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.

Background:Autism Spectrum Disorder (ASD) and Attention-Deficit/Hyperactivity Disorder (ADHD) exhibit high clinical overlap, but categorical diagnostic boundaries obscure their shared, dynamic physiological vulnerabilities during real-world sensory processing. Methods:We analyzed multimodal eye-tracking synchrony in a large transdiagnostic pediatric cohort (N = 2,026) during naturalistic viewing of four distinct media paradigms. A novel 2D complex correlation framework captured gaze inter-subject correlation (ISC) magnitude and spatiotemporal phase divergence, while 1D pupil ISC measured autonomic arousal synchrony. Linear models evaluated dimensional (RDoC) and categorical (2×2 ANCOVA) diagnostic frameworks alongside rigorous medication and severity controls. Results:Dimensional models revealed a domain-general vulnerability: autistic traits independently predicted widespread reductions across gaze synchrony in all media contexts, and pupillary synchrony in narrative-driven contexts, whereas continuous ADHD traits showed minimal independent effects. In contrast, severe spatiotemporal misalignment (phase divergence) did not scale dimensionally but emerged strictly at clinical boundaries, reflecting highly idiosyncratic spatial locking in isolated ASD. Furthermore, categorical models demonstrated a robust, non-additive interaction: the clinical co-occurrence of ADHD paradoxically buffered against this severe spatiotemporal decoupling. Crucially, this protective phenotype was localized strictly to character-driven social narratives and remained highly significant after rigorously adjusting for daily stimulant medication, outlier instability, and baseline autism trait severity. Conclusions:These findings validate model-free physiological synchrony as a candidate transdiagnostic biomarker. Rather than compounding impairment, comorbid ASD and ADHD reflect competing, non-additive neurocognitive strategies that yield distinct, context-dependent visual phenotypes.

Abstract Background Autism Spectrum Disorder (ASD) and Attention-Deficit/Hyperactivity Disorder (ADHD) share substantial clinical and physiological overlap. While naturalistic and sensory-driven paradigms increasingly capture evoked neurophysiological responses, the intrinsic baseline physiology of these conditions remains poorly defined. We characterized resting-state pupillary volatility and oculomotor stability across the ASD-ADHD spectrum using dimensional and categorical (DSM-5) frameworks. Methods We analyzed resting-state eye-tracking data from a large pediatric cohort (N = 2,315) from the Healthy Brain Network, extracting Pupil Relative Volatility (Coefficient of Variation [CV]) and Bivariate Contour Ellipse Area (BCEA) to index pupillary and spatial gaze stability. Data were evaluated via continuous regressions against Social Responsiveness Scale (SRS) and SWAN inventories, then 2×2 factorial ANCOVAs based on clinical diagnoses, with sensitivity analyses for extreme values, hardware heterogeneity, and psychostimulant medication. Results Dimensional models revealed no significant association between either metric and continuous ASD or ADHD trait severity, apart from a modest sub-additive SRS × SWAN interaction on pupillary CV; an apparent pooled BCEA-trait association proved attributable to hardware differences and was absent within hardware-consistent subgroups. Categorical models showed robust diagnostic differentiation for both metrics: isolated ASD was associated with elevated CV and impaired BCEA, with ADHD additionally impairing BCEA. Comorbid ASD+ADHD produced a sub-additive CV interaction robust across outlier-resistant, hardware-stratified, and medication-adjusted analyses; an analogous BCEA interaction was directionally consistent but not significant under standard estimation. Conclusions Categorical diagnostic status was consistently associated with baseline pupillary and oculomotor physiology across the ASD-ADHD spectrum in this cohort, with a shared sub-additive comorbidity effect most robustly expressed in pupillary volatility; continuous trait severity showed little independent association with either metric. These findings, obtained using a rigorously validated, hardware-heterogeneous multi-site sample, establish a foundation for future naturalistic and sensory-evoked investigations of the shared physiological architecture underlying ASD-ADHD comorbidity.

