The research focused on the isointense stage of infant brain development, when gray and white matter have similar intensities in both T1- and T2-weighted MRI. This low contrast, together with myelination, noise and motion-related errors, makes automated tissue separation difficult. The authors said the benchmark was particularly challenging because ongoing myelination causes substantial overlap between the intensity distributions of gray and white matter.

The study proposed a 3D patch-wise U-Net with weighted-average reconstruction. Each MRI volume was divided into fixed-size 64 × 64 × 64-voxel patches with a stride of 32 voxels along each dimension, creating 50% overlap between adjacent patches. Patch-level segmentation reduced the memory required to process a complete high-resolution volume while retaining local anatomical information.

The pipeline comprised MRI loading and depth-wise cropping, patch creation, intensity normalization and label mapping, patch-wise segmentation, and reconstruction of the complete 3D volume. Normalized image intensities were scaled to the range from zero to one by dividing each patch by its maximum voxel value. Labels were remapped to consecutive indices for cerebrospinal fluid, gray matter and white matter.

The U-Net had a four-level encoder-decoder structure with skip connections. Its encoder used paired 3 × 3 × 3 convolutions, batch normalization and ReLU activation, followed by 2 × 2 × 2 max pooling. Feature channels increased from 8 to 64 across the encoder, while a two-layer bottleneck operated at the lowest resolution. The decoder used 3D transposed convolutions to restore spatial resolution and concatenated encoder features through skip connections. A final 1 × 1 × 1 softmax layer produced probabilities for background, gray matter, white matter and cerebrospinal fluid.

After separate patch predictions were made, overlapping regions were combined with spatial weights based on each voxel’s distance from the center of the patches covering it. The paper states that voxels near patch centers received greater weight, while voxels near patch boundaries received less, producing smoother transitions and reducing stitching artifacts. This reconstruction method was designed to address boundary discontinuities that can arise when separately processed patches are joined.

Training and evaluation used co-registered T1- and T2-weighted MRI from six-month-old infants acquired on a Siemens 3T scanner at 1 mm isotropic resolution. The dataset contains 10 training participants and 13 official test subjects, with expert manual segmentations of the three tissues. The training participants were divided subject-wise into nine for training and one for validation, keeping all patches from each participant within a single set. Although both modalities were used in preprocessing, segmentation was conducted primarily on T1-weighted images.

The model was trained for 1,000 epochs with a combined Dice and cross-entropy loss, the AdamW optimizer and a cosine-annealed learning rate. Training used mixed-precision arithmetic on an NVIDIA GeForce RTX 4080 with 12 GB of memory. The reported training footprint was about 1.9 GB, rising to 4.2 GB when activations and gradients were included, and convergence took approximately 50 to 65 hours. Inference for a 256 × 256 × 192 volume took about 2.5 seconds, with another 0.5 seconds for patch reconstruction.

On the 13-subject test set, the model achieved 95% overall voxel-wise accuracy, a weighted-average F1 score of 0.95 and a macro-average F1 score of 0.87. Gray matter had precision of 0.93, recall of 0.90 and F1 of 0.92; white matter had precision of 0.80, recall of 0.82 and F1 of 0.81; and cerebrospinal fluid had precision of 0.75, recall of 0.73 and F1 of 0.74. The reported 95% confidence ranges for the tissue-specific F1 scores did not overlap, and the paper reported statistically significant differences for all pairwise comparisons at p < 0.05.

The results indicate a trade-off between volumetric overlap and boundary precision. Some comparison methods reported higher Dice scores for white matter and cerebrospinal fluid, but the proposed model produced smaller HD95 and average surface distances in the comparisons presented. Performance was weaker for cerebrospinal fluid, whose thin and irregular structures are affected by partial-volume effects and ambiguous boundaries. Generalizability remains untested beyond the benchmark’s single scanner, single age point and small cohort, with no direct validation in other age groups, scanners or multicenter data. The authors proposed future work on patch and stride sensitivity, interpretability, hybrid CNN-transformer models and workflow integration, but none of those extensions had been evaluated in this study.

Subject of Research: Automated segmentation of infant brain tissues in six-month-old MRI using a patch-wise 3D U-Net

Article Title: Deep learning model for infant brain tissue segmentation using 3D patch-wise U-Net with weighted-average reconstruction

Article References: Mhlanga, S., & Viriri, S. (2026). Deep learning model for infant brain tissue segmentation using 3D patch-wise U-Net with weighted-average reconstruction. Discover Artificial Intelligence, 6(1), Article 1277. https://doi.org/10.1007/s44163-026-01982-x

Image Credits: AI Generated

DOI: 10.1007/s44163-026-01982-x

Keywords: infant brain MRI, tissue segmentation, 3D U-Net, patch-wise learning, weighted reconstruction, iSeg-2017, gray matter, white matter, cerebrospinal fluid, boundary accuracy, Deep, learning

Researchers built an automated pipeline that first delineates MS plaques on fluid-attenuated inversion recovery, or FLAIR, MRI and then examines diffusion tensor imaging, or DTI, along fiber pathways passing through those lesions. The DTI analysis used fractional anisotropy, which describes how directional diffusion is organized, and mean diffusivity, which reflects the overall mobility of water. Together, these measures can reveal tissue changes that are not apparent from lesion size alone, including axonal loss, myelin damage and edema, although they do not directly identify a specific biological process.

The lesion-detection component was a modified 2.5D U-Net trained and evaluated on 115 FLAIR scan series from 75 people with MS. It processed three neighboring slices together, combining local anatomical context with lower computational demand than a fully three-dimensional network. Residual blocks, attention gates and deep supervision were incorporated to improve feature propagation, emphasize subtle lesions and learn patterns at multiple scales. Patient-level splitting prevented scans from the same individual entering both training and testing sets, an important safeguard when some participants contributed several examinations.

At inference, the model generated 24 rotated and flipped predictions and combined them through a self-ensembled lesion-fusion strategy. The final configuration used a 0.5 intersection threshold, a 0.4 connected-union threshold and a minimum lesion size of 15 voxels. On the held-out test set, the network achieved a Dice similarity coefficient of 0.732, with a 95% confidence interval of 0.689 to 0.777. Its voxel-wise true-positive rate was 0.781 and lesion-wise true-positive rate was 0.831, while volume correlation reached 0.954, with a 95% confidence interval of 0.914 to 0.995.

Those detection figures were nominally higher than values reported for a contextual benchmark: a Dice coefficient of 0.682, voxel-wise true-positive rate of 0.613, lesion-wise true-positive rate of 0.542 and volume correlation of 0.864. The comparison was descriptive rather than a head-to-head trial, and the stronger sensitivity came with a trade-off. Positive predictive value was 0.710 in the proposed model, and the lesion-wise false-positive rate was 0.285. The authors therefore emphasized the high volume correlation, which indicates that the extra detections were relatively small and did not substantially distort estimated whole-brain lesion load.

For the microstructural analysis, the researchers reconstructed whole-brain fiber pathways using constrained spherical deconvolution and deterministic tractography. Streamlines were seeded in white matter with fractional anisotropy above 0.3, tracked with a 0.5-mm step size and a maximum turning angle of 25 degrees, and terminated when they entered voxels with fractional anisotropy below 0.20. Only streamlines between 30 and 250 mm were retained. The diffusion maps and reconstructed pathways were then registered into FLAIR space so that each intralesional segment could be assigned the mean fractional anisotropy and mean diffusivity of the voxels it traversed.

To define what counted as reference-like, the investigators used a subject-specific internal-capsule segment of the corticospinal tract. Atlas registration located this region, and any reconstructed pathway intersecting an MS lesion was excluded, leaving lesion-spared streamlines as the internal reference. Because the two DTI measures vary naturally between people, fractional anisotropy and mean diffusivity were normalized within each participant using the combined distribution of lesion-spared reference and intralesional segments. Euclidean distance was then used within that individualized feature space, rather than as a direct cross-patient biological scale.

Six operational approaches were evaluated: K-means, Gaussian mixture models, agglomerative clustering, spectral clustering, density-based clustering with noise, or DBSCAN, and an adaptive threshold method based on the reference distribution. After fitting each model, the cluster whose centroid was closest to the subject’s lesion-spared reference was designated the reference-like intralesional segment set. The percentage of lesion-crossing segments assigned to that set became the proposed measure. Results depended markedly on the algorithm: median reference-like proportions were 26% for K-means and 22% for the Gaussian mixture model, whereas DBSCAN and agglomerative clustering sometimes produced estimates near either 0% or 100%.

