Purpose An estimated 30%-50% of male individuals with fragile X syndrome (FXS) meet criteria for autism spectrum disorder (ASD), indicating phenotypic overlap but potentially distinct neurobiology. Here, we aimed to characterize shared and divergent cortical features between FXS and ASD.Method High-resolution, motion-corrected quantitative MRI was used to compare cortical morphometry and relaxometry in 61 male participants (9-18 years) with FXS or ASD. To increase power, ASD participants were pooled from multiple MPnRAGE studies and harmonized across protocols using ComBat. Cortical thickness and R1 (longitudinal relaxation rate; proxy for myelination) were computed across the cerebral cortex.Finding Relative to ASD, FXS exhibited greater cortical thickness predominantly in early sensory cortices implicated in low-level visual and auditory processing spanning occipital, parietal, and temporal regions. No significant group differences in R1 were found.Conclusion Thicker cortex in FXS within primary and early associative sensory areas suggests divergent early sensory processing mechanisms between FXS and ASD. Characterizing different neuroanatomical features between the two disorders provides a grounding to develop more disorder-specific interventions despite similar behavioral difficulties. Future work should test developmental trajectories, include females and comorbidities, and link imaging markers to individual sensory/clinical profiles to inform and improve personalized therapies and interventions.

This study assessed relations between white matter microstructure at one month of age and visuospatial processing abilities at six months of age. Participants included ninety-one infants (n = 48 female) born after singleton non-complicated pregnancy. The infants underwent MRI at one month of age and a behavioral assessment at six months of age. Results revealed that the intensity of toy play was associated with fractional anisotropy (FA) of the right (t = 3.002, adjusted p = 0.038) and left superior cerebellar peduncle (t = 2.799, adjusted p = 0.038) and right cingulate gyrus (t = 2.908, adjusted p = 0.0038) while Gaze Shifting was associated with mean diffusivity and axial diffusivity of the right superior cerebellar peduncle (t = 3.099, adjusted p = 0.048). No other correlations, including those studied interactions with biological sex, were significant after adjusting for multiple comparisons. The findings support the notion that developing white matter microstructure, as measured by diffusion MRI metrics, plays a role in the development of visuospatial processing. Future studies should explore how these relationships develop and predict visuospatial ability and working memory later in life.

IntroductionParkinson’s Disease (PD) is diagnosed based on motor symptoms (bradykinesia, resting tremor, rigidity); yet non-motor symptoms such as sleep abnormalities, autonomic dysfunction, and cognitive changes often precede motor signs, fulfilling the criteria for prodromal PD. How motor and non-motor symptoms emerge from dopamine depletion and whether they involve separable neural substrates remains unclear.MethodsWe applied correlational tractography based on multi-shell, diffusion-weighted magnetic resonance imaging in early-stage PD to assess microstructural changes throughout the brain. Eight participants with early-stage PD and 5 healthy controls underwent motor, cognitive, and mood assessments, followed by structural and multi-shell, diffusion-weighted magnetic resonance imaging. Their groupwise differences in white matter integrity associated with PD status were quantified using correlational tractography, with and without age correction.ResultsCorrelational tractography delineated both microstructural changes that held either a significant positive or negative association with PD status, where the statistical maps of these changes linked differentially to motor and non-motor symptoms. Quantitative anisotropy (QA) extracted from positively associated fibers significantly correlated with cognitive function, while QA of negatively associated fibers correlated with motor function—independent of the effect of age. Of note, QA of positively associated fibers correlated with depressive mood only in the age-uncorrected analyses, suggesting a strong age-related effect.ConclusionIn early-stage PD, motor and non-motor symptoms are mapped to anatomically distinct pathways, suggesting separable pathophysiological mechanisms. These findings further suggest that correlational tractography is appropriate to evaluate changes in structural connectivity in neurodegenerative diseases and, potentially, their therapeutic interventions.,

