Imaging Neuroscience

High spatio-temporal resolution is crucial for neuroimaging techniques to improve our understanding of human brain function. While the fMRI signal is slow and shows a spread in latencies over space, the precision of hemodynamic response latency for each voxel is preserved and has been shown to be able to detect oscillatory hemodynamic changes approaching 1 Hz, suggesting its potential to reveal rapid neural dynamics. To examine how fast neural information can be derived from fMRI signals, we performed experiments that acquire high-field (7T) fMRI signals at an ultra-fast sampling rate (TR = 125 ms) from the visual cortex while participants observed naturalistic object stimuli. We applied multivariate pattern decoders to extract presented object-category information from the acquired signals at each sampling time after stimulus onset. Results showed that decoding accuracy rose above statistical significance less than 2 s after signal onset, faster than the peak latency of the hemodynamic response. The peak latency of the decoding accuracy was independent of variations in the hemodynamic latency of voxels used for decoding. The application of sparse decoders further revealed that rapid and accurate decoding was possible by pruning vein-rich voxels off from the multivariate voxel input to the decoders. These results suggest that a combination of ultra-fast sampling and multivariate decoding allows fast and temporally precise analysis of neural activity using fMRI signals. Significance statementFunctional magnetic resonance imaging (fMRI) is the most successful method to evaluate human brain function at fine spatial scales but is thought to lack temporal resolution because of slow hemodynamics. We challenge this conventional notion by fast sampling of fMRI signals, combined with multivariate decoding to extract information content represented in the fMRI signals. Results showed that the information content can be extracted faster than the magnitude of hemodynamic response rises, independent of the large spatial variation of hemodynamic latencies across individual voxels, and even after removing venous voxels from decoders. Our method is thus effective at filtering out slow and variable hemodynamic components, leading to the extraction of rapid and temporally precise components reflecting neuronal activity in human fMRI signals.

Brain stimulation with novel 40 Hz invisible spectral flicker (ISF) has been proposed as a therapy for Alzheimers disease, with leading hypotheses suggesting local promotion of glymphatic clearance of amyloid as the mechanism of action. Neural signals in the gamma range and their spatial propagation over the brain can be tracked using magnetoencephalography (MEG) with high temporal and good spatial detail. However, stimulation with 40 Hz ISF requires specialised hardware which causes electromagnetic interference (EMI) with MEG equipment. MEG measures the tiny magnetic fields of the brain, which are easily distorted by external magnetic fields. Using MEG to track the propagation of 40 Hz ISF requires multiple modifications to the experimental setup. These include, at least 1) an experimental design that promotes modulation of the neural signal, but not the artifact, 2) removal of all electronics not strictly necessary for light production from the stimulators, and 3) signal processing for 40 Hz EMI artifact suppression. Here, we present an MEG study on the cortical propagation of 40 Hz ISF. In two experiments, we investigate the modulation that visual and non-visual cognitive tasks have on the power and propagation of cortical activity induced by 40 Hz ISF. With the chosen experimental setup and design, the 40 Hz EMI artifact could not be entirely disentangled from the neural signal of interest, thus rendering inference on the spatial propagation of the 40 Hz signal impossible. Further improvements need to be implemented in a follow-up experimental design. We present potential solutions to allow for future investigation of 40 Hz ISF with MEG.

The reliance of fMRI research on between-person comparisons is limited by low test-retest reliability and inability to explain within-person processes. Intraindividual studies are needed to understand how changes in brain functioning relate to changes in behavior. Here, we present open data and analysis of a novel intensively sampled fMRI study, the Night Owls Scan Club. This precision imaging dataset includes 44 sessions acquired across four participants at a twice-weekly rate. In each session, participants completed multiple reward-related tasks, mood and alertness ratings, and a behavioral mood manipulation. We examined how the reward response reflects between-person or within-person variance. Test-retest-reliability of the reward response was very low and not explained by measurement error, suggesting little utility for between-person comparisons. At an intraindividual level, the mood induction showed small increases in the reward anticipation response. Additionally, mood and alertness explained notable intraindividual variance of the reward response, including as much as 31% for one participant. Overall, results suggest that BOLD activation to reward tasks - and likely other fMRI tasks - is more appropriate for within-person study than between-person study, highlighting a need for intensive longitudinal neuroimaging designs.