Across all six methods, no association with macroscopic burden survived statistical correction. For total lesion volume, Spearman correlation coefficients ranged from −0.04 to 0.25, but all adjusted p-values were at least 0.161. For lesion count, the coefficients ranged from −0.01 to 0.20, with all adjusted p-values at least 0.212. K-means and Gaussian mixture models produced the most stable proportions, while spectral clustering showed a bimodal-like pattern and sometimes classified every tract as reference-like. The study therefore supports a dissociation between the amount of visible lesion and the operationally measured microstructural similarity within lesions, rather than proving that either quantity predicts clinical outcomes.

Several limitations constrain how the result should be interpreted. The study was a retrospective secondary analysis of two public, de-identified imaging datasets, and the DTI cohort contained no patient-level disability scores or cognitive assessments, preventing direct testing of the proposed index against symptoms. Its sample of 100 people may also limit generalizability. In addition, comparison with the internal-capsule reference measures similarity to a highly intact, strongly anisotropic white-matter region; it is not a universal measure of healthy tract biology. Uncorrected spatial variations in diffusion encoding can bias mean diffusivity and tensor orientation, while segmentation false positives could influence downstream tract measurements. The investigators identified longitudinal studies with clinical outcomes, test–retest assessment and spatial diffusion calibration as necessary next steps before the metric can be evaluated as a predictor of disability or disease progression.

Subject of Research: Automated segmentation and diffusion-tensor analysis of microstructural heterogeneity in multiple sclerosis lesions

Article Title: Revisiting the MS paradox: automated 2.5D FLAIR segmentation and DTI microstructural profiling reveals dissociation between lesion burden and reference-like intralesional tract profiles

Article References: Toscano-Flores, E. A., Mireles-Ramírez, M. A., Guerrero-García, J. D. J., & Alvarez-Padilla, F. J. (2026). Revisiting the MS paradox: automated 2.5D FLAIR segmentation and DTI microstructural profiling reveals dissociation between lesion burden and reference-like intralesional tract profiles. Frontiers in Human Neuroscience, 20, Article 1901702. https://doi.org/10.3389/fnhum.2026.1901702

Image Credits: AI Generated

DOI: 10.3389/fnhum.2026.1901702

Keywords: multiple sclerosis, MRI lesion segmentation, diffusion tensor imaging, FLAIR, tractography, microstructure, clinico-radiological paradox, deep learning, white matter, Revisiting, paradox, automated

The raw material for the project came from the Adolescent Brain Cognitive Development Study, or ABCD, the largest long-term study of brain development and child health in the United States, funded by the National Institutes of Health. Because ABCD follows thousands of children over multiple years, it collects repeated scans from the same participants at different points in time, which is essential for distinguishing true developmental change from individual differences between people. The Penn-led team took the more than 24,000 scans gathered from over 10,000 adolescents across those longitudinal visits and subjected them to a rigorous, standardized processing pipeline, transforming raw images into analyzable measures of brain structure. The result is a curated dataset known as the ABCD-BIDS Community Collection, which the researchers have released so that scientists anywhere in the world can study adolescent brain development without repeating the enormous technical effort of processing the images themselves.

At the heart of the study is a technique called diffusion MRI, which does not simply photograph the brain but probes its microstructure. By measuring how water molecules move through tissue, diffusion MRI reveals the organization of white matter, the bundles of nerve fibers that connect distant brain regions and allow them to communicate. Because water diffuses more easily along the length of an axon than across it, the direction and speed of water movement encode information about fiber density, coherence, and integrity. Adolescence is a critical period for this wiring: the long-range connections that support attention, learning, emotion regulation, and executive function continue to mature well into the teenage years, and disruptions in these trajectories have been linked to the emergence of psychiatric illness during this window of life.

Conventional analyses of diffusion MRI typically rely on relatively simple summary measures, such as fractional anisotropy, which captures the overall directionality of water diffusion in a region of white matter. These measures have been workhorses of developmental neuroscience for decades, but they are inherently limited, because they collapse complex microstructural information into a single number. The Penn team instead applied more advanced diffusion models that extract richer, multidimensional features of white matter organization. What they found was striking: these sophisticated measures uncovered developmental changes across adolescence that were far less apparent with traditional metrics. In other words, some aspects of teenage brain maturation have been hiding beneath the resolution of the field’s most commonly used tools.

The technical achievement underlying these findings should not be understated. When tens of thousands of scans are collected on different scanners, running different software versions, at dozens of sites spread across the country, the technical differences between images can dwarf the subtle biological changes scientists hope to measure. As Meisler explained, those site and scanner effects can be larger than the developmental signals themselves, threatening to swamp any genuine pattern. To overcome this, the researchers turned to advanced open-source image-processing and statistical methods developed at Penn specifically for large, multisite neuroimaging studies. These pipelines harmonize the data, statistically accounting for differences in acquisition and quality so that scans from different machines can be meaningfully compared.

The payoff from this harmonization was clear. After correcting for technical differences, the team examined how multiple features of white matter changed across adolescence and found widespread, highly consistent developmental patterns that held up across all 21 study locations. Satterthwaite emphasized that despite substantial variation in image quality across tens of thousands of scans acquired by different scanners, the patterns of white matter development were remarkably consistent. This consistency matters enormously for the credibility of large-scale developmental neuroscience, because inconsistent findings across sites have long been a weakness of multisite imaging studies. Demonstrating that a carefully harmonized pipeline can produce reproducible developmental trajectories at this scale offers reassurance that the biological signals are real and not artifacts of measurement.

Beyond the scientific findings themselves, a central ambition of the project was reproducibility. Large research consortia generate datasets of extraordinary scale, but the processing required to make those data scientifically useful is often difficult, sometimes impossible, for individual laboratories to reproduce on their own. Meisler noted that by releasing both the processed data and the open-source tools used to create them, the team hopes other researchers can build on this work rather than starting from scratch. This philosophy reflects a broader movement in neuroscience toward open science, in which shared data, shared code, and shared pipelines allow results to be verified, extended, and reanalyzed in ways that closed research simply cannot permit. The dataset is openly accessible, lowering the barrier for researchers worldwide who study adolescent development but lack the computing infrastructure or consortium membership to generate such resources independently.

The implications of the resource extend well beyond white matter biology. Researchers in the PennLINC laboratory are already using the collection to investigate how white matter development relates to environmental exposures and other aspects of adolescent health. Because the ABCD Study also tracks participants’ environments, experiences, and behaviors in fine detail, the processed imaging data can be linked to a rich web of developmental context, from socioeconomic factors to substance use to academic outcomes. This opens the door to asking precisely how the events of adolescence, both beneficial and harmful, shape the maturing brain’s connectivity during a period when it is especially plastic. The consistency of the developmental patterns uncovered here also provides a normative backdrop against which atypical development, and the early emergence of psychiatric conditions, can be measured.

For the field of developmental neuroscience, the study signals a methodological turning point. It shows that state-of-the-art diffusion MRI, combined with harmonization methods built for scale, can extract developmental signals from real-world, multisite data that conventional approaches may miss. It also demonstrates that the perceived trade-off between scale and quality, between the statistical power of thousands of participants and the precision of carefully controlled imaging, can be resolved with the right processing tools. As similar open collections are built for other populations and age ranges, researchers anticipate a future in which normative charts of brain development, analogous to the growth curves pediatricians use for height and weight, become a routine benchmark for identifying when a young person’s brain development deviates from the expected path.

The ABCD-BIDS Community Collection thus represents both a dataset and a statement about how twenty-first-century neuroscience should be done: collaboratively, transparently, and at scale. By pairing one of the largest longitudinal imaging collections ever assembled with openly documented, reproducible processing, the Penn-led team has converted tens of thousands of raw scans into a durable scientific infrastructure. The consistent white matter trajectories it reveals illuminate one of the most dynamic periods of human brain maturation, while the resource itself ensures that the next generation of discoveries about the adolescent brain will not have to be built from the ground up. The findings were published in Neuron, and the processed data are freely available to the research community.

Subject of Research: White matter development in the adolescent brain measured with diffusion MRI from a large multiscale longitudinal neuroimaging dataset

Article Title: Advanced MRI dataset maps adolescent brain development

Article References: Advanced MRI dataset maps adolescent brain development. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: adolescent brain development, diffusion MRI, white matter, ABCD Study, neuroimaging, open science, brain connectivity, harmonization, Neuron, longitudinal study, University of Pennsylvania, data sharing

The first strand of evidence comes from toddlers in South Africa who were exposed to tobacco or alcohol before birth. Scholten and colleagues used diffusion tensor imaging, a technique that tracks the movement of water molecules through brain tissue to measure the integrity of white matter pathways, in 93 children aged two to three years. Surprisingly, children with prenatal substance exposure scored just as well as unexposed peers on standard language assessments. Yet their brains told a different story. The exposed toddlers showed lower mean diffusivity in the splenium of the corpus callosum, a major bundle of fibers connecting the brain’s hemispheres, and the relationship between white matter properties and language scores was altered across seven tracts. In exposed children, fractional anisotropy, a marker of fiber organization, showed more positive links with language, while mean diffusivity showed more negative ones, suggesting that prenatal exposure may reshape how white matter supports linguistic processing, possibly through premature maturation or compensatory reorganization.