The human brain depends on dynamic interactions among modular networks, where connector and provincial hubs facilitate efficient information integration. Most previous studies have relied on single metrics or qualitative labels to identify hubs, overlooking multi-metric integration and the quantitative contributions of nodes. Here, we introduce the Multi-Indicator Entropy Hub Score (MIEHS), which integrates six graph-theoretical metrics to quantify hub properties. Validated on benchmark and simulated networks as well as resting-state fMRI data from the Midnight Scan Club dataset, MIEHS reliably identifies hubs. High-scoring connector hubs were localized in the attention network, whereas high-scoring provincial hubs were concentrated in the default mode network. Gradient mapping further revealed that connector hubs bridge unimodal and transmodal regions, supporting information transfer from primary sensory areas to higher-order cognitive regions, while provincial hubs primarily sustain intra-network communication. Null model analyses highlighted the stability of hubs within the default mode and limbic networks. Although hubs are widely studied, they have not yet been established as robust clinical biomarkers. Using Partial Least Squares analysis in the UCLA dataset (HC = 110, ADHD = 37, BD = 40, SCHZ = 37), we observed significant associations between hub alterations in the DMN, SMN, limbic, DAN, and control networks and measures of cognitive flexibility, abstract reasoning, and verbal expression. Together, these findings demonstrate that MIEHS provides a robust and versatile framework for mapping brain network organization and characterizing functional reconfiguration.

BACKGROUND:The glymphatic system, a key fluid clearance pathway in the central nervous system, is emerging as a potential therapeutic target for synucleinopathies. Dysregulation of this system may contribute to spinocerebellar ataxia type 3 (SCA3) pathogenesis, in which the accumulation of misfolded proteins acts as a central driver. OBJECTIVES:The goal was to investigate glymphatic system function in SCA3 patients and evaluate its relationship with brain damage and clinical disability. METHODS:Ninety-two SCA3 patients (14 with premanifest SCA3 and 78 with manifest SCA3) and 98 healthy controls underwent clinical evaluation and magnetic resonance imaging (MRI) scans. MRI parameters, including the diffusion along the perivascular space (DTI-ALPS) index (a proxy for glymphatic function); cerebral, cerebellar, and subcortical gray matter volumes; and whole-brain microstructural properties of white matter, were calculated. RESULTS:Patients with premanifest and manifest SCA3 had lower ALPS indices compared with healthy controls, and patients with manifest SCA3 had a lower ALPS index compared with premanifest SCA3. In SCA3 patients, lower ALPS index was associated with more severe disability and longer disease duration. A negative correlation between ALPS and disease duration emerged after 3 years, with no significant association observed before the 3-year cutoff. Moreover, lower ALPS index was correlated with more pronounced cortical and subcortical gray matter atrophy, decreased fractional anisotropy, and elevated mean diffusivity in white matter. CONCLUSIONS:Our findings demonstrate that glymphatic function is impaired, particularly in the presymptomatic stage of SCA3, and this impairment is associated with disability, neurodegeneration, and demyelination. Therefore, glymphatic dysfunction may contribute to the pathogenesis of SCA3. © 2026 International Parkinson and Movement Disorder Society.

Eye gaze and eye movements provide important indices of perceptual and cognitive processes, particularly under naturalistic conditions such as movie viewing. However, concurrent eye tracking is often unavailable in functional MRI (fMRI) studies due to technical and logistical constraints. Recent deep learning approaches have made it possible to estimate eye gaze directly from eyeball signals in fMRI data, offering a potential alternative to camera-based eye tracking. Here, we applied a pre-trained fMRI-based deep neural network model (DeepMReye) to estimate eye gaze during movie watching across three independent fMRI datasets. Model performance was evaluated by comparison with camera-based eye-tracking data when available, as well as by assessing inter-individual correlations, a commonly used benchmark in naturalistic fMRI research. At the individual level, predicted gaze showed modest correspondence with measured data (r ≈ -0.38 to 0.67). In contrast, group-averaged gaze predictions exhibited substantially higher correlations (r ≈ 0.7-0.8), indicating improved reliability at the group level. We further derived eye-movement-related time series from the predicted gaze signals and examined their associated brain activity. Consistent with differences in prediction accuracy, individual-level analyses yielded activation patterns largely restricted to visual cortex, whereas group-averaged predictions revealed more widespread activation, including established oculomotor control regions such as frontal and parietal eye fields. Exploratory analyses indicated age-related effects on gaze prediction accuracy and eye-movement-related brain activity, although these effects were not consistent across datasets. Together, these findings demonstrate that group-averaged fMRI-based gaze estimation can support the investigation of eye-movement-related brain activity in naturalistic paradigms, while highlighting current limitations for individual-level inference. The results provide a methodological assessment of fMRI-based gaze prediction and inform its appropriate use in future neuroimaging studies.

