Brian Edlow
Vice Chair of Research
Make your next move.
Prepare your application
Review drafts before you send them.
Explore related searches
Articles (28)
United States Special Operations Forces (SOF) personnel endure repeated blasts throughout training and combat. Recent human postmortem, neuroimaging, and blood proteomic work suggests that tau pathology is present following repeated blast exposure. This study aimed to determine whether blood tau markers are associated with brain tau paired helical filaments (PHF) in SOF personnel. Twenty-eight active-duty SOF completed a positron emission tomography-magnetic resonance imaging scan with the PHF-specific radiotracer fluorine-18 MK6240 ([18F]MK6240) and provided blood samples to measure total tau and phosphorylated tau181 (p-tau181). Whole brain voxel-wise analysis showed that higher total tau in the blood was associated with higher [18F]MK6240 uptake in the left temporal cortex, parahippocampal gyrus and hippocampus. We performed post hoc analyses to assess whether brain or blood tau measures were associated with memory performance. Higher levels of blood total tau and p-tau181 were associated with a longer response time and lower throughput (i.e., fewer accurate responses per minute) during the Code Substitution-Delayed test, a visual memory task in the Automated Neuropsychological Assessment Metric (ANAM). This study provides preliminary evidence in active-duty SOF that blood total tau is associated with regional [18F]MK6240 uptake in the brain and that blood total tau and p-tau181 are associated with memory performance.
Year:
2026
Clinical interventions and neuroimaging in the subcortex require anatomical definitions that exceed the resolution and anatomical detail of currently available deformable brain atlases. Here, we introduce a high-resolution human brain atlas comprising 95 manually segmented grey and white matter structures as well as 82 white matter tracts compiled from a multitude of resources including ex-vivo MRI, histology, fibre dissections, and neuroanatomy textbooks. The atlas is defined at an isotropic resolution of 100 μm and can be precisely deformed to individual subject brain anatomy. By providing precise definitions of both grey and white matter structures within and around the basal ganglia, thalamus, subthalamus, midbrain and cerebellum, the atlas provides a foundational resource for stereotactic surgery and subcortical brain imaging research, as well as for development of next-generation neuromodulation strategies.
Year:
2026
Brain network mapping plays a crucial role in advancing our understanding of human brain organization and the neuroanatomic foundations of cognition. Historically, the identification of large-scale brain networks has focused on the cerebral cortex. Some cortical networks have been fractionated into subnetworks, yielding valuable insights into their domain-specific cognitive functions. In contrast, functional mapping of large-scale brain networks within subcortical regions remains an emerging and challenging field, hindered by a low signal-to-noise ratio in subcortical functional MRI data and an inability to distinguish networks with substantial spatiotemporal overlap. In this study, we fractionated and identified fifteen spatially overlapped and temporally correlated subnetworks, which can be categorized into four large-scale brain networks. The subcortical functional connectivity patterns of these subnetworks exhibited distinct, yet overlapping, spatial configurations, with widely connected hub nodes identified in the caudate, putamen, hippocampus, and thalamus. These subnetworks are highly reproducible across healthy human brains and provide normative functional atlases, which we release here as a resource for the academic community. As a proof-of-principle demonstration of how the atlases can be used to elucidate the pathophysiology of neuropsychiatric disorders, we show that the spatial patterns of the subnetworks predict the level of consciousness in patients with severe traumatic brain injury. These observations indicate a highly conserved and spatially overlapped subcortical functional architecture in the human brain, providing opportunities to elucidate pathophysiologic mechanisms and develop new neuromodulatory therapies for a broad spectrum of neuropsychiatric diseases.