The significance of this finding lies in the dissociation between brain and behavior. Neural differences were detectable even though no behavioral language deficits were present. That means neuroimaging may be able to flag subtle alterations in developmental trajectories before they surface as measurable delays, opening a window for intervention when the brain is at its most plastic. For clinicians, this is a powerful argument for using brain-based measures to identify children who could benefit from early support even when conventional assessments look reassuring.

Experience, too, leaves its mark on the young brain. Cook and colleagues examined how early bilingual exposure shapes prefrontal organization during cognitive control. Using functional near-infrared spectroscopy, which measures changes in blood oxygenation as a proxy for neural activity, they compared 13 bilingual and 13 monolingual preschoolers performing an interference suppression task. Behaviorally, the two groups were indistinguishable in both accuracy and reaction time. But the pattern of brain activity diverged. Bilingual children showed more localized prefrontal connectivity with reduced activation in key decision-making regions, a pattern the authors interpret as greater neural efficiency. The constant practice of switching between two languages, the findings suggest, may refine executive control circuits even before any behavioral advantage becomes apparent, demonstrating that language experience itself acts as a sculptor of functional brain networks.

While structural and functional imaging reveal how biology and experience shape the brain’s hardware, language learning ultimately unfolds in social contexts. Eulau and Hirsh-Pasek synthesized decades of behavioral research with emerging neuroimaging to show how caregiver-child interaction builds linguistic and cognitive development. Their review traces an arc from classic work on joint attention to modern hyperscanning studies, in which the brain activity of two interacting people is recorded simultaneously. The timing of caregiver responses turns out to be critical. Infants show enhanced brain responses to verbal input delivered within 300 to 400 milliseconds, and caregiver responsiveness within roughly two seconds supports symbol-referent mapping. When a toddler gazes at or picks up an object, a caregiver who labels that object within this narrow window allows the child to bind the word to the object while it still dominates their multisensory attention.

The review’s boldest claim is that social contingency shapes development well beyond language. Dual-recording studies show that caregiver-child neural synchrony during interaction predicts not only language outcomes but also attention, emotion regulation, and problem-solving. Infants whose caregivers track their attentional focus more precisely show enhanced sustained attention, and preschoolers who solve problems cooperatively with caregivers display increased prefrontal synchrony linked to better performance. Synchrony measures add value that behavior alone cannot capture: they operate at sub-second timescales, reveal which specific behaviors activate or dampen neural responses, provide genuinely dyadic rather than individual-level data, and illuminate interdependencies between constructs like attention and language. The picture that emerges is of a socially gated brain, one whose developmental trajectory is calibrated by contingent, precisely timed exchanges with caregivers.

Can neural measures also predict which children are at risk for language difficulties? Chen and colleagues addressed this question with magnetoencephalography, a technique that records magnetic fields generated by brain activity with millisecond resolution. They studied 30 three- and four-year-olds with an elevated likelihood of intellectual disability or developmental delay, using a passive auditory task that required no active participation. Evaluable data were obtained from 23 children, a 77 percent success rate in a notoriously difficult population, thanks to the researchers’ MEG-PLAN protocol. The key measure was the M50 response, an early cortical evoked response reflecting the speed of auditory encoding in the brain’s auditory cortex.

The results were striking. Longer M50 latencies in the left hemisphere, indicating slower auditory processing, predicted lower language scores. Crucially, the marker predicted language specifically, not general cognitive ability, pointing to domain-specific relevance. The left auditory cortex matures more slowly than the right throughout development, and this protracted maturation may be especially sensitive to factors that affect language. The findings extend earlier work in older children with autism and minimal verbal abilities, showing that auditory cortex response latencies can forecast language outcomes even in very young children at developmental risk. Because the task is passive, it can be deployed across a wide range of ability levels, making it a realistic candidate for clinical translation.

Taken together, the studies converge on three themes that define the current landscape of developmental language neuroscience. First, neural measures reveal risk and mechanism before behavioral deficits emerge: altered brain-language relationships in exposed toddlers, reorganized prefrontal networks in bilingual preschoolers, and predictive auditory latencies in at-risk children all appeared alongside normal or equivalent behavioral performance. Second, multiple imaging modalities provide complementary insights. Diffusion MRI maps the structural wiring that constrains or enables function, MEG captures the millisecond timing of auditory processing, task-based fNIRS tracks how brain regions coordinate during live cognitive challenges, and dual-recording fNIRS and EEG measure the interactive, dyadic processes that single-brain recordings cannot. Researchers now strategically match the technique to the question rather than relying on any single tool. Third, social and biological factors interact rather than compete. Prenatal exposures alter the structural substrate, bilingual experience reshapes functional networks, contingent caregiving sculpts neural synchrony, and individual differences in neural processing predict outcomes even against a backdrop of genetic and biological risk. Effective intervention, the authors argue, must address all of these as interacting components of one developmental system.

Challenges remain. Sample sizes in developmental neuroimaging are small, as the 26-participant bilingual study and the 23 evaluable MEG participants illustrate, and multi-site collaborations with data sharing will be needed to build adequately powered datasets. Accessibility and representativeness are also concerns: neuroimaging research remains concentrated in well-resourced settings, and most samples do not reflect global diversity, although the South African toddler study marks valuable progress. Bilingualism research, meanwhile, must grapple with the heterogeneity of bilingual populations, including language pairs, age of acquisition, proficiency, and cultural context. Translating neural markers like M50 latency into clinical tools will require prospective longitudinal studies to confirm whether they predict later outcomes and whether interventions can modify both the neural measure and the behavioral trajectory. Finally, the field must move toward integrated, multi-measure approaches with standardized protocols, shared analysis pipelines, and rich behavioral and environmental data. If that happens, the promise of this research is profound: detecting the earliest neural signatures of language risk and harnessing the power of contingent social interaction to keep every young mind’s wiring on course.

Subject of Research: Neural correlates of language development in early childhood

Article Title: Editorial: Wiring the young mind: neural correlates of language development in early childhood

Article References: Mathur, A., & Wang, Y. (2026). Editorial: Wiring the young mind: neural correlates of language development in early childhood. Frontiers in Integrative Neuroscience, 20, Article 1941106. https://doi.org/10.3389/fnint.2026.1941106

Image Credits: AI Generated

DOI: 10.3389/fnint.2026.1941106

Keywords: language development, developmental neuroscience, diffusion tensor imaging, magnetoencephalography, fNIRS, bilingualism, prenatal substance exposure, neural synchrony, caregiver interaction, white matter, auditory processing, early intervention

The study’s scale is one of its strengths. The team analyzed resting-state fMRI scans from 415 cognitively normal individuals, 283 people with mild cognitive impairment (MCI), and 91 people with Alzheimer’s disease, all drawn from ADNI’s standardized 3 Tesla imaging protocols. Rather than treating white matter as noise to be discarded — the traditional approach — the researchers applied a template-guided independent component analysis (ICA) framework that simultaneously modeled 202 intrinsic connectivity networks: 97 white matter networks and 105 gray matter networks. Each participant’s scan was decomposed into these networks, yielding a timecourse for each one, and the correlations between all pairs of timecourses produced a 202-by-202 functional network connectivity (FNC) matrix per person.

The technical foundation matters here. Independent component analysis is a data-driven method that separates mixed fMRI signals into statistically independent spatial patterns, each representing a coherent network of brain activity. The guided version used in this study, called MOO-ICAR, imposes soft spatial constraints drawn from validated templates while still adapting to each individual’s anatomy. The white matter template itself was derived from more than 100,000 fMRI scans and validated across independent datasets, with each of the 97 white matter networks selected for spatial specificity, reproducibility across model orders (spatial correlation greater than 0.8), and minimal overlap with gray matter and cerebrospinal fluid. This addresses the long-standing concern that white matter fMRI signals are merely artifacts of vascular effects or partial-volume contamination from neighboring gray matter.

Why has white matter been so neglected in functional imaging? The answer lies in signal physics. The blood-oxygenation level-dependent (BOLD) signal that fMRI measures is weaker in white matter because of its lower blood supply and different vascular architecture, and early analyses worried that white matter fluctuations reflected head motion or physiological noise rather than genuine neural activity. But a growing body of evidence — including work by Peer and colleagues in 2017 and a review by Gore and colleagues in 2019 — has demonstrated that white matter exhibits structured, low-frequency BOLD fluctuations that are temporally coherent, spatially organized, and reproducible across sites and scanners. The new study builds directly on that foundation, asking whether these fluctuations carry information about disease.