Inscapes is a low demand abstract animation used as an alternative to eyes open rest in neuroimaging studies, particularly with pediatric and clinical populations prone to head motion. Although prior work has established that functional connectivity patterns during Inscapes closely resemble those during rest, no study has examined whether the two conditions differ in aperiodic neural activity, a broadband feature of the power spectrum linked to excitation/inhibition balance. Here we used magnetoencephalography (MEG) in 54 healthy adults to compare spectrally parameterised aperiodic and periodic measures between eyes open rest and Inscapes viewing (visual component only, without audio). At the sensor level, both the aperiodic exponent and offset were significantly higher during rest than during Inscapes across widespread frontoparietal and occipital distributions in both magnetometers and gradiometers. Source level analyses at both the parcellation and vertex levels largely supported these patterns. The pericalcarine cortex was a notable exception, where both aperiodic measures were higher during Inscapes than during rest, indicating a regionally specific reversal in primary visual cortex. These results demonstrate that Inscapes and eyes open rest produce distinct aperiodic spectral profiles, indicating that the two conditions are not interchangeable for analyses involving broadband spectral dynamics or excitation/inhibition balance estimation.

Extracorporeal membrane oxygenation (ECMO) provides life support for severe, reversible cardiac or respiratory failure but carries substantial risk of neurological complications. Pediatric ECMO patients are particularly vulnerable, with over half of survivors exhibiting abnormal neuroimaging findings following discharge. Currently available clinical neuroimaging tools are limited, offering either static anatomical snapshots (ultrasound, computed tomography) or sparse functional monitoring (electroencephalography, functional near infrared spectroscopy) with limited spatial specificity. High-density diffuse optical tomography (HD-DOT) addresses these limitations to provide noninvasive, longitudinal, wide-field measurements of changes in cortical oxygenation at the beside. Here, we investigate safety and feasibility for bedside longitudinal HD-DOT monitoring of cerebral oxygenation focusing on data collected over 20 days in seven pediatric ECMO patients. Results confirm the reliable acquisition of high-quality HD-DOT data without adverse events, establishing HD-DOT as a promising tool for continuous, and safe bedside neuroimaging in this population.

Establishing the reliability of spinal cord functional magnetic resonance imaging (fMRI) is critical before employing it to assess experimental or clinical interventions. Previous studies have mapped human motor activity primarily to the ipsilateral ventral horn, aligning with myotomal and dermatomal projections. Despite these insights, the test-retest reliability of spinal fMRI remains under-investigated. Here we assessed spinal cord activation during a sensorimotor paradigm involving right-hand grasping and grip force estimation in 30 healthy volunteers. Participants completed two identical scanning visits, each time performing the same task twice, enabling the investigation of test-retest reliability both within a single experimental visit and between visits performed on different days. Aggregating all task runs, motor-evoked activation was observed in ipsilateral ventro-dorsal regions of spinal segmental levels C5-T1, as well as in medial regions of levels C2-C3. Despite highly reliable task performance (grip force) and fMRI signal quality (temporal signal-to-noise ratio), the reliability of motor activation was predominantly poor-to-fair both within and between visits, with notable variability in spatial distribution observed across task runs. Increasing the number of task runs per individual improved the robustness of group-level activation, as indexed by higher activated voxel count, larger cluster spatial extent, and attenuated t-statistic distribution. Although we demonstrated that motor-evoked activation corresponds to the known neuroanatomical organisation of motor circuits, its low test-retest reliability presents a challenge for wider applications of spinal fMRI. Understanding the drivers of low reliability in functional imaging is warranted, but we suggest that looking beyond measurement error is required, including careful consideration of inherent within-individual variability underpinned by neurophysiological and psychological factors.

Laminar fMRI using GE-BOLD is vulnerable to spatial blurring from intracortical veins, while multivariate pattern analysis (MVPA) is often assumed to mitigate these biases. Yet, this assumption has not been systematically investigated. We thus developed a mechanistic laminar response model that simulates voxel-wise patterns across cortical depths, incorporating a vascular draining model. We conducted simulations in which the ground-truth signal originated in a single, several, or across all layers, and applied standard MVPA decoding before and after deconvolution of the draining effect. Decoding accuracies were consistently influenced by draining veins: deep-origin signals yielded above-chance decoding in superficial layers, and null scenarios produced false positives in middle or deep layers. Vascular deconvolution enhanced specificity in single-layer cases but did not resolve ambiguities in null decoding profiles. Simulating six thinner layers improved decoding accuracies, especially in the deconvolved signal scenarios. These findings demonstrate that multivariate techniques are not inherently immune to vascular biases, but also demonstrate that careful modeling can help correct draining effects.