Year:
2025
Objective This study evaluates contusion volume and location as predictors of PTE while considering mortality as a competing risk using automated segmentation of acute CT images. Methods Adult TBI patients who visited our center (2014-2025) were identified and categorized into three outcome groups: PTE, post-TBI mortality, and event-free. Clinical covariates were extracted from health records, and CT scans were processed using the BLAST-CT algorithm to segment intraparenchymal hemorrhage (IPH), edema, intraventricular (IVH), and extra-axial hemorrhage (EAH). Five-fold nested cross-validation was performed, with L1-regularized regression used for feature selection within each training fold and model performance evaluated in the corresponding test fold. Within each fold, the selected features were then used to fit separate cause-specific multivariable Cox models for PTE and mortality. Contusion location associations were assessed using both multivariable Cox and sparse canonical correlation analysis. Results Of 1017 identified patients, 6.1% developed PTE, 10.5% died, and 83.4% had no event. In multivariable analysis, total contusion volumes (p<0.001) were independently predicted both PTE and mortality, meanwhile and EAH volume (p<0.01) predicted mortality only. When regionalized, frontal (p<0.01) and temporal (p<0.01) contusions independently predicted PTE, whereas frontal (p=0.04) contusions also independently predicted mortality. Sparse canonical correlation analysis identified an association of frontotemporal and insular contusions with PTE but not mortality. Interpretation Automatically measured contusion segmentation from acute CT imaging could help predict post-traumatic epilepsy. A competing risk framework highlighted that contusion volume is associated with both PTE and mortality risk, whereas unique contusion location patterns may aid in post-traumatic epilepsy assessment in a broad TBI population.
Year:
2025
Predicting outcome for patients with acute severe traumatic brain injury (TBI) is imprecise, relying on neurological examination, structural neuroimaging, and resting-state electroencephalography (EEG) to serve as proxies for brain function. We implemented a multimodal assessment protocol to determine whether detection of consciousness using advanced tools, including standardized behavioral evaluation, advanced EEG, and functional magnetic resonance imaging (fMRI), is associated with functional outcome in patients with acute severe TBI admitted to the intensive care unit (ICU). We tested the association between 6-month Disability Rating Scale (DRS) scores and acute level of consciousness on the Coma Recovery Scale-Revised (CRS-R); responses to active-motor-imagery and passive-language EEG and fMRI; and resting-state fMRI default mode network (DMN) connectivity. We consecutively enrolled 55 patients with acute severe TBI. Six-month outcome was available in 45 (45.2 ± 20.7 years old, 70% male), of whom 10 died, all due to withdrawal of life-sustaining treatment (WLST). All deaths occurred in participants who were not behaviorally following commands. Lower (i.e., better) 6-month DRS scores were observed in participants who were younger (ρ = 0.401; 95% confidence interval [CI: 0.078, 0.649]; p = 0.006]) and who were conscious (median DRS score difference [95% CI]: −11 [−20, −2]; p = 0.003) or following commands (−9 [−20, −1]; p = 0.011) on CRS-R assessment in the ICU. Evidence of command-following on EEG or fMRI did not strengthen this relationship. In participants without command-following on the CRS-R, we detected responses to active-motor-imagery (i.e., cognitive motor dissociation [CMD]) in 6/34 (18%) of participants on EEG and 8/24 (33%) participants on fMRI. We detected responses to passive-language stimuli (i.e., covert cortical processing [CCP]) in 30/33 (91%) of participants on EEG and 18/24 (75%) on fMRI. However, neither CMD nor CCP was associated with outcome on the DRS or several secondary outcome measures (e.g., dichotomous DRS, Glasgow Outcome Scale-Extended), an unexpected result that may reflect the modest sample size and high rate of WLST. In exploratory analyses, an intact DMN and the magnitude of DMN connectivity were associated with 6-month outcome on secondary outcome measures. Collectively, our results suggest that the level of consciousness in the ICU, assessed with a standardized behavioral measure, may predict recovery from severe TBI. Further research, conducted at multiple sites and with larger samples, is required to determine whether integrating behavioral, EEG, and fMRI biomarkers of consciousness is more predictive than behavioral assessment alone.