Background: Naturalistic fMRI provides an ecologically valid window into social brain function, yet binary diagnostic labels may obscure neural signatures linked to the continuous spectrum of social deficits. We investigated whether social brain alterations in autism spectrum disorder (ASD) follow a categorical, dimensional, or "dual-track" architecture. Methods: We analyzed fMRI data from 428 youth (262 ASD, 166 typically developing; ages 5-22) watching two films: The Present and Despicable Me. Using Principal Component Analysis (PCA) to quantify primary (PC1) and secondary (PC2) synchronization, we employed variance partitioning to disentangle the contributions of categorical diagnosis from continuous symptom severity (Social Responsiveness Scale-2, SRS-2). Results: During The Present, reduced synchronization was widespread. In social-motivational hubs (medial prefrontal cortex, caudate), reductions were largely explained by variance shared between diagnosis and SRS-2 scores. In contrast, the left amygdala exhibited a unique dimensional association with SRS-2 scores independent of categorical diagnosis. Secondary response patterns (PC2), reflecting complex temporal integration, revealed further unique dimensional effects in the cuneus. Notably, these signatures were stimulus-dependent, manifesting during the emotionally complex narrative of The Present but not during the slapstick-oriented Despicable Me. Conclusions: While core social-motivational hubs reflect overlapping diagnostic and dimensional deficits, the amygdala and secondary visual patterns provide distinct, dimension-specific signatures of social impairment. This variance partitioning approach supports a Research Domain Criteria (RDoC) framework, highlighting the necessity of integrating dimensional assessments and narrative complexity to characterize the neural architecture of autism.

Autism spectrum disorder (ASD) is increasingly conceptualized as a disorder of large-scale functional brain network organization rather than isolated regional abnormalities. Graph-theoretical analysis provides a principled framework for characterizing such distributed network reconfiguration. Here, we investigated global, nodal, and system-level functional network topology in ASD using a large, multi-site resting-state fMRI dataset. Resting-state fMRI data from 996 participants (428 ASD, 568 healthy controls) were obtained from the ABIDE I and II data repositories. Whole-brain weighted resting state functional networks were constructed using Pearson correlation. To improve robustness and reduce threshold-selection bias, graph-theoretical metrics were computed across a range of network sparsity thresholds and summarized using an area-under-the-curve (AUC) approach. At the global level, ASD was associated with reduced assortativity, and local efficiency, along with altered normalized characteristic path length (λ), indicating local information processing and subtle deviations in network integration relative to an optimal small-world topology. Nodal analyses revealed non-random, region-specific alterations predominantly affecting higher-order associative systems. Increased nodal centrality and hub-like properties were observed in frontal and parietal regions within the frontoparietal control and dorsal attention networks, whereas reduced nodal efficiency and centrality were primarily localized to limbic and anterior temporal regions, including the temporal pole. System-level analyses, controlling for age, sex, and acquisition site, further demonstrated network-specific topological reorganization across multiple functional systems. Clinical correlation analyses identified modest but significant associations between nodal topology and core ASD symptom severity, particularly within default mode, limbic, and attention networks. Together, these findings indicate that ASD is characterized by subtle yet reproducible multi-scale reorganization of functional brain network topology, supporting a systems-level account of ASD neurobiology and highlighting the clinical relevance of large-scale network architecture. ### Competing Interest Statement The authors have declared no competing interest. ### Funding Statement This work was supported by the China MOST2030 Brain Project (2022ZD0208500). ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: The resting-state fMRI data used in this study are publicly available from the Autism Brain Imaging Data Exchange (ABIDE) repository at http://fcon_1000.projects.nitrc.org/indi/abide/. I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes All data produced are available online at http://fcon_1000.projects.nitrc.org/indi/abide/.

