Abstract INTRODUCTION Subjective cognitive complaints (SCC) can precede Alzheimer's disease and related dementias. SCC in the absence of objective impairment is termed subjective cognitive decline (SCD). This study aimed to characterize SCC and SCD among a sample of aging individuals with and without prior repetitive head impact (RHI) exposure, the former of whom are at risk for chronic traumatic encephalopathy (CTE). METHODS RHI‐exposed (N = 167) and non–RHI‐exposed (N = 317) Boston University Alzheimer's Disease Research Center (ADRC) participants and their informants completed subjective measures assessing memory and executive function. Participants completed objective tests of these domains. RESULTS RHI exposure was associated with greater self‐ and informant‐reported SCC and with over four‐fold increased odds of SCD among cognitively unimpaired participants (odds ratio = 4.10, p < 0.001). SCC was associated with objective measures in RHI and non‐RHI participants. DISCUSSION Among RHI‐exposed individuals, SCC align with objective cognitive performance. SCD warrants investigation as potential early indicator of RHI‐related neuropathologies.

INTRODUCTION: Accurate MRI-based identification of Alzheimer's disease (AD), mild cognitive impairment (MCI), and related dementias remains challenging because disease-related structural changes are often subtle and heterogeneous. We developed NeuroBridge, a clinically guided multi-task MRI framework for neurodegenerative disease diagnosis. METHODS: NeuroBridge integrates large-scale self-supervised MRI pretraining with hippocampal segmentation, hippocampal atrophy classification, and reconstruction objectives, followed by gated fusion fine-tuning. Performance was evaluated across ADNI and OASIS cohorts, including cross-cohort transfer, probability-based analysis, and opportunistic screening. RESULTS: NeuroBridge achieved the highest performance across evaluated classification tasks, reaching 88.17

IntroductionTransformer-based deep learning has shown great potential in medical imaging, but its real-world applicability remains limited due to the scarcity of annotated data. This study aims to develop a practical framework for the few-shot deployment of pretrained MRI transformers across diverse brain imaging tasks.MethodsWe employ a Masked Autoencoder (MAE) pretraining strategy on a large-scale, multi-cohort brain MRI dataset comprising over 31 million 2D slices to learn transferable representations. For classification tasks, a frozen MAE encoder with a lightweight linear head (MAE-classify) is used. For segmentation, we propose MAE-FUnet, a hybrid architecture that fuses pretrained MAE embeddings with multi-scale CNN features. Extensive evaluations are conducted on multiple datasets, including NACC, ADNI, OASIS, NFBS, SynthStrip, and MRBrainS18, under controlled few-shot settings.ResultsThe proposed framework achieves state-of-the-art performance in MRI sequence classification, reaching an accuracy of 99.24% with only 6,152 trainable parameters. For segmentation tasks, MAE-FUnet consistently outperforms strong baselines, achieving superior Dice and IoU scores across skull stripping and multi-class anatomical segmentation benchmarks. The model also demonstrates enhanced robustness and stability under data-limited conditions, with lower performance variance compared to competing methods.DiscussionThese results highlight the effectiveness of pretrained MAE representations for few-shot medical imaging tasks. The proposed framework enables efficient, scalable, and adaptable deployment of transformer-based models in data-constrained clinical environments. The fusion of global transformer embeddings with local CNN features provides a generalizable design paradigm for a wide range of medical imaging applications.

Accurate differential diagnosis of dementia is essential for guiding timely treatment, particularly as anti-amyloid therapies become more widely available and require precise patient characterization. Here, we developed a radiomics-based machine learning (ML) approach to enhance neuroimaging assessments in distinguishing Alzheimer's disease (AD) from other imaging-evident dementias (OIED). We retrospectively analyzed 1041 individuals from the National Alzheimer's Coordinating Center with confirmed dementia diagnoses and at least one T1 or T2/FLAIR MRI scan. Using FastSurfer and a Lesion Prediction Algorithm, we extracted volumetric and lesion features, which were then used to train ML models. Model performance was compared to the independent evaluations of seven fellowship-trained neuroradiologists. The classifier achieved an AUROC of 0.79 ± 0.01 for AD and 0.66 ± 0.03 for OIED, performing comparably to expert assessments. Interpretation using SHAP values showed strong alignment with imaging features known to align with AD or OIED, respectively. These findings highlight the potential of radiomics to augment neuroimaging workflows.