The answer, emphatically, is yes. When the researchers compared connectivity patterns across diagnostic groups, they found widespread alterations in white matter-to-white matter and white matter-to-gray matter connectivity in Alzheimer’s disease, particularly involving subcortical, frontal, sensorimotor, and occipitotemporal white matter networks. In the comparison between Alzheimer’s patients and cognitively normal controls, hypoconnectivity emerged within the occipitotemporal white matter subdomain and its connections to other white matter networks — a finding that dovetails with diffusion MRI studies showing microstructural degradation, including myelin loss and axonal damage, in posterior white matter tracts in Alzheimer’s. At the same time, hyperconnectivity appeared between subcortical and sensorimotor white matter domains, a pattern the authors suggest may reflect altered subcortical-sensorimotor communication rather than simple degeneration.

Perhaps the most striking findings involve cross-tissue interactions — the functional dialogue between gray matter networks and the white matter pathways that link them. In Alzheimer’s disease relative to controls, connectivity between cerebellar gray matter networks and frontal white matter networks was reduced, while coupling between hippocampal gray matter networks and frontal white matter networks was increased. This reweighting of inputs to frontal executive pathways — diminished cerebellar and subcortical influence alongside heightened hippocampal engagement — suggests that Alzheimer’s-related network disruption is not confined to gray matter but extends prominently to the interfaces where gray and white matter systems communicate. The pattern generalizes earlier gray matter-only reports of disrupted fronto-hippocampal coupling by showing that part of that disruption is expressed at the tissue boundary itself.

Critically, these disruptions were not limited to full-blown dementia. In the comparison between MCI and cognitively normal participants, white matter-to-gray matter dysconnectivity was already present, most prominently between cerebellar and subcortical gray matter regions and frontal white matter subdomains. This means the functional disruption of white matter circuits emerges before overt dementia — at the prodromal stage when interventions might still alter the disease trajectory. Within the gray matter domain, the MCI group also showed reduced connectivity within the cerebellum itself, accompanied by increased coupling between cerebellar and visual networks, an opposing pattern that may reflect early reorganization as pathology takes hold.

The progression from MCI to Alzheimer’s disease revealed further reorganization. Connectivity between temporal and semantic gray matter regions — such as the temporal pole and superior temporal sulcus — and frontal and insular white matter networks increased, while connections between sensorimotor and hippocampal gray matter regions, and between visual gray matter and sensorimotor white matter, weakened. White matter-to-white matter integration also declined, particularly between occipitotemporal and sensorimotor domains, pointing to ongoing disconnection of posterior and sensorimotor pathways even between adjacent clinical stages. Taken together, the sequence suggests that cerebellar and subcortical-to-frontal white matter interactions are altered earliest, with posterior occipitotemporal and sensorimotor disruption following as the disease advances.

Statistically, the team was careful to control for confounds. Age, sex, and mean framewise displacement — a measure of head motion during scanning — were regressed out of each participant’s connectivity values before group comparisons. Multiple comparisons across the thousands of possible network pairs were controlled using false discovery rate correction at q < 0.05, and effect sizes were quantified with Hedges’ g, a bias-corrected standardized measure. The authors also note limitations: the analysis captured static connectivity rather than time-varying dynamics, the cross-sectional design cannot establish temporal causality, and multi-site acquisition introduces scanner heterogeneity that covariate adjustment may not fully eliminate. Future work integrating longitudinal data, amyloid and tau biomarkers, and diffusion MRI will be needed to determine whether the observed functional changes are secondary to structural degeneration or represent independent — possibly compensatory — mechanisms.

The broader implication is a conceptual shift. If white matter functional connectivity is both measurable and selectively altered across the Alzheimer’s continuum, it offers a systems-level signature that complements established gray matter-based measures. Structural imaging can show that white matter tracts are degrading, but it cannot quantify how communication between distributed brain systems is functionally rewired. Hyperconnectivity patterns, in particular, cannot be explained by structural loss alone and may reflect compensatory reorganization or maladaptive reconfiguration. As the authors conclude, white matter functional network connectivity is a promising complement to gray matter measures for understanding disease progression — and, potentially, a tool for earlier diagnosis, patient stratification, and monitoring of one of medicine’s most stubborn diseases. The brain’s wiring, it turns out, has been talking all along; scientists are only now learning to listen.

Subject of Research: White matter functional network connectivity alterations across the Alzheimer's disease spectrum measured with resting-state fMRI

Article Title: Investigating white matter functional network connectivity across the Alzheimer’s disease spectrum using resting-state fMRI

Article References: Itkyal, V. S., LaGrow, T. J., Jensen, K. M., Iraji, A., & Calhoun, V. D. (2026). Investigating white matter functional network connectivity across the Alzheimer’s disease spectrum using resting-state fMRI. Frontiers in Neuroimaging, 5, Article 1796824. https://doi.org/10.3389/fnimg.2026.1796824

Image Credits: AI Generated

DOI: 10.3389/fnimg.2026.1796824

Keywords: Alzheimer's disease, white matter, functional connectivity, resting-state fMRI, mild cognitive impairment, independent component analysis, ADNI, gray matter, biomarker, neurodegeneration, brain networks, dysconnectivity

Late preterm birth occupies an unusual position in developmental medicine. It is common, accounting for a substantial share of all preterm deliveries, yet it falls into a gray zone: too early to be fully term, too late to attract the intensive monitoring given to very premature infants. What makes the window biologically consequential is timing. Delivery at 34 to 36 weeks coincides with a period of extraordinarily rapid brain construction, when white matter expands substantially and the cerebral cortex is still busy with synapse formation, dendritic growth, cortical folding, and expansion of the brain’s surface. Previous studies of children born in this window have produced mixed findings, with some reporting developmental vulnerabilities and others finding few lasting effects, leaving clinicians without a clear picture of what, if anything, late preterm birth does to the long-term trajectory of brain maturation.

To resolve that uncertainty at scale, the Fukui team, which also included doctoral student Ryoko Yoshinare, Assistant Professor Masatoshi Yamashita, and Professor Yoshifumi Mizuno, turned to the ABCD Study, the largest long-term study of brain development and child health in the United States. The researchers compared 1,093 children born late preterm with 8,984 children born at full term, all between the ages of 9 and 11 years at the study’s baseline. Rather than relying on a single snapshot, they exploited the longitudinal design of the ABCD Study, examining the same children across two timepoints separated by roughly two years. Their battery of measures was unusually comprehensive: standardized assessments of behavioral and emotional problems, tests of cognitive function, structural magnetic resonance imaging (MRI) to map regional brain volumes and cortical dimensions, diffusion tensor imaging (DTI) to probe the microstructural integrity of white matter, and resting-state functional MRI to chart how brain networks communicate.

The first surprise was what the team did not find. Across the behavioral and emotional measures, most cognitive outcomes, cortical thickness, cortical surface area, and resting-state functional connectivity, the late preterm and full-term groups were statistically indistinguishable. There were no broad differences that survived correction for multiple comparisons, and no significant timepoint or group-by-timepoint effects in these domains. For parents and clinicians, that absence is itself meaningful: on the measures that most directly touch daily life, children born a few weeks early did not differ from their full-term classmates in any sweeping way.

Structural MRI told a more nuanced story. The analysis identified significant group differences in the volumes of six specific regions: the left postcentral gyrus, which processes bodily sensation; the right fusiform gyrus, involved in recognizing faces and objects; the right middle temporal region, important for motion and language processing; the right pars orbitalis, part of the language-producing inferior frontal cortex; the right supramarginal gyrus, which supports language and spatial attention; and the right insula, a hub for interoception and emotional awareness. Notably, however, these volumetric differences showed no significant change over time and no group-by-timepoint interaction after false discovery rate correction, suggesting they were stable features of the baseline scan rather than diverging developmental trajectories.

The clearest and most dynamic differences emerged from DTI, a technique that tracks the diffusion of water molecules through brain tissue to reveal the organization of its microscopic architecture. Two derived measures carry most of the signal: fractional anisotropy (FA), which is higher when water diffuses preferentially along organized axonal pathways, and mean diffusivity (MD), which reflects the overall freedom of water movement and is influenced by tissue properties such as myelination, axonal density, and cellularity. The researchers found group effects in selected white matter tracts, cortical regions, and subcortical structures for both FA and MD. Among the tracts, the inferior longitudinal fasciculus, a major pathway connecting occipital visual areas with temporal regions involved in memory and language, showed group effects for both FA and MD, although in different hemispheres. More strikingly, the longitudinal analysis detected group-by-timepoint interactions across several white matter tracts and selected cortical and subcortical regions, indicating that the two groups were not simply different at one moment but were maturing along measurably different microstructural courses.