While resting-state brain dysfunctions have been extensively investigated in Alzheimer's disease (AD), the dynamic alterations of functional systems remain poorly understood. We employed co-activation pattern (CAP) analysis to characterize the functional-state alterations in 243 participants using resting-state fMRI data and applied graph theory analysis to estimate corresponding topological properties. The CAP analysis identified five distinct brain states across groups: State 1 (limbic network dominated), State 2 (dorsal attention network (DAN) and central executive network dominated), State 3 (default mode network and central executive network dominated), State 4 (somatomotor network and ventral attention network dominated), and State 5 (DAN, sensorimotor, and visual networks dominated). Compared to cognitively unimpaired individuals, State 3 demonstrated significantly reduced persistence and resilience in both mild cognitive impairment (MCI) and AD groups. Additionally, both clinical groups (MCI and AD) exhibited decreased transitions from State 2 to State 5 and reduced self-transitions within State 3. Graph theory analysis revealed that compared to cognitively unimpaired individuals, MCI and AD individuals had increased node degree centrality and node efficiency, alongside decreased node local efficiency in regions within the default mode network (DAN) and visual network, which corresponded well with CAP analysis results. Our findings provide a multiscale framework linking dynamic state instability to static network reorganization, advancing understanding of the dynamic functional alterations underlying cognitive decline in AD spectrum disorders.

Spinal cord injury (SCI) is associated with cardiovascular deficits that affect cerebral blood flow, cerebral perfusion, and cerebrovascular control. While several studies use neuroimaging techniques such as functional magnetic resonance imaging (fMRI) to understand neuroplasticity following SCI, more work needs to be done to evaluate the cerebrovascular changes following SCI. Understanding these effects using neuroimaging is essential as these deficits also affect neurovascular coupling and how we interpret neuroplasticity measured based on neuroimaging. Hence, we conducted a pilot study in twelve healthy males and thirteen males with thoracolumbar SCI using functional near-infrared spectroscopy (fNIRS) to understand the effects of breath-holding induced hypercapnia on the hemodynamics of the sensorimotor cortex and prefrontal cortex (PFC) after SCI. Participants performed 30 seconds of regular breathing alternated by 15 seconds of breath-holding for 5 minutes. Compared to controls, the SCI group presented with a greater initial decrease in oxy-hemoglobin concentration change and a delayed subsequent increase in oxy-hemoglobin concentration change in response to hypercapnia at p<. Additionally, the net increase in oxy-hemoglobin concentration change following BH in the PFC was negatively correlated with the level of injury at p=0.005, where higher levels of injury were associated with a smaller increase in oxy-hemoglobin concentration following hypercapnia. These findings confirm that a) SCI, including lower levels of injury (below T6) are associated with cerebrovascular changes that are quantifiable using fNIRS, and b) fNIRS could be a robust tool to understand the neuroplastic and cerebrovascular changes in people with SCI. ### Competing Interest Statement The authors have declared no competing interest. ### Funding Statement This study was funded by a fellowship from the New Jersey Commission on Spinal Cord Research (NJCSCR; CSCR15FEL002) to KDK and BBB. ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: Ethics committee/IRB of New Jersey Institute of Technology gave ethical approval for this work. I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes All data produced in the present study are available upon reasonable request to the authors.