Diffusion time-dependence, defined as variations in diffusivity and/or diffusional kurtosis with diffusion time, has emerged as a valuable non-invasive imaging marker for characterizing tissue microstructural features, such as cell size, density, packing disorder, and membrane permeability. In white matter, diffusion time-dependent changes between the short diffusion time and long diffusion time in radial diffusivity (RD), defined as the diffusivity perpendicular to fiber tracts, were demonstrated to correlate strongly with mean axon diameter in ex vivo spinal cord tissues, and to reveal demyelination in mouse corpus callosum. Despite their potential to non-invasively unveil neuronal microstructures to improve the assessment and targeted therapy of neurological diseases, these novel image contrasts obtained at short diffusion times using oscillating gradient spin echo (OGSE) have only recently become feasible for human in vivo studies with high-performance gradient MRI systems. In this preliminary study, we characterized time-dependent RD with OGSE encoding in the human brain in vivo. The change in radial diffusivity between short diffusion time and long diffusion time (ΔRD) consistently exhibited high values in the corticospinal tract, indicating high sensitivity of ΔRD to large axon diameter in human brains. Imaging at a high OGSE frequency of 100 Hz and a moderate b-value of 800 s/mm2 produced the highest ΔRD in the corticospinal tract. This study established a baseline for future investigations of neuronal microstructural alterations in neurological disorders and diseases.

Purpose :T1-weighted images with different weighting and T1 quantification can be useful in the differentiation of both native and pathologic tissue; however, acquiring multiple images requires substantial time. A rapid and robust brain imaging sequence was developed to simultaneously provide quantitative T1 maps and multiple T1-weighted images. Methods: A novel 3D Cartesian MPnRAGE sequence was developed that collects multiple frames at different inversion times with an interleaved variable density Poisson sampling. Varying flip angles were used to enable B1 corrected T1 quantification from the multiple frames. Test-retest of quantitative T1 and a comparison between Cartesian MPnRAGE and reference values were performed in a phantom. Cartesian MPnRAGE was then acquired in vivo for five human participants. Results: In under 5 min, Cartesian MPnRAGE generated 10 high-quality T1-weighted images of different contrasts across the inversion recovery curve, including white matter nulled, gray matter nulled, and cerebral spinal fluid nulled frames. The 10 inversion times at varying flip angles produced accurate T1 values compared to gold standard T1 fitting, and test-retest scans resulted in low variation. In vivo scan results demonstrate the feasibility of Cartesian MPnRAGE acquisition. Conclusion: This work shows the ability of the Cartesian MPnRAGE sequence to provide accurate and efficient whole brain quantitative T1 mapping and multiple image contrasts in a reasonable acquisition time, offering an alternative acquisition strategy to existing approaches.

Objective: Experiencing discrimination is associated with faster biological aging, as reflected in telomere shortening and DNA methylation. However, the impact of discrimination on brain aging processes remains unclear. Here, we tested whether individuals who reported at least one major lifetime discrimination event would exhibit steeper age-related associations in microstructural metrics within whole-brain white matter and the hippocampus, consistent with accelerated brain microstructural aging, compared with those with no such experiences. Methods: We analyzed multi-shell diffusion-weighted MRI data from the Midlife in the United States (MIDUS) cohort (n=147, mean age=65 years, range: 48 to 95 years) to assess brain microstructure using complementary statistical and biophysical diffusion models. Diffusion kurtosis imaging representation was used to derive diffusion tensor imaging (DTI) and white matter tract integrity (WMTI) measures. Additional microstructural health indices were derived using the neurite orientation dispersion and density imaging (NODDI) model. Permutation analyses of linear models were run within the whole-brain white matter and bilateral hippocampi, adjusting for sex, race, and education. Results: Participants who reported at least one major discriminatory experience during their lifetime exhibited accelerated age-associated changes in white matter microstructural measures, including higher mean and radial diffusivities, extra-axonal radial diffusivity, and free water fraction compared with those with no such experiences. Conclusions: These converging findings from complementary measures of brain microstructure suggest that major discrimination experiences may contribute to accelerated brain microstructural aging.