Recent advancements in deep learning have enabled the development of generalizable models that achieve state-of-the-art performance across various imaging tasks. Vision Transformer (ViT)-based architectures, in particular, have demonstrated strong feature extraction capabilities when pre-trained on large-scale datasets. In this work, we introduce the Magnetic Resonance Image Processing Transformer (MR-IPT), a ViT-based image-domain framework designed to enhance the generalizability and robustness of accelerated MRI restoration. Unlike conventional deep learning models that require separate training for different acceleration factors, MR-IPT is pre-trained on a large-scale dataset encompassing multiple undersampling patterns and acceleration settings, enabling a unified framework. By leveraging a shared transformer backbone, MR-IPT effectively learns universal feature representations, allowing it to generalize across diverse restoration tasks. Extensive experiments demonstrate that MR-IPT outperforms both CNN-based and existing transformer-based methods, achieving superior quality across varying acceleration factors and sampling masks. Moreover, MR-IPT exhibits strong robustness, maintaining high performance even under unseen acquisition setups, highlighting its potential as a scalable and efficient solution for accelerated MRI. Our findings suggest that transformer-based general models can significantly advance MRI restoration, offering improved adaptability and stability compared to traditional deep learning approaches.

Recent advances in MRI reconstruction have demonstrated remarkable success through deep learning-based models. However, most existing methods rely heavily on large-scale, task-specific datasets, making reconstruction in data-limited settings a critical yet underexplored challenge. While regularization by denoising (RED) leverages denoisers as priors for reconstruction, we propose Regularization by Neural Style Transfer (RNST), a novel framework that integrates a neural style transfer (NST) engine with a denoiser to enable magnetic field-transfer reconstruction. RNST generates high-field-quality images from low-field inputs without requiring paired training data, leveraging style priors to address limited-data settings. Our experiment results demonstrate RNST's ability to reconstruct high-quality images across diverse anatomical planes (axial, coronal, sagittal) and noise levels, achieving superior clarity, contrast, and structural fidelity compared to lower-field references. Crucially, RNST maintains robustness even when style and content images lack exact alignment, broadening its applicability in clinical environments where precise reference matches are unavailable. By combining the strengths of NST and denoising, RNST offers a scalable, data-efficient solution for MRI field-transfer reconstruction, demonstrating significant potential for resource-limited settings.

Enlarged perivascular spaces (ePVS) on MRI can signal impaired cerebral fluid clearance and predict dementia risk. Risk factors and biological correlates of ePVS are uncertain partially due to the lack of pathological correlation studies. Repetitive head impacts (RHI) from contact sports might represent one risk factor for ePVS, given their association with vascular pathologies and chronic traumatic encephalopathy (CTE), a neurodegenerative disease characterized by perivascular p-tau aggregates. We examined risk factors, neuropathological, and clinical correlates of antemortem MRI ePVS among brain donors exposed to RHI. The sample included 104 brain donors exposed to RHI from the UNITE brain bank. Clinical MRIs were obtained through medical record requests. A stroke neurologist used established visual rating scales (0=no ePVS, 4= >40 ePVS) to rate ePVS in the centrum semiovale (CS-ePVS) and basal ganglia (BG-ePVS) on axial T2 (n=11, 10.6%) or T1 (n=93, 89%). ePVS were coded as low (1/2) or high burden (3/4). Neuropathological diagnoses were made using established criteria. Years of football served as a proxy for duration of RHI. Antemortem dementia diagnoses were made through consensus conferences. Regression-based analyses tested the association between ePVS with years of football play, various neuropathologies, dementia status and FAQ. Analyses controlled for years from MRI scan to death. Sample characteristics are in Table 1. The most common pathological diagnosis was CTE (n=71, 68%). 34 (32.7%) and 13 (12.5%) had high CS-ePVS and BG-ePVS burden, respectively. More years of football play was associated with greater CS- (OR=1.12, 95% CI=1.03-1.22, p=0.01) and BG-ePVS (OR=1.12, 95 CI=1.01-1.24, p=0.03). Effect sizes remained when age at MRI was included for both CS- and BG-ePVS (OR=1.09) but statistical significance was diminished. Greater CS-ePVS was associated with more severe CTE, arteriosclerosis, atherosclerosis, cerebral amyloid angiopathy and Lewy body disease (Table 2). BG-ePVS was only associated with arteriosclerosis. Greater CS-ePVS were associated with increased odds for having dementia diagnosis (OR=5.41, 95% CI=1.42-20.54, p=0.01). Clinical and pathological correlations were not statistically significant when age at MRI was included. Enlarged PVS might be long-term MRI consequences of exposure to RHI, but their age independent pathological and clinical correlates in this setting remain uncertain.