Increasing attention has been paid to the nonlinear functional activity of human brain regions. This paper extends Chatterjee's correlation coefficient (CCC) method to model nonlinear relationships in brain functional networks explicitly. Specifically, the reliability and effectiveness comparisons between CCC and the Pearson correlation coefficient (PCC) are demonstrated using simulated data and two real resting-state functional magnetic resonance imaging (rs-fMRI) datasets: The Midnight Scan Club dataset and the UCLA dataset. The results demonstrate that CCC accounts for both linear and nonlinear dependencies and that its reliability is better than that of PCC. Additionally, from a whole-brain perspective, the number of connections in different brain regions was observed in the following order: bipolar disorder (BP) and healthy controls (HC) > adult attention-deficit/hyperactivity disorder (ADHD) and HC > schizophrenia (SZ) and HC. The commonalities among the three psychiatric disorders compared to HC were differences in occipital, default, cerebellum, and the regions connected to the occipital. Besides, using CCC: occipital performed classification best (AUC: 0.657) between ADHD and HC, and (AUC: 0.622) between BP and HC, but fronto_parietal performed classification best (AUC: 0.700) between SZ and HC. This method enhances sensitivity to group differences and may provide new insights for exploring functional networks based on fMRI in the future.

Naturalistic stimuli, such as movies and narratives, are increasingly used in cognitive neuroscience to map cognitive and affective processes onto brain activity measured with functional MRI (fMRI). Features extracted from movies span multiple levels, from computational visual and auditory inputs to physiological signals and subjective ratings. However, the temporal alignment between these features and the blood-oxygen-level-dependent (BOLD) response varies considerably, and the commonly used canonical hemodynamic response function (HRF) with temporal derivatives may not adequately capture these delays. In this study, we analyzed three movie-watching datasets using cross-correlation and finite impulse response (FIR) deconvolution to map the unconstrained temporal dynamics of visual, auditory, pupillary, and Theory of Mind (ToM) features across the brain. Our results demonstrate that while the canonical HRF effectively captures basic sensory features, it introduces systematic misalignments for inherently delayed signals. Because physiological markers (pupil size) and subject reports (ToM) intrinsically lag the underlying neural events, standard HRF convolution overcompensates for their biological latency, introducing a redundant phase mismatch or "double-delay." Furthermore, our unconstrained FIR models revealed distinct inter-regional temporal hierarchies across the cortex. Given the inherent collinearity of real-world stimuli, these estimated profiles capture the bundled, multi-dimensional dynamics of naturalistic processing rather than perfectly isolated feature effects. Overall, these findings highlight the necessity of flexible, reliability-tested temporal modeling to accurately map the complex processing timescales engaged during naturalistic viewing.

Oxytocin (OT) is a neuropeptide widely implicated in emotional regulation and social cognition. However, its effects on dynamic brain connectivity remain poorly understood. In this study, we applied co-activation pattern (CAP) analysis to resting-state fMRI data to examine how a single intranasal dose of OT modulates whole-brain functional dynamics. Participants included healthy young (18-31 years) and older (63-81 years) adults, with analyses conducted at both the group level and across age subgroups. OT significantly altered temporal properties of brain states, including increased frequency, in-degree, and out-degree in multiple CAPs, indicating enhanced network flexibility and switching. Notably, OT modulated states involving the amygdala, medial prefrontal cortex, and salience network, regions critical for emotion regulation, and increased self-transition probabilities, suggesting greater within-state stability. Age-stratified analysis revealed differential sensitivity: young adults exhibited more pronounced modulation and greater dynamic flexibility, while older adults showed more sustained engagement with emotion-related states. Importantly, only in the elderly OT and combined young subgroups did time spent in these states significantly correlate with cognitive performance on the Digit Symbol Substitution Test, suggesting that OT-enhanced engagement in these networks supports compensatory mechanisms during aging. No such correlations were found in young participants or in either age group under placebo, highlighting the specificity of oxytocin's functional relevance in older adults. Meta-analytic decoding using Neurosynth confirmed that OT-modulated regions are closely associated with emotion, memory, and social cognition. These findings demonstrate that OT shapes transient brain dynamics in age- and function-specific ways. CAP analysis provides a powerful approach for capturing such neuromodulatory effects.

Alzheimer's disease (AD) has traditionally been regarded as a disorder primarily affecting gray matter, while growing evidence highlights the significant role of white matter pathology in its progression. This review aims to assess the current state of knowledge regarding white matter abnormalities and elucidate the potential impact of white matter on the pathogenesis and progression of AD. White matter alterations, including inflammation, hyperintensities, structural and functional changes, often precede gray matter atrophy and cognitive decline during AD progression. Advanced imaging and histopathological studies suggest that white matter degeneration is not merely a downstream consequence of gray matter pathology; it may represent an independent, perhaps initiating, pathological pathway in AD progression. Moreover, white matter lesions in individuals with AD may be modifiable by both pharmacological and non-pharmacological interventions, supporting the potential for reducing white matter damage and improving cognitive functions.