Background Disrupted brain connectivity is central to understanding the neurobiological basis of autism spectrum disorder (ASD). Diffusion tensor imaging (DTI) has been extensively used to study brain microstructure in ASD, often revealing reduced fractional anisotropy (FA) and increased mean diffusivity (MD) in white matter. Methods We conducted a multivariate random-effects meta-analysis with a multilevel structure to evaluate the extent to which FA and MD measures of white matter microstructure are altered in individuals with ASD compared to typically developing (TD) individuals, as well as how publication year, participant characteristics (age, sex, IQ), and laterality moderate the magnitude of the estimated differences. Our analysis included 881 effect sizes from 66 studies (NASD = 2231, NTD = 1856; Mage = 2-50 years) across 12 white matter tracts. Results We found a significant moderate summary effect size for FA in nine tracts (corpus callosum, corticospinal tract, thalamic radiation, arcuate fasciculus, inferior fronto-occipital fasciculus [IFOF], inferior longitudinal fasciculus [ILF], superior longitudinal fasciculus [SLF], uncinate fasciculus, cingulum; Hedges’ g = -0.30 to -0.50), suggesting that individuals with ASD have lower FA compared to TD controls. Additionally, we found a significant moderate-to-large summary effect size for MD in eight tracts (corpus callosum, corona radiata, corticospinal tract, arcuate fasciculus, IFOF, ILF, SLF, uncinate fasciculus; Hedges’ g = 0.29 to 0.97), suggesting that individuals with ASD have higher MD compared to TD controls. Furthermore, moderators demonstrated tract- and metric-specific effects. Conclusions Our findings highlight the complex, multidimensional nature of white matter microstructure alterations observed in ASD.

Divergent age-related functional brain connectivity in autism spectrum disorder (ASD) has been observed using resting-state fMRI, although the specific findings are inconsistent across studies. Common statistical regression approaches that fit identical models across functional brain networks may contribute to these inconsistencies. Relationships among functional networks have been reported to follow unique nonlinear developmental trajectories, suggesting the need for flexible modeling. Here we apply generalized additive models (GAMs) to flexibly adapt to distinct network trajectories and simultaneously describe divergent age-related changes from childhood into mid-adulthood in ASD. 1107 males, aged 5–40, from the ABIDE I II cross-sectional datasets were analyzed. Functional connectivity was extracted using a network-based template. Connectivity values were harmonized using COMBAT-GAM. Connectivity-age relationships were assessed with thin-plate spline GAMs. Post-hoc analyses defined the age-ranges of divergent aging in ASD. Typically developing (TD) and ASD groups shared 15 brain connections that significantly changed with age (FDR-corrected p < 0.05). Network connectivity exhibited diverse nonlinear age-related trajectories across the functional connectome. Comparing ASD and TD groups, default mode to central executive between-network connectivity followed similar nonlinear paths with no group differences. Contrarily, the ASD group had chronic hypoconnectivity throughout default mode-ventral attentional (salience) and default mode-somatomotor aging trajectories. Within-network somatomotor connectivity was similar between groups in childhood but diverged in adolescence with the ASD group showing decreased within-network connectivity. Network connectivity between the somatomotor network and various other functional networks had fully disrupted age-related pathways in ASD compared to TD, displaying significantly different model curvatures and fits. The present analysis includes only male participants and has a restricted age range, limiting analysis of early development and later life aging, years 40 and beyond. Additionally, our analysis is limited to large-scale network cortical functional parcellation. To parse more specificity of brain region connectivity, a fine-grained functional parcellation including subcortical areas may be warranted. Flexible non-linear modeling minimizes statistical assumptions and allows diagnosis-related brain connections to follow independent data-driven age-related pathways. Using GAMs, we describe complex age-related pathways throughout the human connectome and observe distinct periods of divergence in autism.