The direction of those interactions was telling. The analyses suggested that many of them reflected decreases in DTI measures among full-term children over the two-year interval, while the late preterm group showed changes that were smaller, statistically nonsignificant, or less consistent. In other words, some aspects of microstructural development that typically shift during early adolescence appeared to unfold differently in children born late preterm, not as a dramatic deviation but as a dampening or reordering of the usual developmental rhythm. Because FA and MD are influenced by multiple tissue properties simultaneously, the authors are careful to note that the results do not establish a specific mechanism such as altered myelination or axonal organization; they identify where and when development differs, not precisely how.

Professor Tomoda framed the findings with deliberate caution. “These findings suggest that the effects of LP birth may be subtle and may not necessarily appear as broad behavioral or cognitive difficulties,” she said. “Instead, brain imaging may reveal regional differences in maturation that are not immediately evident in everyday functioning.” That framing captures the study’s central tension: the imaging data reveal real, statistically robust differences in how the brain matures, yet those differences do not translate into observable deficits on the behavioral and cognitive instruments used here. Neuroimaging, in this sense, functions as a sensitive early instrument, capable of detecting developmental variation before it would surface in clinical or educational measures, if it ever does.

The reassuring headline for families is that the absence of widespread behavioral or cognitive differences suggests late preterm birth, on its own, is not a sentence of developmental difficulty. But the DTI results argue against complacency in the opposite direction. “LP birth should not be viewed as a prediction of poor developmental outcomes,” Professor Tomoda added. “Rather, our results support continued awareness of individual developmental needs, so that children who experience emerging difficulties can receive appropriate assessment and support.” The imaging differences do not establish that every late preterm child will develop difficulties, nor do they demonstrate that the observed brain differences cause functional problems. What they provide is a biological rationale for attentive, individualized follow-up rather than either alarm or blanket reassurance.

The researchers are equally candid about the limits of what the study can claim. Substantial attrition between the two assessment timepoints, the availability of only two measurement occasions, retrospective reporting of gestational age by parents, and the observational design of the ABCD Study all constrain interpretation; observational data cannot rule out the influence of factors that correlate with both gestational length and brain development. What the study establishes is that late preterm birth is associated with selected regional brain-volume differences and, particularly in longitudinal analysis, with differences in DTI measures during early adolescence. Whether those subtle neuroimaging signatures persist, normalize, or eventually relate to cognitive, behavioral, or educational outcomes is a question only further longitudinal research can answer. For now, the study offers a refined picture of the late preterm brain: largely typical in its outward function, quietly distinctive in its microscopic maturation, and best understood not as a category of risk but as a reminder that development unfolds on individual timetables that deserve to be watched with care.

Subject of Research: Neurodevelopmental and brain imaging outcomes in children born late preterm

Article Title: Study reveals subtle neuroimaging differences after late preterm birth

Article References: Study reveals subtle neuroimaging differences after late preterm birth. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: late preterm birth, ABCD Study, neuroimaging, diffusion tensor imaging, brain development, white matter, fractional anisotropy, mean diffusivity, pediatric neuroscience, Frontiers in Pediatrics, University of Fukui, child development

The research team, led by Suzanne M. de la Monte of the Departments of Pathology and Laboratory Medicine, Neurology, Neurosurgery, and Medicine at Rhode Island Hospital, together with Ming Tong, obtained six-millimeter cores of the anterior cerebellar vermis from the New South Wales Brain Tissue Research Centre in Sydney, Australia. The samples came from six male donors with alcohol use disorder and six matched male controls, all between fifty and seventy years of age. The groups were closely matched for age, years of drinking, postmortem interval, and brain pH, but differed dramatically in lifetime alcohol exposure: the AUD donors had consumed an average of roughly 2,784 kilograms of alcohol over their lifetimes, compared with about 42 kilograms in the controls. Consistent with the toxic burden, the AUD brains weighed significantly less, averaging 1,335 grams against 1,505 grams in controls.

What makes this study technically distinctive is its methodological breadth. Rather than relying on a single assay, the investigators processed the fresh frozen tissue through parallel pipelines: duplex enzyme-linked immunosorbent assays to quantify oligodendrocyte and astrocyte proteins, magnetic bead-based multiplex panels to measure twenty-one cytokines, chemokines, and trophic factors, seven-plex assays for total and phosphorylated Akt pathway signaling proteins, and custom QuantiGene 2.0 multiplex RNA hybridization panels to quantify messenger RNA transcripts for glial genes, the insulin and insulin-like growth factor system, and the Notch pathway. Every assay was run in triplicate under code, normalized to internal reference standards, and analyzed with two-way analysis of variance and post hoc multiple comparison tests.

The oligodendrocyte findings were striking. Protein analysis revealed significantly elevated levels of CNPase, an enzyme associated with immature myelin, and myelin oligodendrocyte glycoprotein, or MOG, a marker of mature oligodendrocytes, in the AUD cerebellar white matter. The RNA analysis was even more revealing, detecting broad upregulation of immature and mature myelin genes alike: KLK6, a serine protease linked to myelin degradation, along with CNPase, myelin basic protein, proteolipid protein, and MOG were all significantly increased. At first glance, elevated myelin gene expression might seem like good news, but the authors interpret it as a sign of dysfunction rather than health. In experimental models of alcohol-related brain damage, similar upregulation appears in subacute stages of injury, before the later collapse of myelin gene expression. The pattern suggests oligodendrocytes struggling to compensate for damaged myelin, much like a factory running overtime to replace a product that keeps failing.

The elevation of KLK6 is particularly telling. Previous work has shown that inhibiting this protease in oligodendrocytes is associated with cellular maturation and increased myelin thickness and volume. Its upregulation in the AUD samples, the authors argue, corresponds with a failure of oligodendrocytes to fully mature, which would compromise the synthesis and maintenance of the myelin sheaths that give white matter its integrity. This interpretation dovetails with earlier lipidomic mass spectrometry studies showing altered myelin lipid composition in alcohol-related brain damage, and with diffusion tensor imaging studies demonstrating disruption of white matter microstructural integrity in living patients with alcohol use disorder.

Beneath these cellular changes, the study uncovered disruptions in two master signaling systems. The first is the insulin and insulin-like growth factor pathway, which supports cell growth, survival, and energy metabolism throughout the brain. While mRNA levels of insulin pathway genes showed only modest, statistically insignificant shifts, the protein-level analysis told a sharper story: total Akt, a central kinase in the pathway, was significantly reduced, as was GSK-3β and its serine-9 phosphorylated form. These reductions suggest impaired signaling through the PI3K-Akt axis, which experimental models of chronic ethanol exposure have repeatedly implicated in neurodegeneration. The second system is the Notch pathway, a developmental signaling network that also regulates oligodendrocyte myelin synthesis and maintenance in the adult brain. Here the RNA assays revealed significant reductions in ASPH, aspartyl-asparaginyl-beta-hydroxylase, and HES1, a downstream Notch transcriptional target. ASPH is the enzyme that hydroxylates Notch receptors’ EGF-like domains, enabling the cleavage and nuclear translocation that activates the pathway. Its suppression, the authors propose, would throttle Notch signaling and thereby cripple the oligodendrocyte functions needed to sustain central nervous system myelin.

Perhaps the most surprising result concerns neuroinflammation, long considered a primary driver of alcohol-related brain damage. The comprehensive cytokine and chemokine survey found that among twenty-one inflammatory mediators measured, only one, the pro-inflammatory cytokine IL-16, differed significantly between groups, and it was reduced, not elevated, in the AUD tissue. IL-16, a chemoattractant for CD4-positive T cells that has been implicated in inflammatory demyelinating diseases, was also the most abundantly expressed inflammatory factor in the cerebellar samples. The authors note that these findings contrast with studies linking increased neuroinflammation to short-term, binge, or developmental alcohol exposures, but align with other experimental data showing that chronic ethanol consumption can suppress immune and cytokine-mediated responses. On balance, the results do not support persistent chronic neuroinflammation as the prime mediator of white matter degeneration in this cohort.

One trophic factor did stand out: hepatocyte growth factor, or HGF, was significantly reduced in the AUD cerebellum. HGF plays a key role in preventing neuronal death and promoting cell survival, and its loss corresponds with the neuronal loss observed in the cerebellar cortex in alcohol-related brain damage. Given HGF’s pro-angiogenic and anti-inflammatory effects, its reduction may also bear on the blood-brain barrier disruption documented in this condition. The other growth factors measured, including nerve growth factor, vascular endothelial growth factor, stem cell factor, fibroblast growth factor, and stromal cell-derived factor, showed no significant AUD-related changes.