Understanding the complex mechanism of human brain function requires innovative approaches that capture the intricate interplay of electrical, chemical, and hemodynamic activities. We developed the PMEEN system, named for its concurrent integration of PET, MRI, EEG, Eye Tracking, and fNIRS modalities by successfully addressing the electromagnetic and gamma ray interference among these modalities and incorporation of centralized clock control to allow simultaneous spatial registration and temporal synchronization. Here we show that PMEEN enables the concurrent acquisition of metabolic, structural, electrical, hemodynamic, and behavioral data, providing a comprehensive view of brain function that was previously inaccessible. By successfully validating the system in healthy volunteers and clinical cases involving patients with major depressive disorder, Alzheimers disease, and epilepsy, we demonstrate its potential for advancing both research and clinical diagnostics. These findings represent a significant technological step forward in neuroimaging, allowing for a holistic understanding of the neural mechanisms underlying cognition and neurological disorders, and setting the stage for future breakthroughs in brain research and therapy. SummaryWe introduce the PMEEN system, a groundbreaking multi-modal neuroimaging platform integrating PET, MRI, EEG, Eye tracking, and fNIRS. This system allows concurrent acquisition of metabolic, structural, electrical, hemodynamic, and behavioral data, providing unprecedented insights into brain function. Through advanced electromagnetic compatibility design and synchronized data acquisition, PMEEN overcomes the technical challenges of integrating these modalities. Validated in phantom, healthy volunteers and patients with major depressive disorder, Alzheimers disease, and epilepsy, PMEEN has demonstrated its capability for capturing comprehensive neural dynamics. This system paves the way for new research opportunities and clinical applications in neurological and psychiatric diagnostics.

Ultrahigh-density electrocorticography (ECoG) provides unprecedented spatial resolution for recording cortical electrical activity. This study uses simulated scalp projections from an ECoG recording to challenge the assumption that channel-level electroencephalography (EEG) reflects only local field potentials near the recording electrode. Using a 1024-electrode ECoG array placed on the primary motor cortex during finger movements, we applied Adaptive Mixture Independent Component Analysis (AMICA) to decompose activity into maximally independent grid activity components and projected these to 207 simulated EEG scalp electrode channels using a high-definition MR image-based electrical forward-problem head model. Our findings demonstrate how cortical surface-recorded potentials propagate to scalp electrodes both far from and near to the generating location. This work has significant implications for interpreting both EEG and ECoG data in clinical and research applications. Clinical RelevanceThis study provides insights for interpreting scalp EEG data, demonstrating that scalp channel activity represents a complex mixture of distributed cortical source activities rather than primarily activity generated nearest to the scalp electrodes. These findings may hopefully spur improvement in EEG-based diagnostics for neurological disorders.

The perinatal period involves transitioning from an intra- to an extrauterine environment, which requires a complex adaptation of the brain. This period is marked with dynamic and multifaceted cortical changes in both structure and function. Most studies to date have focused either on the fetal or postnatal period, independently. To the best of our knowledge, this is the first neurodevelopmental study targeting the cortical trajectory of typically developing perinatal subjects, combining MRIs from both fetal and postnatal participants. Prior to analysis, preprocessing and segmentation parameters were harmonized across all subjects in order to overcome methodological limitations that arise when studying such different populations. We conducted a normative modeling analysis on a sample of 607 subjects, age ranged 24 to 45 weeks post-conception, to observe changes that arise as participants traverse the birth barrier. We observed that the trajectories of global surface area and several volumetric features, including total gray matter, white matter, brainstem, cerebellum and hippocampi, follow distinct but continuous patterns during this transition. We further report three features presenting a discontinuity in their neurodevelopmental trajectories as participants traverse from a fetal to a postnatal environment: the extra-cerebrospinal fluid volume, the ventricular volume and global gyrification. The current study demonstrates the presence of unique neurodevelopmental patterns for several structural features during the perinatal period, and confirms that not all features are affected in the same way as they cross the birth barrier. SIGNIFICANCE STATEMENTThe perinatal phase comprises the fetal and immediate postnatal period, and is generally described as the time surrounding birth. Comprehensively understanding this period is crucial due to the presence of dynamic and multifaceted brain changes. What makes this investigation unique is that it is the first neurodevelopmental study, to the best of our knowledge, focused on the cortical trajectory of typically developing perinatal subjects through the combination of both fetal and postnatal participants into one analysis. We report that certain brain feature trajectories change drastically as fetuses become newborns, while other features remain continuous. These observations are relevant in both the isolation of biomarkers for later cognitive and physiological disorders and in the understanding of typical cerebral development.