Introduction:Chronic widespread musculoskeletal pain (CMP) is a primary condition of Veterans who were deployed to the Persian Gulf War. The mechanisms that underlie CMP in these Veterans are unknown and few efficacious treatment options exist. This study tested the effects of 16 weeks of resistance exercise training (RET) on gray matter (GM) volume and white matter (WM) microstructure in Gulf War Veterans (GWVs) with CMP compared to GWV waitlist controls (WLC). Methods:Fifty-four GWVs were randomly assigned to 16 weeks of RET (n = 28) or WLC (n = 26). Training involved 10 resistance exercises to involve the whole body, was supervised and individually tailored, and progressed slowly to avoid symptom exacerbation. Outcomes assessed at baseline, 6, 11 and 17 weeks and 6- and 12-months post-intervention included GM volume (voxel-based morphometry), WM microstructure (diffusion tensor imaging), pain [short form McGill Pain Questionnaire (SF-MPQ) and 0-100 visual analog scale (VAS)], fatigue (0-100 VAS), and mood (Profile of Mood States). Muscular strength was assessed at baseline, 8 and 16 weeks, and training volume was tracked throughout the 16-week intervention. Primary analyses used linear mixed effects models with Group, Time, and the Group*Time interaction as fixed factors and subject and slope as random factors to test the differential effects of RET and WLC on brain structure and symptoms. All neuroimaging analyses used the False Discovery Rate to correct for multiple comparisons at an alpha of 0.05. Results:Strength increased significantly across the trial for the RET group (p < 0.001). There were significant Group*Time interaction effects for pain ratings (SF-MPQ total; p < 0.01) and the Profile of Mood States total mood disturbance score (p < 0.01). There were no Group or Group*Time effects for GM volume or WM microstructure. There were no significant associations between strength, symptoms, and brain structure (p > 0.05). Conclusion:Sixteen weeks of low-to-moderate intensity RET (i) improved musculoskeletal strength and (ii) did not exacerbate symptoms, but (iii) was insufficient to alter brain structure in GWVs with CMP.

Residence in highly socioeconomically disadvantaged neighborhoods has recently been associated with Alzheimer’s disease (AD) neuropathology at autopsy, cognitive decline, and magnetic resonance imaging (MRI) markers of volumetric brain atrophy in cognitively unimpaired adults. Furthermore, there is mounting evidence that markers of brain microstructure derived from diffusion-weighted MRI (DWI), including neurite density index (NDI), orientation dispersion index (ODI), and isotropic volume fraction (ISO), are sensitive to AD-related neurodegeneration. In this study, we used linear mixed-effects (LME) modeling to investigate the hypothesis that neighborhood-level disadvantage is associated with mixed-longitudinal trajectories of microstructural neurodegeneration in 539 late-middle-aged participants across the AD continuum. 539 participants (Table 1) from the Wisconsin Registry for Alzheimer’s Prevention and the Wisconsin Alzheimer’s Disease Research Center were imaged between 1 and 5 times with multi-shell DWI (constituting 865 total scans). For each scan, average NDI, ODI, and ISO values were extracted from the hippocampus, anterior parahippocampal gyrus, and whole brain cortical gray matter. Geocoded participant addresses were linked to neighborhood disadvantage as measured by the Area Deprivation Index, a marker derived from 17 census indicators of education, employment, poverty, and housing quality. Statewide ADI was binarized for each participant (the highest quintile/most disadvantaged or lowest 4 quintiles/least disadvantaged) and used as a predictor of DWI metrics in LME models with age, sex, diagnosis, education level, and interactions between ADI and age and ADI and diagnosis as covariates. NDI—a measure sensitive to axonal and dendritic loss—was significantly lower (denoting more neurodegeneration) in the most disadvantaged group in all three assessed brain regions (P Corrected < 0.05), and there was a significant interaction between ADI and age (P Corrected < 0.05) for hippocampal NDI (Figures 1-2). No other ADI-DWI associations or ADI interactions survived Benjamini-Hochberg correction for multiple comparisons. Our findings suggest that living in a disadvantaged neighborhood is associated with neuronal degeneration, evidenced by the apparent loss of neurites, and that this process accelerates with age across the AD continuum. As a result, addressing disparities in the social determinants of health may have the potential to reduce the likelihood of neuronal injury in aging adults.