Background:Autism spectrum disorder (ASD) is characterized by heterogeneous developmental trajectories, yet it remains unclear whether frequency-specific resting-state functional magnetic resonance imaging (rs-fMRI) features can distinguish age-defined developmental stages within the condition. Methods:We analyzed rs-fMRI data from 251 participants with ASD, comprising 146 children and 105 adolescents aggregated from ten sites in the Autism Brain Imaging Data Exchange (ABIDE). ALFF and ReHo were computed across three frequency bands: Conventional (0.01-0.08 Hz), slow-4 (0.027-0.073 Hz), and slow-5 (0.01-0.027 Hz). Region-of-interest features were extracted using the 246-region Brainnetome Atlas. To ensure rigorous generalization, participants were divided into a stratified training set (80%, n = 200) and a held-out test set (20%, n = 51), with stratification based on the child-adolescent group label and a fixed random seed of 42. CovBat harmonization parameters, feature-scaling parameters, LASSO feature selection, and classifier hyperparameters were estimated using the training data only and subsequently applied to the held-out test data. Final model performance was evaluated once on the held-out test set. Performance was evaluated using Logistic Regression (LR), Support Vector Machine, and Random Forest classifiers, with Shapley Additive Explanations (SHAP) used to characterized interpret feature contributions. Results:The slow-4 and Conventional-band features showed higher held-out ASD test-set performance than slow-5 features. The best single-metric model by area under the receiver operating characteristic curve (AUC) was slow-4 ReHo Logistic Regression, which achieved an AUC of 0.811 and accuracy of 0.745. The exploratory combined model using slow-4 ALFF and ReHo features achieved the highest overall AUC of 0.819 (accuracy = 0.725). SHAP analysis identified distributed model-contributing regions in the slow-4 ReHo model, including the inferior parietal lobule, lateral occipital cortex, middle and inferior frontal gyri, basal ganglia, and thalamus. Conclusion:Frequency-specific resting-state features, particularly local synchronization in the slow-4 band, capture developmental-stage-related variation within ASD. The involvement of frontoparietal, visual, and subcortical networks suggests that developmental heterogeneity in ASD is supported by distributed reorganization of intrinsic brain activity. These findings highlight the potential of frequency-specific rs-fMRI metrics as candidate markers for characterizing neurodevelopmental stages in ASD, warranting further validation in longitudinal and independent cohorts.

Background/Objectives: Autism Spectrum Disorder (ASD) is a complex neurodevelopmental condition characterized by differences in social communications and restricted, repetitive patterns of behaviors and interests, affecting approximately 1% of children globally. While functional magnetic resonance imaging (fMRI) has provided insights into altered brain connectivity patterns in ASD, classification based on neuroimaging remains a challenging due to the heterogeneity of the disorder and variability in imaging data across sites. This study employs a network-based approach using large-scale, multi-site rs-fMRI dataset from the Autism Brain Imaging Data Exchange (ABIDE I and II) to classify ASD and healthy controls using machine learning. Methods: A semi-blind Independent Component Analysis method, specifically the spatial constraint reference ICA, is applied to identify functional brain networks, and the ComBat harmonization technique is used to address site-specific variability across 11 independent datasets, ensuring consistency in feature representation. Support Vector Machines (SVMs) are employed for classification, focusing on three key networks: the Default Mode Network (DMN), Sensorimotor Network (SMN), and Visual Sensory Network (VSN). Results: The results demonstrate high classification accuracy, with the VSN achieving the highest performance (83.23% accuracy, 87.90% AUC), followed by the DMN (81.43% accuracy, 84.53% AUC) and the SMN (80.52% accuracy, 84.96% AUC), positioned with their recognized roles in social cognition and sensory-motor processing, respectively. Conclusions: The integration of ICA-based feature extraction with ComBat harmonization significantly improved classification accuracy compared to previous studies. These findings point out the potential of network-based approaches in ASD classification and point out the importance of integrating multi-site neuroimaging data for identifying reproduceable network-level features.