The study has clear limitations that the authors acknowledge candidly. Six brains per group is a modest sample, constrained by the scarcity of donors with uncomplicated alcohol-related brain damage free of other substance use disorders. All donors were male, reflecting the male predominance of heavy drinking in Australia, and the analysis captured only the endpoint of disease rather than its progression. The authors call for future work examining earlier timepoints and including donors with periods of abstinence, which would help assess whether the cerebellar white matter pathology is reversible. Encouragingly, prior neuroimaging studies have shown that cerebellar atrophy can be partially reversed with abstinence, making the identification of actionable disease stages clinically meaningful.

The translational horizon is perhaps the most exciting aspect of the work. Because oligodendrocytes and other brain cells release cell-type-specific membrane vesicles called exosomes into the bloodstream, the authors propose that the molecular signatures identified here could eventually be detected and monitored non-invasively through liquid biopsy approaches, analyzing exosomes isolated from peripheral blood. Such a test could allow clinicians to track alcohol-related oligodendrocyte pathology in living patients, stage the disease, and evaluate whether interventions, from abstinence to potential therapies targeting insulin/IGF-Akt signaling or Notch networks, are working. For a disorder that is among the most common and preventable causes of cerebellar degeneration worldwide, moving diagnosis from the autopsy table to the blood draw would represent a genuine leap forward in the fight against alcohol-related brain damage.

Subject of Research: Molecular mechanisms of alcohol-related cerebellar white matter degeneration in humans

Article Title: Molecular and biochemical pathologies in human alcohol-related cerebellar white matter degeneration

Article References: de la Monte, S. M., & Tong, M. (2025). Molecular and biochemical pathologies in human alcohol-related cerebellar white matter degeneration. Advances in Drug and Alcohol Research, 5, Article 15342. https://doi.org/10.3389/adar.2025.15342

Image Credits: AI Generated

DOI: 10.3389/adar.2025.15342

Keywords: alcohol use disorder, cerebellum, white matter, oligodendrocytes, myelin, insulin/IGF signaling, Akt, GSK-3β, Notch pathway, ASPH, neuroinflammation, postmortem brain tissue

Myelin is built by oligodendrocytes, specialized glial cells that each form multiple sheaths around several axons. During development, oligodendrocyte precursor cells migrate through the central nervous system and mature, generating a dynamic period of sheath formation and loss followed by pruning. Neural activity shapes this process: increased firing enhances precursor proliferation and maturation, boosting myelination. For a long time, mature myelin was considered essentially stable, changing only when new oligodendrocytes appeared or in disease. More recent experimental evidence has overturned that view, showing that myelination remains dynamic throughout adulthood. Studies in the rodent cortex reveal that learning, experience, and behavior alter sheath length, number, thickness, and organization, and there is even growing evidence for activity-dependent myelin retraction. Neurons and oligodendrocytes form a bidirectional feedback loop that shapes the myelination program of the brain in real time.

The functional significance of this loop is profound. By remodeling myelin, oligodendrocytes tune axonal conduction velocity and thereby the delays with which action potentials arrive at their postsynaptic targets. This introduces a timing-based plasticity mechanism that operates in parallel to changes in synaptic strength, supporting learning, memory, and sensory and motor function. In the healthy brain, ADM optimizes and coordinates neural communication. But abnormal activity can hijack it. In certain forms of epilepsy, seizure activity promotes maladaptive hypermyelination that reinforces seizure patterns and drives disease progression, a feedback loop of dysfunction. In demyelinating disorders such as multiple sclerosis, compromised myelination underlies the sensory, motor, and cognitive deficits patients experience. Despite this significance, ADM lacks the strong theoretical framework that synaptic plasticity enjoys, largely because it involves multiple mechanisms interacting across widely different time scales: action potentials propagate in milliseconds, whereas myelin changes unfold over hours to days, and the relevant processes span subcellular signaling, single axons, gray matter microcircuits, and large-scale white matter networks.

Nguyen and Lefebvre organize the emerging computational literature into three classes of models, each operating at a distinct spatial scale. The first class focuses on individual axons, where neuron-glia interactions occur, and examines how changes in myelin structure map onto changes in conduction timing. Because structural remodeling in vivo often alters multiple myelin properties simultaneously, computational models can isolate the impact of individual variables and determine where causal relationships exist. These models are often biophysically detailed, explicitly representing action potential dynamics along myelinated axons composed of sheath and node compartments, frequently relying on Hodgkin-Huxley-like kinetics to simulate realistic ionic currents.

A landmark example comes from combined experimental-computational work by Ford and colleagues, who fit their model to geometrical measurements and electrophysiological recordings from gerbil auditory fibers. Building on the Halter and Clark multi-compartment model of the myelinated axon, which independently models intra-axonal, peri-axonal, and extra-axonal potentials, they established how the ratio between internode length and axon diameter influences conduction velocity. The smaller ratio observed in low-frequency auditory fibers alone accounts for a 15 percent increase in conduction velocity, and both axon types exhibit an optimal ratio that maximizes speed: internodes that are too short lack overall myelination, while depolarizations are not efficiently transferred between nodes when internodes are too long. Follow-up modeling of node length, fit to rat optic nerve and cortex measurements, revealed that with a constant number of ion channels per node, increasing node length almost linearly decreases conduction velocity due to increased capacitance and axial resistance, yet with constant channel density there is an optimal node length. Micrometer-scale changes in node length can increase conduction velocity by up to 20 percent in the optic nerve, and nodal remodeling presents a more metabolically efficient strategy for tuning conduction than altering internode architecture.

Axonal models have also examined how spatial heterogeneity in myelin coverage impacts signaling, injury susceptibility, and recovery. Myelin sheaths vary in length and are often non-uniformly arranged along axons, a patchy organization that is difficult to precisely manipulate in vivo but easy to control in simulation. Work by Scurfield and Latimer showed that, assuming the same number of remyelinated segments, transitions between short and long internodes slow conduction more than if all shorter internodes occupy one half of the axon and all longer ones the other. More recent modeling by Talidou and Lefebvre, using mouse imaging data to generate physiologically realistic sheath lengths and arrangements, demonstrated that axons with uniform myelin show negligible variability in conduction delays, whereas heterogeneous motifs show high, nonlinear, path-dependent variability, and that spatial heterogeneity increases the likelihood of conduction failure, making some motifs more vulnerable to demyelination. The major limitation of this class of models is that they do not explicitly model ADM; they treat myelin microstructure as a static parameter and model different myelination states as a proxy for its outcome. Still, they provide evidence that specific structural configurations can optimize spike timing, which is a primary goal of ADM.

The second class operates at the microcircuit level, where conduction velocity serves as a proxy for myelination and the feedback between neural activity and myelin adaptation is modeled directly. Talidou and colleagues built a recurrent network of excitatory spiking neurons in which conduction delays are state-dependent, using a Hebbian, use-dependent rule in which conduction velocity scales with firing rate along a given axon. Simulations showed that ADM acts as a homeostatic mechanism of gain control, temporally coordinating action potentials and up-regulating correlated activity, and that it can bidirectionally and reversibly alter conduction velocities in response to external stimuli, suggesting myelin plasticity responds dynamically to different brain states such as learning versus rest. Because synaptic connectivity was fixed, the results highlight how timing plasticity can complement synaptic plasticity in maintaining optimal brain function.

Pajevic and colleagues took a step further by asking how oligodendrocytes might integrate local spike timing as feedback to tune conduction velocities. In their framework, each oligodendrocyte releases a transient global intracellular signaling factor in response to action potentials propagating along any of the multiple axons it myelinates, while each axon carries a unique local myelination factor whose release is proportional to the global signal. This combination of global and local signals allows selective, bidirectional modulation of myelination along each axon. Model simulations reveal that oligodendrocytes can actively tune conduction velocities to selectively synchronize correlated spike trains while leaving uncorrelated trains unaffected, with synchronization greatest when firing rates are 10 Hz or less and transient responses span 10 to 40 milliseconds. This proposes a biologically plausible mechanism in which oligodendrocytes act as coincidence detectors of neural activity. The limitations are clear, however: existing microcircuit models rely on phenomenological learning rules, bypass explicit myelin structure, consider only excitatory networks, and compress the timescales of myelination for computational tractability.

