. RESULTS:At baseline, both patient groups exhibited lower alpha power compared with HC group, with region-specific deficits in the LA group and bilateral reductions in the HA group. Following iTBS, the LA group demonstrated the HAMD, HAMA, and SHAPS scores and alpha power recovery to the state of HC group and increased frontal/central-posterior connectivity. Conversely, although HAMD and HAMA scores improved in the HA group, SHAPS scores remained higher and alpha power deficits persisted and exhibited decreased bilateral connectivity, particularly in the right hemisphere. CONCLUSIONS:Remission of depressive and anxiety symptoms does not necessarily indicate recovery from anhedonia. Distinct neurophysiological responses to iTBS suggest divergent pathophysiological mechanisms in anhedonia subtypes, which may inform personalized treatment strategies for MDD.

BACKGROUND:Subcortical regions are widely implicated in the pathological mechanisms and treatment of schizophrenia, and accumulating evidence, including our prior findings, suggests that subcortical functional dysconnectivity is closely associated with treatment response. Accordingly, the present study aimed to examine the relationship between the subcortical functional connectivity (FC) and treatment outcomes in schizophrenia using multivariate analytical approaches and machine learning algorithms. METHODS:One hundred and nineteen individuals with first-episode schizophrenia were recruited for this study. All patients underwent MRI scanning and completed assessments with the Positive and Negative Syndrome Scale (PANSS) at baseline and at follow-up after 12 weeks of antipsychotic medication. We employed partial least squares analysis to explore the multivariate associations between changes in subcortical FC (∆FC) and changes in symptom severity (∆PANSS). In addition, a machine learning algorithm was used to predict the antipsychotic treatment outcome based on the distinctive subcortical FC pattern at baseline. RESULTS:We identified a distinctive subcortical FC pattern dominated by the striatum that was associated with overall treatment outcomes in first-episode schizophrenia. Furthermore, the reduction in PANSS total scores predicted using baseline subcortical FC patterns was positively correlated with the actual reduction in PANSS total scores following antipsychotic treatment. CONCLUSION:These results indicate that the distinctive subcortical FC pattern holds promise as a biomarker for schizophrenia, supporting individualized treatment approaches and facilitating early intervention to improve clinical outcomes.

Background Capturing multimodal acute stress responses is important for understanding how stress is expressed across psychophysiological systems. Pupil diameter and electroencephalogram (EEG) are important indices for assessing acute stress responses. However, the relationships among behavioral, pupillary, and EEG responses during sustained stress exposure remain less well understood. Methods We collected multimodal data, including self-reported stress, salivary cortisol, electrocardiogram (ECG), pupil diameter, behavioral performance, and EEG from 33 healthy young adults during the Montreal Imaging Stress Task (MIST). Differences in responses between the training and testing conditions across difficulty levels were analyzed to characterize changes in the multimodal dynamics of acute stress responses across MIST blocks, and exploratory correlation analyses were conducted to examine associations between pupil diameter changes and relative EEG power changes Results The testing condition successfully elicited an acute stress response profile, reflected by increased self-reported stress, salivary cortisol, heart rate, and pupil diameter, alongside decreased behavioral performance and a shift in relative EEG power from lower to higher frequency bands. During the testing condition, behavioral performance improved and pupil diameter decreased from the medium to the hard difficulty block, accompanied by changes in relative alpha, beta (beta1, beta2, beta3), and gamma power. These findings may reflect adaptive-like changes over the course of the testing condition. Exploratory analyses showed limited associations between pupil diameter changes and relative EEG power changes, restricted to changes from the medium to the hard difficulty block in the testing condition and mainly involving frontal relative beta2 power. Conclusion Our findings suggest that acute stress involves coordinated but partly dissociable multimodal dynamics across MIST blocks and highlight the value of integrating behavioral, pupillary, and EEG measures to characterize acute stress responses.

The mechanisms underlying the impact of metabolic syndrome on cognitive dysfunction in patients with schizophrenia remain unclear. The present study employed a two-factor factorial design to investigate the effects of metabolic syndrome on white matter microstructure in schizophrenia and its association with cognitive function. A total of 187 participants were included and classified into four groups based on the diagnoses of schizophrenia and metabolic syndrome: schizophrenia patients with metabolic syndrome (SZ-wMS), schizophrenia patients without metabolic syndrome, healthy controls with metabolic syndrome, and healthy controls without metabolic syndrome. Diffusion tensor imaging data were acquired. Using diffusion tensor model, fractional anisotropy (FA) was calculated to characterize the microstructural integrity of white matter. Peripheral metabolic indices and multiple domains of cognitive function were also assessed. The SZ-wMS group showed further reduced FA in the right sagittal stratum. Within the two-factor analytical framework, an interaction effect between metabolic syndrome and schizophrenia on FA in the right sagittal stratum was identified. Correlation analyses revealed that reduced FA in the right sagittal stratum was associated with impaired language function in patients with schizophrenia. Moreover, mediation analysis indicated that body mass index might indirectly affect language function by influencing FA in the right sagittal stratum. In summary, reduced integrity of white matter fibers in sagittal stratum may represent a potential neural mechanism underlying the comorbidity of schizophrenia and metabolic syndrome and may be associated with language dysfunction in schizophrenia.

Previous studies showed abnormalities in both visual motion perception (VMP) and occipital cortex activity in subjects suffering from major depressive disorder (MDD). Can the psychophysical and/or neural markers of visual perception serve for clinical identification of MDD subgroups? To address this yet unresolved issue, we develope a novel analytical framework combining visual perceptual measurement with machine learning clustering to identify MDD subgroups. Within a cohort of 272 individuals with acute MDD, this approach reveals a VMP-positive (VMP-P) biotype characterized by impaired VMP. Clinically, this biotype exhibits significant cognitive deficits. We validate the robustness of this biotype through cross-validation and confirm its generalizability in an independent sample (n = 63). Furthermore, 7T MRI implicate aberrant neural activity in the occipital cortex as a mechanism underlying the VMP-P biotype. Our findings establish a novel, clinically translatable path for stratifying MDD, which can guide treatment development and advance precision medicine.

Objective.Sleep is hypothesized to restore near-critical dynamics in large-scale brain networks, whereas insomnia may disrupt this self-organizing process. This study aimed to determine whether insomnia alters neural avalanche dynamics and criticality-based EEG metrics, and whether these metrics enhance prediction of sleep fragmentation compared with conventional spectral measures.Approach.Overnight high-density electroencephalography was recorded from 50 participants aged 16-69 years, including healthy sleepers and individuals with insomnia. Neural avalanches were detected as clusters of significant amplitude excursions. The branching parameter (σ) quantified temporal propagation within avalanches, while the deviation-from-criticality coefficient (DCC) indexed the system's distance from the critical state. These criticality features were contrasted with spectral power measures in predictive models of non-rapid eye movement (NREM) sleep fragmentation.Main results.Participants with insomnia exhibited reduced avalanche density and diminished slow-wave activity, accompanied by significant deviations ofσfrom the critical value and elevated DCC across the night. Criticality-based metrics captured fragmentation dynamics more sensitively than spectral features. In predictive modeling, criticality measures significantly outperformed spectral power in forecasting NREM fragmentation (F1-score = 0.69 vs. 0.62), with the strongest gains in mild and severe insomnia subgroups.Significance.Insomnia is characterized by a persistent deviation from near-critical neural dynamics, reflecting compromised stability and recovery during sleep. Criticality-based EEG features provide a more mechanistic and predictive framework for identifying sleep fragmentation and may offer novel biomarkers for quantifying disrupted sleep physiology in clinical insomnia.