Brain activity comprises both rhythmic (periodic) and arrhythmic (aperiodic) components. These signal elements vary across healthy aging, and disease, and may make distinct contributions to conscious perception. Despite pioneering techniques to parameterize rhythmic and arrhythmic neural components based on power spectra, the methodology for quantifying rhythmic activity remains in its infancy. Previous work has relied on parametric estimates of rhythmic power extracted from specparam, or estimates of rhythmic power obtained after detrending neural spectra. Variation in analytical choices for isolating brain rhythms from background arrhythmic activity makes interpreting findings across studies difficult. Whether these current approaches can accurately recover the independent contribution of these neural signal elements remains to be established. Here, using simulation and parameter recovery approaches, we show that power estimates obtained from detrended spectra conflate these two neurophysiological components, yielding spurious correlations between spectral model parameters. In contrast, modelled rhythmic power obtained from specparam, which detrends the power spectra and parametrizes brain rhythms, independently recovers the rhythmic and arrhythmic components in simulated neural time series, minimising spurious relationships. We validate these methods using resting-state recordings from a large cohort. Based on our findings, we recommend modelled rhythmic power estimates from specparam for the robust independent quantification of rhythmic and arrhythmic signal components for cognitive neuroscience.

Physiological time-series data, like electroencephalography (EEG), are vulnerable to motion, ocular, and muscle artifacts that hinder real-time inference and bias offline analyses. We present the Minds AI Filter: a lightweight, physics-based, sensor-fusion method that exploits multichannel spatial structure and band-aware synchrony to enhance neural activity while suppressing non-neural noise. The nomenclature "AI" reflects integration within a larger artificial-intelligence pipeline; the filter itself requires no prior training or deep learning. A single tuning parameter controls filter strength. The design supports streaming windows ({approx}1 s) with minimal added latency and extends naturally to longer offline segments; leveraging a sensor-fusion design across channels, it suggests applicability to other neurophysiological time-series, such as MEG and ECoG, pending further validation; exploratory incorporation of EOG/ECG as auxiliary signals is a potential avenue for future filter advancements. We evaluate the approach across multiple devices and public datasets, assessing both down-stream AI classification performance and real-time signal-quality metrics. In both real-time and offline settings, the filter performed better on dynamic artifacts and noise than baseline and commonly used alternatives in our evaluations. When applied in conjunction with other methods, it was only observed to improve downstream accuracy, never reduce it, when any effect was present. Denoising is quantified using SNR-like measures, and ablations isolate the roles of spatial coupling and band weighting. Artifact-specific analyses (ocular bursts, head tilt, jaw clench) and latency profiling on commodity hardware are included. These results indicate that a lightweight, synchrony-aware filter can robustly stabilize real-time EEG and systematically improve downstream AI classification. The method is compatible with standard preprocessing but does not depend on it.

Functional connectivity magnetic resonance imaging (FC-MRI) has been widely used to investigate neurodevelopment. However, FC-MRI is vulnerable to head motion, which is associated with age and distorts FC estimates. Numerous preprocessing strategies have been developed to mitigate confounds, each with advantages and drawbacks. Preprocessing strategies for FC-MRI have typically been validated and compared using resting state data from adults. However, FC-MRI in young children presents a unique challenge due to relatively high head motion and a growing use of passive viewing paradigms to mitigate motion. This highlights a need to compare processing choices in pediatric samples. To this end, we leveraged longitudinal, passive viewing fMRI data collected from 4 to 8-year-old children. We systematically investigated combinations of widely used and debated preprocessing strategies, namely global signal regression, volume censoring, ICA-AROMA, and bandpass filtering. We implemented commonly used metrics of noise removal (i.e. quality control-functional connectivity), metrics sensitive to individual differences (i.e. connectome fingerprinting), and, because data was collected during a passive viewing task, we also assessed the impact on stimulus-evoked responses (i.e. intersubject correlations; ISC). We found that the most efficacious pipeline included censoring, global signal regression, bandpass filtering, and head motion parameter regression. Despite the drawbacks of noise-mitigation steps, our findings show benefits for both noise removal and information retention in a high-motion early childhood sample. Highlights- We evaluated 27 preprocessing pipelines in passive viewing data from young children - Pipelines were evaluated on noise-removed and information retained - Pipelines that included censoring and GSR outperformed alternatives across benchmarks - For high-motion scans, preprocessing choices substantially alter connectomes