The third class addresses the whole brain. Most inter-regional traffic flows through white matter, a heavily myelinated fiber network occupying roughly half of the mammalian brain volume, and large-scale models adopt phenomenological neural mass formulations that link directly to neuroimaging data such as fMRI, EEG, and MEG. Pajevic and colleagues emphasized that myelin-dependent changes in conduction velocity alter time delays, which can strongly influence brain oscillations and synchrony, drawing on the phenomenon of amplitude death, whereby dispersed delays suppress oscillatory activity. Using time-delayed Kuramoto oscillator networks, they showed that millisecond-scale changes in delays can impede the ability of neural assemblies to synchronize, and that delays directly alter oscillation frequency. Noori and colleagues then implemented a phenomenological learning rule in which conduction velocity along each axon changes according to the oscillatory phase difference between nodes, informed by macaque white matter structural connectivity. Their simulations revealed that ADM enables an initially incoherent network to self-regulate conduction velocities and achieve a globally phase-locked state, compensating for differences in tract lengths to maintain similar delays, and that plastic velocities enable synchronization at lower synaptic strengths and across wider frequency bands, even restoring synchrony after perturbations or injury. Park and Lefebvre confirmed mathematically that phase-dependent delays increase the stability of globally synchronous states and promote resilience to the loss of axonal connections.

Toward a unified framework, Nguyen and Lefebvre argue that the field remains fragmented and that future progress depends on interdisciplinary iteration between models and experiments. The physiological mechanisms underlying ADM remain unclear, and current models change myelination based on firing rate, individual spike timings, or synchrony, while ignoring oligodendrocyte heterogeneity and the contributions of astrocytes, precursor cells, and microglia. A particularly promising mechanistic candidate is calcium-dependent signaling: recent in vivo imaging shows that increases in intracellular calcium transients within developing oligodendrocytes trigger myelin sheath elongation, while sheath shortening is associated with lower transient frequency, suggesting a calcium-mediated feedback loop that could temporally bridge fast neural activity and slower myelin changes. Multiscale modeling offers a path across spatial scales, as illustrated by recent work linking a biophysically detailed axonal model of rhesus monkey prefrontal neurons to a spiking network showing that demyelination-induced propagation failures impair working memory but can be restored with remyelination. The authors conclude that future models should include dynamic conduction delays, analogous to the widespread adoption of variable synaptic weights, and that, like synaptic plasticity models, ADM models may clarify this form of plasticity in healthy and pathological brain states while inspiring new applications in engineering, machine learning, and artificial intelligence.

Subject of Research: Neuroscience

Article Title: A computational roadmap to activity-dependent myelination

Article References: Nguyen, T., & Lefebvre, J. (2026). A computational roadmap to activity-dependent myelination. Frontiers in Computational Neuroscience, 20, Article 1895171. https://doi.org/10.3389/fncom.2026.1895171

Image Credits: AI Generated

DOI: 10.3389/fncom.2026.1895171

Keywords: activity-dependent myelination, myelin plasticity, oligodendrocytes, conduction velocity, neural synchrony, computational neuroscience, white matter, brain plasticity, computational, roadmap, activity-dependent, myelination

The research team enrolled 82 participants in a cross-sectional design: 44 people carrying carotid vulnerable plaques, 15 people with carotid stable plaques, and 23 healthy controls. Vulnerable plaques are those with features such as lipid-rich cores, thin or ruptured fibrous caps, or intraplaque hemorrhage, characteristics that mark them as more likely to rupture and trigger embolic events. Stable plaques, by contrast, tend to be fibrotic and calcified, and are considered less likely to shed debris into the bloodstream. All participants underwent brain magnetic resonance imaging with a multi-shell diffusion protocol, a technique that sensitizes the MRI signal to the random motion of water molecules within tissue. Because water diffuses more freely along the length of nerve fibers than across them, the pattern of diffusion reveals the integrity of the brain’s underlying cabling.

The study employed two complementary diffusion modeling approaches. Diffusion tensor imaging, or DTI, the older and more widely used method, characterizes diffusion with a simple ellipsoid model and yields metrics such as fractional anisotropy, mean diffusivity, axial diffusivity, and radial diffusivity. Diffusion kurtosis imaging, or DKI, extends this framework by measuring the degree to which water diffusion deviates from the ideal Gaussian behavior assumed by DTI. This non-Gaussian component, quantified as kurtosis, is thought to be sensitive to tissue complexity at the cellular level, including restrictions imposed by cell membranes, myelin sheaths, and organelles. By combining both modalities in a single multi-shell acquisition, the researchers aimed to strengthen confidence that the abnormalities they detected reflected genuine microstructural injury rather than artifacts of a single model.

To analyze the imaging data, the team applied tract-based spatial statistics, a widely used whole-brain approach that aligns all participants’ images to a common template and projects diffusion metrics onto the centerlines of major white matter tracts. This voxel-wise method allows researchers to detect subtle, spatially extended changes that might escape notice in conventional region-of-interest measurements. The investigators complemented the whole-brain analysis with targeted region-of-interest comparisons across a battery of named tracts, including the corona radiata, the corpus callosum, the internal capsule, the superior longitudinal fasciculus, and the cingulum, among others. Blood samples were collected to measure two systemic inflammatory markers: high-sensitivity C-reactive protein, an acute-phase protein produced by the liver, and interleukin-6, a signaling molecule secreted by immune cells that drives much of the acute-phase response.

The results were striking in their asymmetry. Compared with healthy controls, participants with vulnerable plaques showed widespread white matter microstructural damage, concentrated in the posterior corona radiata and the superior longitudinal fasciculus. The damage pattern was internally consistent: mean diffusivity, axial diffusivity, and radial diffusivity were all elevated, indicating increased freedom of water movement and therefore disruption of the barriers that normally constrain it, while mean kurtosis and axial kurtosis were reduced, consistent with a loss of cellular complexity and architectural order. Radial diffusivity, in particular, is often interpreted as a marker of myelin injury, whereas axial diffusivity changes have been linked to axonal damage. The co-occurrence of these signatures suggests that the vulnerable plaque group carried genuine structural injury in the long fiber pathways that connect distant brain regions.

Participants with stable plaques told a very different story. When the researchers compared the stable plaque group with healthy controls, only a handful of regions showed statistically significant differences in diffusion metrics, and the overall burden of abnormality was far lower than in the vulnerable plaque group. This dissociation is the central finding of the study: two groups of people, both carrying carotid plaques that had never caused symptoms, showed profoundly different brain microstructure depending on the biological character of the plaque itself. The finding implies that the vulnerable plaque is not merely a stroke risk waiting to happen but may be an active source of ongoing, subclinical brain injury.

The inflammatory connection sharpened the picture further. In the vulnerable plaque group, the degree of white matter damage correlated significantly with blood levels of both high-sensitivity C-reactive protein and interleukin-6. Participants with higher systemic inflammatory burden tended to show more severe microstructural abnormalities in the affected tracts. No such correlation emerged in the stable plaque group or in healthy controls, suggesting that the link between inflammation and brain injury is specific to the vulnerable plaque state rather than a general feature of atherosclerosis. The researchers propose a model in which biologically active plaques sustain a state of systemic inflammation, and that inflammation, through mechanisms that may include endothelial dysfunction, microvascular injury, and direct inflammatory signaling across the blood-brain barrier, contributes to the degradation of white matter integrity.

The technical strength of the study lies in its dual-modality design. DTI metrics alone can be confounded by factors such as crossing fibers, partial volume effects, and free water contamination, which can mimic or mask true microstructural change. Kurtosis metrics, being sensitive to non-Gaussian diffusion, provide an independent check on the tensor findings. The fact that both approaches converged on the same tracts, with diffusivity measures rising and kurtosis measures falling in the posterior corona radiata and superior longitudinal fasciculus, strengthens the argument that the observed changes reflect real tissue injury. The posterior corona radiata carries fibers connecting the thalamus and brainstem to the parietal and occipital cortices, while the superior longitudinal fasciculus is one of the brain’s major long association pathways, linking frontal, parietal, and temporal regions. Damage to these tracts has been associated with cognitive slowing, executive dysfunction, and processing speed deficits in other cerebrovascular populations.

The study has limitations inherent to its cross-sectional design. With 44 vulnerable plaque participants, 15 stable plaque participants, and 23 controls, the sample is modest, and the imbalance between plaque subgroups may limit statistical power for detecting subtler differences in the stable group. Cross-sectional data cannot establish whether the white matter changes precede, accompany, or follow the development of plaque vulnerability, nor can they confirm that inflammation is a cause rather than a correlate of the injury. Longitudinal follow-up will be needed to determine whether people with vulnerable plaques and elevated inflammatory markers progress to overt cognitive decline or stroke at higher rates, and whether reducing inflammation, through statins, anti-inflammatory agents, or lifestyle interventions, can halt or reverse the microstructural damage.