We describe the rationale, methodology, and design of the Boston University Alzheimer's Disease Research Center (BU ADRC) Clinical Core (CC). The CC characterizes a longitudinal cohort of participants with/without brain trauma to characterize the clinical presentation, biomarker profiles, and risk factors of post-traumatic Alzheimer's disease (AD) and AD-related dementias (ADRD), including chronic traumatic encephalopathy (CTE). Participants complete assessments of traumatic brain injury (TBI) and repetitive head impacts (RHIs); annual Uniform Data Set (UDS) and supplementary evaluations; digital phenotyping; annual blood draw; magnetic resonance imaging (MRI) and lumbar puncture every 3 years; electroencephalogram (EEG); and amyloid and/or tau positron emission tomography (PET) on a subset. As of 3/2025, the CC consists of 467 participants (mean age: 65.6, 50.1% female), including 163 RHI and 302 TBI who completed a UDS 3.0 baseline visit. Common sources of RHI included football (n = 95), soccer (n = 26), ice hockey (n = 17), and military service (n = 46). Most TBIs were mild (97.7%). Eighty-nine percent agreed to brain donation. The BU ADRC CC will facilitate research, education, and training on post-traumatic AD/ADRD. HIGHLIGHTS: The Boston University Alzheimer's Disease Research Center (ADRC) Clinical Core facilitates unique research, education, and training on Alzheimer's disease and Alzheimer's disease-related dementias (AD/ADRD) with a focus on post-traumatic AD/ADRD, including chronic traumatic encephalopathy (CTE). We summarize the rationale, mission, study design, and recent updates for the Clinical Core. As of March 2025, the Clinical Core includes a longitudinal cohort of 467 participants enriched for repetitive head impacts (∼1/3) and traumatic brain injury (∼1/3) exposure who span the cognitive continuum, with most having available fluid and neuroimaging biomarker data and agreeing to brain donation (89%).

Repetitive head impacts (RHI) from contact sports can lead to long-term white matter injury visualized on FLAIR scans as white matter hyperintensities (WMH). The goal of this study was to preliminarily characterize the unique pattern and features of WMH in middle aged- to older adults with remote history of exposure to RHI from contact sports. 76 participants (38 with substantial RHI, 38 with minimal or no RHI) from the Boston University Alzheimer’s Disease Research Center had a FLAIR MRI during their annual study visit. The presence of moderate-severe WMH was adjudicated by a panel of clinicians (including of a neurologist and neuroradiologist) during a multidisciplinary diagnostic consensus. Previously, we observed small, spherical, discrete lesions proximal to the gray matter in individuals exposed to RHI. Therefore, the number of discrete, spherical WMH within 1.0cm of the cortex was counted. Regression and analysis of variance models examined group effects as well as years of American football play (proxy for duration of RHl), controlling for age at MRI and vascular risk factors. Sample characteristics are in Table 1. Sources of RHI were American football, hockey, soccer, wrestling, field hockey, lacrosse, mixed martial arts, and rugby. The RHI group was judged to have greater burden of WMH compared to non-RHI (n=15 vs n=5, p=0.042). Years of football play was associated with greater odds for having moderate-severe WMH (OR =1.10, p=0.023). RHI participants had a greater number of unique, distinguishable WMH within 0.5cm (p=0.006) and 1.0cm (p=0.008) of the cortex (Figure 1). Qualitatively, we visualized small, punctate lesions close to the gray matter, clustered around the depths of the sulci in the RHI cohort, distinct from the patterns typically seen with chronic ischemic disease and neuro-inflammatory conditions (Figure 2). We propose RHI induces a unique pattern of WMH characterized by small, spherical, punctate lesions proximal to the deep gray matter distributed throughout the cerebrum. The lesion location corresponds to areas susceptible to RHI and p-tau in chronic traumatic encephalopathy. Future research targets include quantitative analysis to better characterize features and patterns of WMH, clarify specificity to RHI, and study biological correlates.