INTRODUCTION:Adults with Down syndrome (DS) accumulate amyloid beta (Aβ) plaques faster and earlier on average than neurotypical adults with sporadic Alzheimer's disease (AD). White matter (WM) microstructure characterized with diffusion tensor imaging (DTI) can indicate underlying architectural changes in longitudinal studies, suggestive of neurodegeneration. This study investigated relationships between DTI and Aβ in DS along the AD continuum. METHODS:Using longitudinal amyloid Pittsburgh compound B positron emission tomography, Centiloid (CL) and DTI parameters were examined in 35 adults with DS ages 25 to 57. DTI measures of anisotropy and diffusivity were analyzed using tract-based spatial statistics and permutation analysis of linear models, testing for significant correlation between the rates of change for CL and DTI. RESULTS:All rates of DTI and Aβ changes were significantly related. Significant regions included the corpus callosum, corona radiata, and long-association fibers. DISCUSSION:Aβ burden is associated with widespread longitudinal WM changes in DS. This suggests WM microstructure alterations accompany amyloid accumulation. HIGHLIGHTS:A Down syndrome-specific template was created. Longitudinal diffusion tensor imaging (DTI) and amyloid burden rates of change correlate. Longitudinal results show more significant regions than cross-sectional results. DTI and amyloid changes were found over two timepoints, 3.7 years apart on average. DTI and amyloid-PET offer greater sensitivity when tracking microstructural changes.

OBJECTIVES:Elevated intracranial pressure (ICP) is a complication of severe traumatic brain injury (TBI) that carries a risk of secondary brain injury. This study investigated the association between ICP burden and brain injury patterns on MRI in children with severe TBI. DESIGN, SETTING, AND PATIENTS:Secondary analysis of the Approaches and Decisions in Acute Pediatric TBI (ADAPT) study, which included children with severe TBI (Glasgow Coma Scale score < 9) who received a clinical MRI within 30 days of injury. We excluded patients who had ICP monitoring less than 24 hours, were missing ICP data for greater than 40% of monitoring time, or who underwent craniectomy. INTERVENTIONS:None. MEASUREMENTS AND MAIN RESULTS:ICP burden was defined as the trapezoidal area under the curve of hourly ICP greater than 20 mm Hg. ICP was standardized to total monitoring time, and patients were categorized to four levels of ICP burden. MRI was evaluated for number of diffuse axonal injury (DAI) microhemorrhages, intracerebral hemorrhage (ICH) volume, contusion volume, and number of regions with ischemia. Fisher exact or chi-square tests were used to test the independence between ICP burden and MRI injury amount. Of the 220 patients, 156 (71%) had DAI, 31 (14%) had ICH, 161 (73%) had contusions, and 70 (32%) had ischemia on MRI. Most patients (180, 82%) experienced episodes of ICP greater than 20 mm Hg. Contusion volume ( p = 0.02) and number of regions with ischemia ( p = 0.007) were associated with ICP burden, but we failed to identify such an association for DAI or ICH. Severe (but not mild or moderate) ICP burden was associated with presence of ischemia (odds ratio, 4.64 [95% CI, 1.30-19.5]; p = 0.02). CONCLUSIONS:Elevated ICP was prevalent in the ADAPT cohort. Ischemia and contusion were associated with the burden of ICP. Further research is needed to determine temporal relationships between elevated ICP and ischemia.

PURPOSE:To extend and automate a data-consistent, self-navigated motion-correction method for 3D radial T1-weighted imaging. METHODS:This method incorporated rigid-body motion effects into the forward model, solving for parameters that maximize consistency with the data. The method was tested on five datasets with a range of motion types and severities. A separate collection of datasets was used to study the effect that the method has on the test-retest reliability of cortical thickness estimates. RESULTS:Image quality was improved across a wide range of distinct motion types, including some cases that would have been unusable if left uncorrected. The error-based weighting scheme and the increased timing resolution afforded by the proposed method were especially useful in cases of extreme and rapid motions. Moreover, the method improved test-retest reliability of cortical thickness measures in pediatric subjects, decreasing the average coefficient of variation from 2 . 73 % ± 1 . 75 % $$ 2.73\%\pm 1.75\% $$ in uncorrected images (with freesurfer failing on one subject) down to 0 . 88 % ± 0 . 21 % $$ 0.88\%\pm 0.21\% $$ for images corrected at ∼ 2 s $$ \sim 2\kern0.3em \mathrm{s} $$ timing resolution and 0 . 79 % ± 0 . 16 % $$ 0.79\%\pm 0.16\% $$ when corrected at faster temporal rates. CONCLUSION:This method was found to be effective when used on T1-weighted radial data, both qualitatively and quantitatively. The fine-scale timing resolution and error-based weighting afforded by this technique will likely provide only a small benefit, unless one is investigating motion-prone populations or is searching for a very small effect size.