Cognitive functions depend on neuronal communication, which is subserved by the synchronization of neuronal rhythms. Rhythms are characterized by their frequency, power and phase. If the phase of a rhythm just preceding an input is predictive of the neuronal or behavioral response to the input, this provides strong evidence for a functional role of the rhythm. Yet, this requires estimating the phase of a rhythm at the edge of the epoch. This is challenging, because any phase estimation that is spectrally specific requires a finite window length often combined with tapers that de-emphasize the signal close to the edge. To overcome this, we propose a method that builds on previously described approaches based on autoregressive modeling of the data and corresponding extrapolation beyond the edge. In contrast to related previous approaches, the modeling is based on the broadband signals, avoiding filtering-related group delays, and the extrapolation is performed multiple times, allowing averaging and thereby the reduction of extrapolation noise. The new method provided more accurate phase estimation at the edge for most simulated datasets, and for an empirical dataset from awake macaque area V4. We propose that the enhanced phase estimation accuracy at the edge might help to investigate the functional roles of brain rhythms and potentially also to improve phase-specific stimulation for clinical applications.

Multiband (MB) acceleration of functional magnetic resonance imaging has become more widely available to neuroscientists. Here we compare MB factors of 1, 2 and 4 while participants view complex hand actions vs. simpler hand movements to localize the action observation network. While in a previous study, we show that MB4 shows moderate improvements in the group-level statistics, here we explore the impact it has on single subject statistics. We find that MB4 provides an increase in p values at the first level that is of medium effect size compared to MB1, providing moderate evidence across a number of voxels that MB4 indeed improves single subject statistics. This effect was localized mostly within regions that belong to the action observation network. In parallel, we find that Cohens d at the single subject level actually decreases using MB4 compared to MB1. Intriguingly, we find that subsampling MB4 sequences, by only considering every fourth acquired volume, also leads to increased Cohens d values, suggesting that the FAST algorithm we used to correct for temporal auto-correlation may over-penalize sequences with higher temporal autocorrelation, thereby underestimating the potential gains in single subject statistics offered by MB acceleration, and alternative methods should be explored. In summary, considering the moderate gains in statistical values observed both at the group level in our previous study and at the single subject level in this study, we believe that MB technology is now ripe for neuroscientists to start using MB4 acceleration for their studies, be it to accurately map activity in single subjects of interest (e.g. for presurgical planning or to explore rare patients) or for the purpose of group studies.

Harmonisation is widely used to mitigate site- and scanner-related batch variability in multisite neuroimaging studies and is particularly critical in longitudinal clinical trials, where detection of subtle biological or treatment-related changes depends on reliable measurement across scanners and timepoints. However, the effectiveness of harmonisation in small, heterogeneous clinical datasets remains insufficiently understood, particularly in relation to subject-level variability and consistency across acquisition settings, and its impact on both removal of technical variability and preservation of biological variation in pooled multisite analyses. We systematically evaluated a range of image-based and statistical harmonisation methods using a clinically realistic multisite, multiscanner structural T1-weighted (T1w) MRI test-retest dataset comprising three controlled acquisition scenarios: repeatability, intra-scanner reproducibility and inter-scanner reproducibility. Methods were applied under different batch specifications (site, scanner, or both) and performance was assessed within each scenario and in pooled data using a multi-metric framework capturing both technical and biological variability in volumetric imaging-derived phenotypes (IDPs) relevant to aging and dementia research. Across IDPs, before harmonisation variability was lowest in the repeatability scenario (median variability=0.6 to 2.7%, rank consistency {rho} [≥]0.9), with modest increases under intra-scanner reproducibility (0.5 to 3.2%, {rho}=0.5 to 1.0) and substantially greater variability under inter-scanner reproducibility conditions (1.7 to 19.2%, {rho} =-0.1 to 0.9). These results offer important information to consider for multisite study design, including sample size calculation in clinical trials. Harmonisation performance was strongly context dependent, with clearer benefits emerged in inter-scanner scenarios where both variability reduction and improvements in subject-level consistency were observed. In pooled data, approaches that explicitly modelled site as batch and accounted for repeated-measure structure showed greater consistency across IDPs in batch effect mitigation and more accurately reflected underlying biological variation. Our evaluation metrics enabled disentangling the removal of global batch effect while highlighting residual variability at the phenotype-specific or multivariate levels. These findings demonstrate that harmonisation cannot be treated as a one-size-fits-all solution and must be interpreted relative to the acquisition context, dataset structure, and downstream analytic goals. Multi-metric evaluation under realistic clinical constraints is essential to support reliable and translatable neuroimaging inference by ensuring appropriate correction of batch effects while preserving longitudinal biological signals and sensitivity to clinically meaningful change in multisite studies.