Deep learning-based MRI reconstruction models have achieved superior performance these days. Most recently, diffusion models have shown remarkable performance in image generation, in-painting, super-resolution, image editing and more. As a generalized diffusion model, cold diffusion further broadens the scope and considers models built around arbitrary image transformations such as blurring, down-sampling, etc. In this paper, we propose a k-space cold diffusion model that performs image degradation and restoration in k-space without the need for Gaussian noise. We provide comparisons with multiple deep learning-based MRI reconstruction models and perform tests on a well-known large open-source MRI dataset. Our results show that this novel way of performing degradation can generate high-quality reconstruction images for accelerated MRI.

Differential diagnosis of dementia remains a challenge in neurology due to symptom overlap across etiologies, yet it is crucial for formulating early, personalized management strategies. Here, we present an artificial intelligence (AI) model that harnesses a broad array of data, including demographics, individual and family medical history, medication use, neuropsychological assessments, functional evaluations and multimodal neuroimaging, to identify the etiologies contributing to dementia in individuals. The study, drawing on 51,269 participants across 9 independent, geographically diverse datasets, facilitated the identification of 10 distinct dementia etiologies. It aligns diagnoses with similar management strategies, ensuring robust predictions even with incomplete data. Our model achieved a microaveraged area under the receiver operating characteristic curve (AUROC) of 0.94 in classifying individuals with normal cognition, mild cognitive impairment and dementia. Also, the microaveraged AUROC was 0.96 in differentiating the dementia etiologies. Our model demonstrated proficiency in addressing mixed dementia cases, with a mean AUROC of 0.78 for two co-occurring pathologies. In a randomly selected subset of 100 cases, the AUROC of neurologist assessments augmented by our AI model exceeded neurologist-only evaluations by 26.25%. Furthermore, our model predictions aligned with biomarker evidence and its associations with different proteinopathies were substantiated through postmortem findings. Our framework has the potential to be integrated as a screening tool for dementia in clinical settings and drug trials. Further prospective studies are needed to confirm its ability to improve patient care. Drawing on 51,269 participants across 9 independent, geographically diverse datasets, an AI model identifies the etiologies contributing to dementia in individuals, harnessing a broad array of data, including demographics, medical history, medication use, neuropsychological assessments, functional evaluations, and multimodal neuroimaging.

OBJECTIVE:To evaluate whether preterm infants with prenatal opioid exposure had differences in brain size on head ultrasounds (HUS) in comparison to non-exposed infants. STUDY DESIGN:Preterm infants ≤34 weeks with prenatal opioid exposure (n = 47) and matched non-exposed infants (n = 62) with early HUSs were examined. Fifteen brain measurements were made and linear regression models performed to evaluate differences. RESULTS:Brain measurements were smaller in the right ventricular index [β = -0.18 mm (95% CI -0.32, -0.03]), left ventricular index [β = -0.04 mm (95% CI -0.08, -0.003)], left basal ganglia insula [β = -0.10 mm (95% CI -0.15, -0.04)], right basal ganglia insula [β = -0.08 mm (95% CI -0.14, -0.03)], corpus callosum fastigium length [β = -0.16 mm (95% CI -0.25, -0.06)], intracranial height index [β = -0.31 mm (95% CI -0.44, -0.18)], and transcerebellar measurements [β = -0.13 (95% CI -0.25, -0.02)] in the opioid-exposed group. CONCLUSIONS:Preterm infants with prenatal opioid exposure have smaller brain sizes compared to non-exposed infants, potentially increasing their risk for neurodevelopmental abnormalities.

Magnetic Resonance Imaging (MRI) acceleration techniques using k-space sub-sampling (KS) can greatly improve the efficiency of MRI-based stroke diagnosis. Although Deep Neural Networks (DNN) have shown great potential on stroke lesion recognition tasks when the MR images are reconstructed from the full k-space, they are vulnerable to the lower quality MR images generated by KS. In this paper, we propose a Distributionally Robust Learning (DRL) approach to improve the performance of stroke recognition DNN models when the MR images are reconstructed from the sub-sampled k-space. For Convolutional Neural Network (CNN) and Vision Transformer (ViT)-based models, our methods improve the stroke classification AUROC and AUPRC by up to 11.91% and 9.32% on the KS-perturbed brain MR images, respectively, compared against Empirical Risk Minimization (ERM) and other baseline defensive methods. We further show that DRL models can successfully recognize the stroke cases from highly perturbed MR images where clinicians may fail, which provides a solution for improved diagnosis in an accelerated MRI setting.