Year:
2025
Structural brain analysis at the subregion level offers critical insights into healthy aging and neurodegenerative diseases. The NextBrain histological atlas was recently introduced to support such fine-grained investigations, but its existing Bayesian segmentation framework remains computationally prohibitive, particularly for large-scale studies. We present a new, open-source tool that dramatically accelerates segmentation using a hybrid approach combining: machine learning, contrast-adaptive segmentation; target-specific image synthesis; and fast diffeomorphic registration (all three with GPU support). Our method enables highly granular segmentation of brain MRI scans of any resolution and contrast ( in vivo or ex vivo ) at a fraction of the computational cost of the original method ( < 5 minutes on a GPU). We validate our tool on four different modalities ( in vivo MRI, ex vivo MRI, HiP-CT, and photography) across a total of approximately 4,000 brain scans. Our results demonstrate that the accelerated approach achieves comparable accuracy to the original method in terms of Dice scores, while reducing runtime by over an order of magnitude. This work enables high-resolution anatomical analysis at unprecedented scale and flexibility, providing a practical solution for large neuroimaging studies. Our tool is publicly available in FreeSurfer ( https://surfer.nmr.mgh.harvard.edu/fswiki/HistoAtlasSegmentation ).
Year:
2025
Background Acute ischemic stroke (AIS) can lead to profound disturbances in consciousness, including coma, which is associated with poor prognosis and increased mortality. Clarifying the lesion patterns that precipitate loss of consciousness can refine pathophysiological models and guide prognosis. Objectives In this study, we aim to identify the brain regions most commonly affected in comatose AIS and determine whether specific combinations of lesions are necessary and sufficient to produce coma. Methods We retrospectively analyzed 476 AIS patients (52 comatose) using diffusion-weighted imaging. Infarcts were automatically segmented, manually verified, and normalized to MNI space. Support vector regression lesion-symptom mapping (SVR-LSM) quantified voxel-wise associations with coma, controlling for lesion volume. To assess the necessity and sufficiency of lesion combinations, we employed permutation-based nested logistic regression models comparing all subsets of four anatomical predictors: brainstem, thalamus, cerebellum, and the rest of brain lesions. Results SVR-LSM revealed that coma was strongly associated with lesions involving the brainstem, thalamus, and cerebellum, whereas non-comatose patients exhibited predominantly cortical infarcts. Nested model comparisons showed that concurrent lesions to both the brainstem and thalamus were necessary and sufficient for coma. Additional involvement of the cerebellum or cerebral cortex did not improve predictive performance. Conclusions Coma after AIS results from a dual-node subcortical lesion pattern involving both the brainstem and thalamus. Cerebellar and cortical lesions, even when extensive, did not induce coma in the absence of the dual-brainstem and thalamic lesions. These observations emphasize the predominant role of lesion location over lesion volume in the pathogenesis of coma. They also support mechanistic models that position the brainstem and thalamic hubs as central to the neural circuitry underlying arousal. Furthermore, these findings delineate a specific anatomical substrate that may serve as a strategic target for circuit-based neuroprotective and neuromodulatory therapies.
Year:
2025
Objective Emotional dysfunction is a common consequence of severe traumatic brain injury (TBI), yet the mechanisms underlying these symptoms remain poorly understood. This study investigated whether brain network and autonomic mechanisms involved in emotional processing are abnormal in TBI. Methods We conducted a cross-sectional study of chronic severe TBI (n=26) and healthy control participants (n=15). We analysed functional MRI (fMRI) data to assess brain processing of emotionally salient music (joyful and fearful stimuli; n=15 TBI, n=15 controls), and resting-state fMRI (rsfMRI) to measure the functional connectivity of relevant intrinsic brain networks (limbic, salience, and default mode networks; n=16 TBI, n=15 controls). We additionally measured the pupillary light reflect (PLR) to assess parasympathetic and sympathetic function (n=14 TBI, n=11 controls). Results Individuals with severe TBI did not demonstrate the left insula activation elicited by joyful versus fearful musical stimuli seen in healthy controls. rsfMRI revealed decreased connectivity between the salience network, caudate and hippocampus in severe TBI compared to controls. Exploratory analyses identified reduced connectivity between default mode (bilateral medial prefrontal cortex) and limbic (bilateral amygdala) nodes in TBI compared to controls. PLR measurements revealed blunted dark-adaptation responses in individuals with severe TBI compared to controls (F(1,24)=27.4, p<0.001). Interpretation Individuals with chronic severe TBI show reduced insula activation during emotional stimuli processing, resting connectivity abnormalities in salience, limbic and default mode networks, and evidence of sympathetic dysfunction. Brain network and autonomic alterations may be potential neural mechanisms of post-TBI emotional dysregulation.