IntroductionPreterm birth is a major risk factor for disrupted brain development and subsequent neurodevelopmental disorders, yet the underlying mechanisms remain poorly understood. Further, typical neuroimaging analyses are particularly challenging in the neonatal brain: data is frequently low quality and a lack of cellular development violates the assumptions relied on by many commonly-used techniques. In this study, we develop and present an advanced diffusion magnetic resonance imaging method to examine the microstructural organization of white matter in a clinically-acquired cohort of premature neonates. MethodsUsing a novel approach that resolves multiple tissue compartments within the brain, we provide highly detailed orientation and quantification of white matter fibers and tissue signal fraction. We also utilize a series of automated segmentation algorithms to identify and measure these metrics across key tracts and subcortical regions. We investigate how these measures relate to postmenstrual age, as well as to clinical factors reflecting neonatal illness severity. ResultsWe report successful segmentation and reconstruction of numerous white matter tracts throughout the neonatal brain. We further demonstrate the utility and functionality of microstructural analysis in a variety of pathologies commonly encountered in the neonatal clinical environment. Our results demonstrate tract-specific developmental trajectories, with early-maturing pathways showing higher microstructural organization. Exploratory analyses suggest that neonatal illness severity has modest, tissue-specific associations with microstructural properties. DiscussionThis work demonstrates that advanced microstructural imaging methods can extract meaningful white matter measurements from clinically-acquired scans, providing a practical framework for studying neonatal brain development in real-world hospital settings. These metrics are able to be calculated at extremely young ages, potentially allowing non-invasive study of vulnerable populations before detailed behavioral or neurological assessments are feasible.

Diffusion tensor imaging (DTI) has advanced our understanding of how brain microstructure evolves over development. However, the proliferation of multi-shell diffusion imaging sequences has coincided with notable advances in the modeling of neuronal diffusion patterns, such as Neurite Orientation Dispersion and Density Imaging (NODDI) and Laplacian-regularized Mean Apparent Propagator MRI (MAPL). The relative utility of these newer diffusion models for understanding brain maturation remains sparsely investigated. Additionally, despite evidence that motion artifact is a major confound for studies of development, the relative vulnerability of these models to in-scanner motion has not been described. Accordingly, in a sample of 123 youth (ages 12-30) we evaluated DTI, NODDI, and MAPL for associations with age and in-scanner head motion at multiple scales, including mean white matter values, voxelwise analyses, and tractography-based structural brain networks. Our results reveal that multi-shell diffusion imaging sequences can be leveraged to robustly characterize neurodevelopment, even within the framework of DTI. However, these metrics of diffusion are variably impacted by motion, highlighting the importance of modeling choices for studies of movement-prone populations. Our findings suggest that while traditional DTI is sensitive to neurodevelopmental trends, contemporary modeling techniques confer key advantages for neurodevelopmental inquiries.

Normative brain growth charts in early life hold great promise for furthering basic and clinical science. We leverage the rapid, substantial development of visual cortex function that is indexed by visual-evoked potentials (VEP) in electroencephalography to create longitudinal normative growth curves of task-related brain function with 1374 observations contributed by 802 infants (57 to 579 days old) from South Africa, Brazil, and the United States. Site-specific models were cross-validated and showed excellent fits to other sites samples, demonstrating functional growth curves generalize across contexts robustly. Deviations from the normative growth models associated with early environmental and behavioral measures such as prenatal exposures and postnatal cognition. These findings demonstrate the utility of using functional growth charts to understand and potentially act on individual neurodevelopmental trajectories. VEP brain function growth charts represent a new direction for EEG research to support healthy brain development globally.

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 {approx} -0.38 to 0.67). In contrast, group-averaged gaze predictions exhibited substantially higher correlations (r {approx} 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.