Objective: To compare visually-rated macrostructural features on MRIs between autopsy-confirmed CTE and participants with AD. Background: Biomarkers that can accurately detect the neurodegenerative disease chronic traumatic encephalopathy (CTE) do not yet exist. Structural MRI is an integral component to the in vivo detection of Alzheimer's disease (AD) and related disorders but currently unclear usefulness in identifying CTE. MRI features of CTE have been previously characterized through comparison to participants with normal cognition. The specificity of those findings to CTE (versus alternative neurodegenerative diseases) is uncertain due to lack of disease comparison groups like AD. Design/Methods: The sample included 63 brain donors with autopsy-confirmed CTE and 35 participants with AD (7 autopsy-confirmed, 28 AD dementia). Participants were all males, ≥60 years. MRIs were obtained through medical records. Three neuroradiologists used visual rating scales (0=absent, 4=severe) to rate atrophy on T1 and microvascular disease on T2-FLAIR. Cavum septum pellucidum (CSP) presence was rated. Majority rating was used; median was used in the absence of majority. Analysis of covariance controlling for age at MRI compared groups on atrophy and microvascular disease ratings. Binary logistic regression was used for absent/present CSP. Results: Of the 63 with CTE, 56 had high stage and donors with CTE were four years younger than AD (71.51, SD=7.7 vs 75.6, SD=7.4). Compared with AD, CTE had higher anterior temporal lobe atrophy ratings (mean diff=1.01, 95%CI=0.08–1.94) and a 5.2X (95% CI=1.02–26.90) increase odds for having a CSP. There were statistical trends for greater dorsolateral frontal atrophy (mean diff=0.88, 95%CI=-0.06–1.81), larger third ventricle (mean diff=0.46, 95%CI=-0.07–0.99), and greater periventricular (mean diff=0.39, 95% CI=-0.05–0.83) and deep white matter hyperintensities (mean diff=0.42, 95%CI=-0.05–0.89) in CTE. There were no effects for parietal-occipital and medial temporal lobes, lateral ventricles, corpus callosum, or microbleeds (p>0.10). Conclusions: Frontotemporal atrophy and CSP on MRI might facilitate differential diagnosis of CTE from AD. Disclosure: Miss Mosaheb has nothing to disclose. Dr. Mian has received personal compensation for serving as an employee of Boston Imaging Core Lab. Dr. Mian has received personal compensation in the range of $0-$499 for serving on a Speakers Bureau for Biogen. Dr. Mian has stock in Boston Imaging Core Lab. Dr. Mian has received intellectual property interests from a discovery or technology relating to health care. Mr. Farris has nothing to disclose. Karen Buch, MD has nothing to disclose. Breton Asken has nothing to disclose. Dr. Rabinovici has received personal compensation in the range of $10,000-$49,999 for serving as a Consultant for Eisai. Dr. Rabinovici has received personal compensation in the range of $500-$4,999 for serving as a Consultant for GE Healthcare. Dr. Rabinovici has received personal compensation in the range of $500-$4,999 for serving as a Consultant for Roche. Dr. Rabinovici has received personal compensation in the range of $500-$4,999 for serving as a Consultant for Genentech. Dr. Rabinovici has received personal compensation in the range of $500-$4,999 for serving as a Consultant for Eli Lilly. Dr. Rabinovici has received personal compensation in the range of $500-$4,999 for serving on a Scientific Advisory or Data Safety Monitoring board for Johnson & Joihnson. Dr. Rabinovici has received personal compensation in the range of $10,000-$49,999 for serving as an Editor, Associate Editor, or Editorial Advisory Board Member for JAMA Neurology. The institution of Dr. Rabinovici has received research support from NIH. The institution of Dr. Rabinovici has received research support from American College of Radiology. The institution of Dr. Rabinovici has received research support from Alzheimer's Association. The institution of Dr. Rabinovici has received research support from Rainwater Charitable Foundation. The institution of Dr. Rabinovici has received research support from Genentech. Dr. Rabinovici has received personal compensation in the range of $5,000-$9,999 for serving as a Topic Chair, Course Director and teacher with AAN. Dr. Rabinovici has received personal compensation in the range of $500-$4,999 for serving as a Grant reviewer with NIH. Madeline Uretsky has nothing to disclose. Yorghos Tripodis, 5406 has nothing to disclose. Brett Martin has nothing to disclose. Joseph Palmisano has nothing to disclose. Dr. Kowall has received publishing royalties from a publication relating to health care. Bertrand Huber has nothing to disclose. Robert Stern, PhD has received personal compensation in the range of $10,000-$49,999 for serving as a Consultant for Biogen. The institution of Robert Stern, PhD has received research support from Eisai. The institution of Robert Stern, PhD has received research support from Lilly. The institution of Robert Stern, PhD has received research support from ATRI/NIA. Robert Stern, PhD has received publishing royalties from a publication relating to health care. Robert Stern, PhD has a non-compensated relationship as a Member with NFLPA Mackey-White Committee that is relevant to AAN interests or activities. Dr. Killiany has nothing to disclose. Dr. Stein has nothing to disclose. Dr. McKee has nothing to disclose. The institution of Dr. Mez has received research support from NIH, DOD. The institution of Michael Alosco, PHD has received research support from NIH.