Precise characterization of stroke lesions from MRI data has immense value in prognosticating clinical and cognitive outcomes following a stroke. Manual stroke lesion segmentation is time-consuming and requires the expertise of neurologists and neuroradiologists. Often, lesions are grossly characterized for their location and overall extent using bounding boxes without specific delineation of their boundaries. While such characterization provides some clinical value, to develop a precise mechanistic understanding of the impact of lesions on post-stroke vascular contributions to cognitive impairments and dementia (VCID), the stroke lesions need to be fully segmented with accurate boundaries. This work introduces the Multi-Stage Cross-Scale Attention (MSCSA) mechanism, applied to the U-Net family, to improve the mapping between brain structural features and lesions of varying sizes. Using the Anatomical Tracings of Lesions After Stroke (ATLAS) v2.0 dataset, MSCSA outperforms all baseline methods in both Dice and F1 scores on a subset focusing on small lesions, while maintaining competitive performance across the entire dataset. Notably, the ensemble strategy incorporating MSCSA achieves the highest scores for Dice and F1 on both the full dataset and the small lesion subset. These results demonstrate the effectiveness of MSCSA in segmenting small lesions and highlight its robustness across different training schemes for large stroke lesions. Our code is available at: https://github.com/nadluru/StrokeLesSeg.

Alzheimer’s disease (AD) has been mainly thought of as a disease involving gray matter changes. However, despite known correlations between white matter integrity and cognition, less is known about how disruptions to white matter during the development of AD underpin cognitive impairment. This study tests the associations between disruptions to white matter along the AD clinical continuum (cognitive unimpaired (CU): cognitive impaired (CI) – Mild Cognitive Impairment (MCI) and AD) and cognition using diffusion tensor imaging (DTI) and multi-tissue neurite and orientation dispersion and density imaging (mtNODDI) models of the multi-shell connectome diffusion MRI (ms-dMRI) data from the Alzheimer’s Disease Connectome Project (ADCP). Multi-shell connectome diffusion MRI data from 80 participants (55 CU (32 F, mean age 66.7 +/- 6.6; 25 CI (8 F, mean age 73.6 +/- 7.3)) in the ADCP were pre-processed using DESIGNER processing guidelines. Cognition was assessed using the Cognition Total Composite score from the NIH toolbox. Permutation testing was performed using threshold free cluster enhancement and family wise error correction (FWE) and 10,000 permutations. Relationships between cognitive impairment and white matter integrity were deemed statistically significant for FWE corrected p < 0.05. All analyses controlled for age and sex. Our analysis revealed a significant interaction between clinical status and cognitive test outcomes on the mtODI (multi-tissue orientation dispersion index) metric, with higher mtODI levels indicating increased neurodegeneration. CU individuals exhibited higher cognitive performance with lower mtODI levels in cerebellar and brainstem regions, whereas CI individuals showed consistently higher mtODI levels in these regions and lower cognitive performance (Figure 1). Moreover, CI individuals presented with lower overall cognitive test scores compared to their CU counterparts. Analysis of clusters uncorrected for multiple comparisons revealed trend-level differences in frontal and temporal regions (Figures 2-3). Results demonstrate that CI individuals who perform worse than CU individuals on cognitive tests also have higher levels of neurodegeneration than CU individuals. Further research is needed to replicate and extend these findings to other tests of cognitive ability, and to help better understand the potential links between cognitive decline in AD with white matter integrity of cerebellar and brainstem regions.