Background:Eye gaze provides crucial insights into perceptual and cognitive processes during naturalistic movie viewing, yet concurrent eye tracking is often unavailable in functional MRI (fMRI) research. While deep learning models can estimate gaze directly from fMRI eyeball signals, their out-of-the-box generalizability across heterogeneous datasets requires empirical evaluation. Methods:We applied a specific pre-trained model from the DeepMReye framework in a zero-shot setting (without dataset-specific fine-tuning) to estimate gaze during movie watching across three independent fMRI datasets. Model accuracy was evaluated against camera-based eye-tracking data and via inter-subject correlations. Furthermore, we derived eye-movement-related time series from the predicted gaze signals to map their associated brain activation. Results:At the individual level, predicted gaze showed modest correspondence with measured ground-truth data (r ≈ 0.24-0.37), yielding brain activation maps largely restricted to the visual cortex. In contrast, group-averaged gaze predictions exhibited substantially higher reliability (r ≈ 0.73-0.84). First-level general linear models (GLMs) derived from group-averaged predictions successfully revealed widespread activation across established oculomotor control regions, including the frontal and parietal eye fields. Exploratory analyses of age-related effects on gaze prediction and brain activity yielded inconsistent results across datasets. Conclusions:Under a zero-shot implementation, the pre-trained model exhibits limitations for individual-level inference, likely reflecting the absence of dataset-specific training. However, group-averaged fMRI-based gaze estimates successfully capture shared viewing behaviors and robustly support the investigation of eye-movement-related brain activity. These findings inform the appropriate use of fMRI-based gaze decoding for naturalistic neuroimaging datasets lacking ground-truth eye-tracking logs.

Neurobiological and neurodegenerative diseases are inherently multifactorial, arising from coupled influences spanning genetic susceptibility, brain alterations, and environmental and behavioral factors. Multimodal modeling has therefore been increasingly adopted for disease diagnosis by integrating complementary evidence across data sources. However, in both large-scale cohorts and real-world clinical workflows, modality coverage is often incomplete, making many multimodal models brittle when one or more modalities are unavailable. Existing approaches to incomplete multimodal diagnosis typically rely on group-wise or static priors, which may fail to capture subject-specific cross-modal dependencies; moreover, many models provide limited interpretability into which evidence sources drive the final decision. To address these limitations, we propose Conditional Evidence Reconstruction and Decomposition (CERD), a framework for interpretable multimodal diagnosis with incomplete modalities. CERD first reconstructs missing modality representations conditioned on each subject's observed inputs, then decomposes diagnostic evidence into shared cross-modal corroboration and modality-specific cues via logit-level attribution. Experiments on the Alzheimer's Disease Neuroimaging Initiative (ADNI) demonstrate that CERD outperforms competitive baselines under incomplete-modality settings while producing structured and clinically aligned evidence attributions for trustworthy decision support.

The cerebellar role in various cognitive functions other than motor coordination has been gradually recognized, while its functional architecture and interaction with white matter functional networks (WM-FNs) remain unclear. The study combined resting-state functional connectivity and K-means clustering methods to obtain nine WM-FNs and seven gray matter functional networks (GM-FNs) by using the test-retest neuroimaging dataset collected from human connectome project. Subsequently, adopting winner-take-all algorithm and two distinct connectivity-based parcellations, we identified two parcellation maps of the cerebellum that corresponded to WM- and GM-FNs, respectively. We observed the cerebellar parcellations with unique spatial distribution pattern, which corresponds to white matter and gray matter functional systems. Additionally, the distinct WM-FNs exhibited high functional connectivity with GM-FNs, which corresponded to the overlapping maps between their cerebellar sub-regions well, indicating that intrinsic functional connectivity has an important constraint on the topological structure of cerebellum. Our finding created a new cerebellar parcellation atlas, advancing the mechanisms’ understanding of interaction between cerebellar organization and functional systems.