The application of compressed sensing (CS)-enabled data reconstruction for accelerating magnetic resonance imaging (MRI) remains a challenging problem. This is due to the fact that the information lost in k-space from the acceleration mask makes it difficult to reconstruct an image similar to the quality of a fully sampled image. Multiple deep learning-based structures have been proposed for MRI reconstruction using CS, in both the k-space and image domains, and using unrolled optimization methods. However, the drawback of these structures is that they are not fully utilizing the information from both domains (k-space and image). Herein, we propose a deep learning-based attention hybrid variational network that performs learning in both the k-space and image domains. We evaluate our method on a well-known open-source MRI dataset (652 brain cases and 1172 knee cases) and a clinical MRI dataset of 243 patients diagnosed with strokes from our institution to demonstrate the performance of our network. Our model achieves an overall peak signal-to-noise ratio/structural similarity of 40.92 ± 0.29/0.9577 ± 0.0025 (fourfold) and 37.03 ± 0.25/0.9365 ± 0.0029 (eightfold) for the brain dataset, 31.09 ± 0.25/0.6901 ± 0.0094 (fourfold) and 29.49 ± 0.22/0.6197 ± 0.0106 (eightfold) for the knee dataset, and 36.32 ± 0.16/0.9199 ± 0.0029 (20-fold) and 33.70 ± 0.15/0.8882 ± 0.0035 (30-fold) for the stroke dataset. In addition to quantitative evaluation, we undertook a blinded comparison of image quality across networks performed by a subspecialty trained radiologist. Overall, we demonstrate that our network achieves a superior performance among others under multiple reconstruction tasks.

A central goal of modern magnetic resonance imaging (MRI) is to reduce the time required to produce high-quality images. Efforts have included hardware and software innovations such as parallel imaging, compressed sensing, and deep learning-based reconstruction. Here, we propose and demonstrate a Bayesian method to build statistical libraries of magnetic resonance (MR) images in k-space and use these libraries to identify optimal subsampling paths and reconstruction processes. Specifically, we compute a multivariate normal distribution based upon Gaussian processes using a publicly available library of T1-weighted images of healthy brains. We combine this library with physics-informed envelope functions to only retain meaningful correlations in k-space. This covariance function is then used to select a series of ring-shaped subsampling paths using Bayesian optimization such that they optimally explore space while remaining practically realizable in commercial MRI systems. Combining optimized subsampling paths found for a range of images, we compute a generalized sampling path that, when used for novel images, produces superlative structural similarity and error in comparison to previously reported reconstruction processes (i.e. 96.3% structural similarity and < 0.003 normalized mean squared error from sampling only 12.5% of the k-space data). Finally, we use this reconstruction process on pathological data without retraining to show that reconstructed images are clinically useful for stroke identification. Since the model trained on images of healthy brains could be directly used for predictions in pathological brains without retraining, it shows the inherent transferability of this approach and opens doors to its widespread use.