The design philosophy of leading interventional devices and software typically employs point- or line-like fiducial markers that are detected via multiple external optical cameras and/or the primary medical imaging system employed in the procedure (MRI, CT, fluoroscopy, etc.). Fiducial markers with characteristic geometric signals encoded into their physical design can provide more robust information on device location and orientation than point or line signals at lower imaging resolutions while leveraging the volumetric imaging capabilities of MRI and CT to circumvent the need for additional external camera hardware in the operating room. This work describes efforts to develop unique geometric fiducial designs and software methods to detect and determine their orientation based on volumetric MR-imaging. Using these methods, a single MR scan can be used to localize and orient specific medical devices autonomously, rather than relying on multimodal imaging or semi-manual detection methods. These results have been validated in a series of MRI scans using human cadaver heads and fruit-based phantoms, successfully detecting and isolating the device fiducial in 8/8 T2w MRI scans, and coregistering the signal to an orthonormal template with a Dice-based evaluation metric of 0.819. These methods have the potential to accelerate imageguided surgeries by improving automated detection and registration methods to reduce the number and/or duration of required scans while minimizing the amount of supplemental hardware (external optical cameras).

Accurate segmentation of small brain lesions in magnetic resonance imaging (MRI) is essential for understanding neurological disorders and guiding clinical decisions. However, detecting small lesions remains challenging due to low contrast and limited size. This study proposes two simple yet effective labeling strategies, Multi-Size Labeling (MSL) and Distance-Based Labeling (DBL), that can seamlessly integrate into existing segmentation networks. MSL groups lesions based on volume to enable size-aware learning, while DBL emphasizes lesion boundaries to enhance structural sensitivity. We evaluate our approach on two benchmark datasets: stroke lesion segmentation using the Anatomical Tracings of Lesions After Stroke (ATLAS) v2.0 dataset and multiple sclerosis lesion segmentation using the MSLesSeg dataset. On ATLAS v2.0, our approach achieved higher Dice (+1.3%), F1 (+2.4%), precision (+7.2%), and recall (+3.6%) scores compared to the top-performing method from a previous challenge. On MSLesSeg, our approach achieved the highest Dice score (0.7146) and ranked first among 16 international teams. Additionally, we examined the effectiveness of attention-based and mamba-based segmentation models but found that our proposed labeling strategies yielded more consistent improvements. These findings demonstrate that MSL and DBL offer a robust and generalizable solution for enhancing small brain lesion segmentation across various tasks and architectures. Our code is available at: https://github.com/nadluru/StrokeLesSeg.

Inhibitory control (IC) develops in stages from infancy through adolescence and is associated with numerous developmental disorders and learning outcomes. This study examined how neural architecture - in particular myelination - underlies brain activation patterns observed during IC tasks in a sample of 28 children aged 4-10 years old. IC was observed using reaction times during go/no-go and flanker IC tasks. Myelination was measured using quantitative longitudinal relaxation rate (R1) mapping obtained from selected white matter regions of interest (ROIs). Brain activation was defined as task-related changes in hemoglobin oxygenation as measured by functional near-infrared spectroscopy (fNIRS) averaged within ROIs. Results indicated that myelination in ROIs was higher in older children and fNIRS activation in frontal channels was significantly and positively associated with go/no-go mean reaction time. Myelination in the corona radiata and superior longitudinal fasciculus was positively associated with frontal fNIRS activation, while myelination was negatively associated with go/no-go and flanker mean reaction times across white matter ROIs. Overall, significance level notably varied across models. Independently of inhibitory control constructs, these regions may be of interest in future structure-function studies across development.