Nevertheless, the findings carry practical implications. Carotid ultrasound is already a routine part of stroke risk assessment, and plaque morphology can be characterized noninvasively. If the results are replicated in larger and more diverse cohorts, the combination of plaque vulnerability assessment with diffusion kurtosis imaging and inflammatory biomarker measurement could offer clinicians a way to identify, among the many people with silent carotid disease, the subset whose brains are already being affected. That would represent a shift from treating asymptomatic carotid atherosclerosis as a single condition toward recognizing it as a spectrum in which plaque biology, systemic inflammation, and brain integrity are intertwined. For now, the study stands as an early but compelling piece of evidence that the brain keeps score of what the carotid artery is doing, even when the patient feels nothing at all.

Subject of Research: White matter microstructural changes associated with vulnerable versus stable asymptomatic carotid plaques and their relationship to inflammatory biomarkers

Article Title: Plaque vulnerability-specific white matter microstructural integrity loss in asymptomatic carotid atherosclerosis and its link to inflammatory biomarkers: a dual-modality DTI/DKI study

Article References: Wang, D., Leng, Y., Liu, W. V., Zhang, Y., Liao, W., & Yang, S. (2026). Plaque vulnerability-specific white matter microstructural integrity loss in asymptomatic carotid atherosclerosis and its link to inflammatory biomarkers: a dual-modality DTI/DKI study. BMC Medical Imaging. https://doi.org/10.1186/s12880-026-02939-3

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02939-3

Keywords: carotid atherosclerosis, vulnerable plaque, white matter, diffusion tensor imaging, diffusion kurtosis imaging, inflammation, C-reactive protein, interleukin-6, stroke risk, brain imaging, microstructural integrity, cerebrovascular disease

The technique, called Tractfinder-constrained Tractography, or TcT, was developed by researchers at UCL Great Ormond Street Institute of Child Health, the Francis Crick Institute and Great Ormond Street Hospital for Children in London. It builds on Tractfinder, an automated method introduced in recent years that outlines white matter tracts in patients whose brain anatomy has been distorted by tumours. Tractfinder’s strength lies in its speed and automation, but its outputs arrive as voxel-based masks, essentially three-dimensional volumes marking where a tract is likely to sit. That format creates a practical bottleneck, because most established tractography tools, visualisation workflows and downstream analysis methods are built around streamlines, the continuous fibre pathways that model tracts as collections of curved lines through the brain.

The distinction matters more than it might sound. A binary mask tells a surgeon where a tract lies but offers little of the geometric detail that streamline representations provide. Streamlines also enable analyses that voxel outputs cannot support, such as along-tract statistical profiling, which examines how tissue properties change along the length of a fibre bundle. By constraining probabilistic tractography to the probability maps that Tractfinder produces, TcT generates streamline representations while preserving the automation that makes the original method clinically appealing. The Tractfinder-derived segmentations serve as binary mask regions of interest, restricting streamline propagation to within the tract volume and ensuring that the reconstructed streamlines remain consistent with the underlying segmentation.

To test the approach, the team applied TcT to ten pre-operative paediatric patients with supratentorial tumours, a group whose diagnoses ranged from pleomorphic xanthoastrocytoma and craniopharyngioma to ependymoma, pilocytic astrocytoma, epidermoid cyst, glioneuronal tumour and low-grade glioma, with some tumours affecting both hemispheres. Three clinically critical tracts were selected for evaluation: the corticospinal tract, which carries motor signals; the arcuate fasciculus, a key language pathway; and the optic radiation, which relays visual information. These pathways are frequently displaced, infiltrated or disrupted by supratentorial tumours, and surgical injury to any of them carries significant functional consequences.

The imaging data came from routine pre-operative assessments performed on a Siemens MAGNETOM Prisma 3 Tesla scanner. Structural T1-weighted images were acquired at isotropic 1 mm resolution, while diffusion-weighted images used a multi-shell scheme with 60 directions at b-values of 1,000 and 2,200 s/mm², plus 13 unweighted volumes, at 2 mm isotropic resolution. After standard preprocessing steps including thermal noise removal, Gibbs ringing correction, motion and distortion correction, and bias field correction, the diffusion data were linearly registered to MNI152 standard space. The researchers deliberately chose affine rather than non-linear registration, reasoning that a global transformation accounts for individual anatomical variability while limiting the streamline distortions that local registration errors can introduce, and that affine methods offer advantages in speed and reproducibility.

Fibre orientation distributions were estimated using multi-shell multi-tissue constrained spherical deconvolution, with tissue response functions derived through an unsupervised algorithm. Where tumour masks were needed, the team employed an nnU-Net-based segmentation framework validated for routinely acquired, heterogeneous clinical data. Critically, that framework was trained on incomplete imaging datasets, so lesion masks could be derived from T1-weighted images alone, reducing both data requirements and processing time. Tractfinder then generated pseudo-probability maps of tract location, computed as voxel-wise inner products of a registered orientation atlas and the subject’s fibre orientation images. Where tumour mass effect had displaced a tract, a tumour deformation model incorporating the lesion mask was applied. Manual adjustment of the deformation parameter was required in only one of the ten cases, suggesting the pipeline is largely robust to automation, though the authors note that visual quality control remains advisable.

The probability maps were thresholded at 0.05, a value determined empirically in the original Tractfinder methodology, and the resulting binary masks were used to constrain targeted probabilistic streamline tractography using the iFOD2 algorithm. Tractography ran with high-density seeding of one million seeds and no fixed streamline count, with MRtrix3 default settings governing step size, angle threshold and fibre orientation cutoff. For comparison, the same datasets underwent conventional targeted probabilistic tractography seeded from manually placed regions of interest, following a published protocol with specific seed, inclusion and exclusion regions for each tract. Conventional streamlines were then post-processed to remove spurious fibres using a track-density-imaging filter that rejected any streamline not passing through voxels meeting a minimum density threshold.

The results were striking. Spatial agreement between the two approaches, quantified with the Bundle Adjacency metric, ranged from 2.1 to 2.6 mm on average across the three tracts, with the optic radiation showing the closest correspondence at 2.10 mm and the arcuate fasciculus the widest at 2.57 mm. Those figures compare favourably against published benchmarks for conventional probabilistic tractography, where mean Bundle Adjacency between different pipelines has been reported at 4.3 mm, and they approach the roughly 2 mm variability seen when the same protocol is repeated. In other words, the automated pipeline agreed with the manual one about as closely as expert users agree with themselves. The agreement held across a clinically heterogeneous cohort encompassing a range of tumour types, locations and degrees of tract displacement, lending confidence to the approach’s generalisability.

The practical gains were equally significant. The TcT pipeline ran as a single automated script on a standard MacBook Air laptop, requiring no dedicated GPU infrastructure, and completed in approximately 5 to 15 minutes per subject, excluding preprocessing. The automated tumour segmentation step added only about 10 to 15 seconds per patient on GPU-accelerated hardware. Conventional tractography, by contrast, took one to two hours per subject and demanded manual placement of seed, inclusion and exclusion regions for every tract, a process that requires considerable neuroanatomical expertise and introduces inter-rater variability. The authors suggest this speed advantage could prove particularly valuable in time-critical scenarios such as intraoperative MRI, where rapid tract delineation is essential for surgical decision-making when tumours sit close to eloquent white matter pathways.

The researchers are careful to frame the findings as agreement rather than validation. Because ground truth tract anatomy is unavailable in this population, TcT was benchmarked against conventional tractography rather than against an independent standard, and the observed agreement may partly reflect shared anatomical assumptions embedded in the atlas-derived priors that constrain both the segmentation and the tractography. The cohort was also small and dominated by low-grade tumours, which limits generalisability to infiltrative gliomas and high-grade lesions associated with extensive peritumoral oedema, where automated delineation may face greater challenges. Bundle Adjacency, moreover, captures spatial proximity but not topological correspondence. Still, the authors argue that TcT represents a meaningful step toward scalable tract delineation, and future work will evaluate the method in larger multicentre cohorts, extend it to additional clinically relevant tracts, formally assess reproducibility across raters and sites, and explore its integration into prospective surgical planning workflows. If those efforts succeed, the minutes-long mapping of a child’s motor, language and visual highways could become a routine part of the operating theatre’s toolkit.

Subject of Research: Automated streamline-based white matter tract segmentation for pre-operative paediatric brain tumour imaging

Article Title: A rapid streamline-based extension of Tractfinder for white matter tract segmentation

Article References: Kanel, D., Young, F., Seunarine, K. K., Clark, C. A., Aquilina, K., & Clayden, J. D. (2026). A rapid streamline-based extension of Tractfinder for white matter tract segmentation. Frontiers in Neuroimaging, 5, Article 1873040. https://doi.org/10.3389/fnimg.2026.1873040

Image Credits: AI Generated

DOI: 10.3389/fnimg.2026.1873040

Keywords: tractography, white matter, diffusion MRI, paediatric brain tumours, Tractfinder, surgical planning, corticospinal tract, arcuate fasciculus, optic radiation, bundle adjacency, neuroimaging, automation