Background: Cerebral small vessel disease (SVD) is common in older people and causes lacunar stroke and vascular cognitive impairment. Risk factors include old age, hypertension and variants in the genes COL4A1/COL4A2 encoding collagen alpha-1(IV) and alpha-2(IV), here termed collagen-IV, which are core components of the basement membrane. We tested the hypothesis that increased vascular collagen-IV associates with clinical hypertension and with SVD in older persons and with chronic hypertension in young and aged primates and genetically hypertensive rats. Methods: We quantified vascular collagen-IV immunolabeling in small arteries in a cohort of older persons with minimal Alzheimer pathology (N=52; 21F/31M, age 82.8±6.95 years). We also studied archive tissue from young (age range 6.2–8.3 years) and older (17.0–22.7 years) primates ( M mulatta ) and compared chronically hypertensive animals (18 months aortic stenosis) with normotensives. We also compared genetically hypertensive and normotensive rats (aged 10–12 months). Results: Collagen-IV immunolabeling in cerebral small arteries of older persons was negatively associated with radiological SVD severity (ρ: −0.427, P =0.005) but was not related to history of hypertension. General linear models confirmed the negative association of lower collagen-IV with radiological SVD ( P <0.017), including age as a covariate and either clinical hypertension ( P <0.030) or neuropathological SVD diagnosis ( P <0.022) as fixed factors. Reduced vascular collagen-IV was accompanied by accumulation of fibrillar collagens (types I and III) as indicated by immunogold electron microscopy. In young and aged primates, brain collagen-IV was elevated in older normotensive relative to young normotensive animals ( P =0.029) but was not associated with hypertension. Genetically hypertensive rats did not differ from normotensive rats in terms of arterial collagen-IV. Conclusions: Our cross-species data provide novel insight into sporadic SVD pathogenesis, supporting insufficient (rather than excessive) arterial collagen-IV in SVD, accompanied by matrix remodeling with elevated fibrillar collagen deposition. They also indicate that hypertension, a major risk factor for SVD, does not act by causing accumulation of brain vascular collagen-IV.

Background and Objectives Late neuropathologies of repetitive head impacts from contact sports can include chronic traumatic encephalopathy (CTE) and white matter degeneration. White matter hyperintensities (WMH) on fluid-attenuated inversion recovery (FLAIR) MRI scans are often viewed as microvascular disease from vascular risk, but might have unique underlying pathologies and risk factors in the setting of repetitive head impacts. We investigated the neuropathologic correlates of antemortem WMH in brain donors exposed to repetitive head impacts. The association between WMH and repetitive head impact exposure and informant-reported cognitive and daily function were tested. Methods This imaging–pathologic correlation study included symptomatic male decedents exposed to repetitive head impacts. Donors had antemortem FLAIR scans from medical records and were without evidence of CNS neoplasm, large vessel infarcts, hemorrhage, or encephalomalacia. WMH were quantified using log-transformed values for total lesion volume (TLV), calculated using the lesion prediction algorithm from the Lesion Segmentation Toolbox. Neuropathologic assessments included semiquantitative ratings of white matter rarefaction, cerebrovascular disease, hyperphosphorylated tau (p-tau) severity (CTE stage, dorsolateral frontal cortex), and β-amyloid (Aβ). Among football players, years of play was a proxy for repetitive head impact exposure. Retrospective informant-reported cognitive and daily function were assessed using the Cognitive Difficulties Scale (CDS) and Functional Activities Questionnaire (FAQ). Regression models controlled for demographics, diabetes, hypertension, and MRI resolution. Statistical significance was defined as p ≤ 0.05. Results The sample included 75 donors: 67 football players and 8 nonfootball contact sport athletes or military veterans. Dementia was the most common MRI indication (64%). Fifty-three (70.7%) had CTE at autopsy. Log TLV was associated with white matter rarefaction (odds ratio [OR] 2.32, 95% confidence interval [CI] 1.03, 5.24; p = 0.04), arteriolosclerosis (OR 2.38, 95% CI 1.02, 5.52; p = 0.04), CTE stage (OR 2.58, 95% CI 1.17, 5.71; p = 0.02), and dorsolateral frontal p-tau severity (OR 3.03, 95% CI 1.32, 6.97; p = 0.01). There was no association with Aβ. More years of football play was associated with log TLV (unstandardized β 0.04, 95% CI 0.01, 0.06; p = 0.01). Greater log TLV correlated with higher FAQ (unstandardized β 4.94, 95% CI 0.42, 8.57; p = 0.03) and CDS scores (unstandardized β 15.35, 95% CI −0.27, 30.97; p = 0.05). Discussion WMH might capture long-term white matter pathologies from repetitive head impacts, including those from white matter rarefaction and p-tau, in addition to microvascular disease. Prospective imaging–pathologic correlation studies are needed. Classification of Evidence This study provides Class IV evidence of associations between FLAIR white matter hyperintensities and neuropathologic changes (white matter rarefaction, arteriolosclerosis, p-tau accumulation), years of American football play, and reported cognitive symptoms in symptomatic brain donors exposed to repetitive head impacts.