While magnetic resonance imaging (MRI) markers of neurodegeneration are nonspecific to Alzheimer’s disease (AD) pathology, they have been correlated with cognitive dysfunction, and therefore, provide important information pertaining to disease staging. Neurodegeneration in AD is commonly assessed with macrostructural measures of brain atrophy, such as hippocampal volume. However, recent investigations have shown that markers of neural microstructure derived from diffusion MRI (DWI) may provide supplementary insight into the progression of AD pathophysiology. Furthermore, DWI measures of neurite density index (NDI), isotropic volume fraction (ISO), return to origin probability (RTOP), and mean squared displacement (MSD) have recently been associated with cerebrospinal fluid markers of senile plaques, neurofibrillary tangles, and AD-related neurodegeneration. In this study, we used linear mixed-effects (LME) modeling to compare the utility of different MRI markers for predicting performance on a three-test Preclinical Alzheimer’s Cognitive Composite (PACC3) in 268 late-middle-aged participants. 268 participants from the Wisconsin Registry for Alzheimer’s Prevention and the Wisconsin Alzheimer’s Disease Research Center (216 unimpaired, 37 MCI, 15 AD) were imaged between 1 and 2 times with T1-weighted MRI and multi-shell DWI (Table 1). For each scan, hippocampal volume was estimated and average NDI, ISO, RTOP, and MSD values were extracted from the hippocampus and anterior parahippocampal gyrus. PACC3 scores were calculated from comprehensive neuropsychological testing within 6 months of imaging. Neuroimaging metrics were used as predictors in separate LME models of the following form: PACC3 = b 0 + b 1 *(metric) + b 2 *(age) + b 3 *(sex) + b 4 *(metric*clinical status) + b 5 *(years of education) + u i DWI metrics consistently outperformed hippocampal volume in predicting PACC3 scores across both brain regions (Table 2). Anterior parahippocampal MSD exhibited the lowest AIC/BIC values, while hippocampal NDI had the strongest association with PACC3 scores among unimpaired participants. Meanwhile, hippocampal volume had the largest AIC/BIC values and the weakest correlation with PACC3 scores (Table 2, Figure 1). Our study provides evidence that neuroimaging markers of brain microstructure may be particularly sensitive to subtle alterations predictive of cognitive performance, especially early in AD. Future work will translate these analyses to larger cohorts and incorporate fluid markers of amyloid and tau.

Accelerated aging is strongly linked to adverse social exposome and accelerated aging of the brain may be a dementia risk factor. Machine-learning can estimate the biological “brain age” from neuroimages, which provides complementary information to the chronological/calendar age. The difference between biological and chronological age is referred to as the “brain age gap.” The Area Deprivation Index is a metric representing 17 indicators of neighborhood level poverty, education, employment, and physical environment, and provides a robust measure of the social exposome. The purpose of this work was to test the relationship between neighborhood-level disadvantage and accelerated brain aging using brain age gap estimates in a Wisconsin cohort. Brain age was estimated using T1-weighted MR images from 1052 participants enrolled in the Wisconsin Registry for Alzheimer’s Prevention or Wisconsin Alzheimer’s Disease Research Center studies. Brain age was estimated using a publicly available deep learning model called the two-stage-age-network (TSAN) that was pre-trained on >4000 scans from OASIS, ADNI-I, and PAC-2019 datasets. Average brain age was used when multiple images were available at the same time point. Participants’ addresses were geolinked to their statewide ranking of neighborhood disadvantage using a time-concordant ADI. Participants were binned by quintile based on state ADI rank, where Quintile 5 comprised of participants with highest ADI (Table 1). In the analyses, Quintile 5 was compared to a grouped category of Quintiles 1-4. A marginal means test was conducted on a fitted one-way ANOVA model to assess differences in mean brain age gaps between Quintile 5 and Quintiles 1-4. The ANOVA model used quintile as a fixed effect predictor variable of brain age gap and a random effect of participant ID to account for participants with multiple visits. The mean brain age gap for Quintile 5 was 0.96 years (-0.79, 2.71), and for Quintiles 1-4 was 0.27 years (-0.12, 0.66) (Figure 1). This analysis found that the mean brain age gap does not significantly differ between Quintile 5 and Quintiles 1-4 (p≤0.45), however statistical power was low given very broad confidence intervals. Future work will test for longitudinal effects and incorporate a larger sample.