Year:
2025
Background Emotional disturbances are common after traumatic brain injury (TBI), yet the neuroanatomic basis of these sequalae remains unclear. We aimed to determine whether emotional wellbeing is associated with the extent and/or location of structural brain injury in individuals with chronic TBI. Methods We studied 188 adults with chronic (≥1 year post-injury) TBI enrolled in the multi-center Late Effects of TBI (LETBI) study. Participants underwent T1-weighted MRI to assess cortical lesion location and volume, and diffusion MRI to measure global mean white matter fractional anisotropy (FA), defined as the mean FA across 40 white matter tracts. The primary outcome was emotional wellbeing, measured as the RAND-36 emotional wellbeing subscale t score. To identify associations between structural injury and the primary outcome, we used multivariate linear models adjusted for age, sex, years of education, time since injury, history of psychiatric illness, and multiple indices of scan quality. Results The cohort was predominantly male (68%) and white (88%), with a mean age of 58 years (SD 15) and median time since injury of 8 years (IQR 14). The presence of cortical lesions was independently associated with greater emotional wellbeing scores (RAND-36 adjusted t score was 4.1; 95% CI: [1.0, 7.0] points higher in participants with versus without lesions; p=0.009). Right temporal and lateral orbitofrontal lesions associated with lower (worse) emotional wellbeing scores, and left hemispheric lesions associated with higher (better) scores. Lower global white matter FA was independently associated with greater emotional wellbeing scores (RAND-36 adjusted t score was 2.7 [0.8, 4.7] points higher per 1 SD decrease in FA; p = 0.008). The strongest associations between individual white matter tract integrity and emotional wellbeing were in the anterior (genu) corpus callosum (β = -3.0, [-1.3, -4.7]; pHolm=0.034) and the left cingulum bundle (β = -3.0, [-1.2, -4.8]; pHolm=0.048). Conclusions In a large cohort of participants with chronic TBI, structural brain injuries were associated with better emotional wellbeing scores. These findings challenge traditional interpretations of emotional disturbances after TBI and motivate future studies to determine whether specific structural brain injury locations may be causally associated with improvements in emotional wellbeing.
Year:
2025
Accumulating evidence of heterogeneous long-term outcomes after traumatic brain injury (TBI) has challenged longstanding approaches to TBI outcome classification that are largely based on global functioning. A lack of studies with clinical and biomarker data from individuals living with chronic (>1 year post-injury) TBI has precluded refinement of long-term outcome classification ontology. Multimodal data in well-characterized TBI cohorts are required to understand the clinical phenotypes and biological underpinnings of persistent symptoms in the chronic phase of TBI. The present cross-sectional study leveraged data from 281 participants with chronic complicated mild-to-severe TBI in the Late Effects of Traumatic Brain Injury Study. Our primary objective was to develop and validate clinical phenotypes using data from 41 TBI measures spanning a comprehensive cognitive battery, motor testing and assessments of mood, health and functioning. We performed a 70/30% split of training (n = 195) and validation (n = 86) datasets and performed principal components analysis to reduce the dimensionality of data. We used Hierarchical Clustering on Principal Components with k-means consolidation to identify clusters, or phenotypes, with shared clinical features. Our secondary objective was to investigate differences in brain volume in seven cortical networks across clinical phenotypes in the subset of 168 participants with brain MRI data. We performed multivariable linear regression models adjusted for age, age-squared, sex, scanner, injury chronicity, injury severity and training/validation set. In the training/validation sets, we observed four phenotypes: (i) mixed cognitive and mood/behavioural deficits (11.8%; 15.1% in the training and validation sets, respectively); (ii) predominant cognitive deficits (20.5%; 23.3%); (iii) predominant mood/behavioural deficits (27.7%; 22.1%); and (iv) few deficits across domains (40%; 39.5%). The predominant cognitive deficit phenotype had lower cortical volumes in executive control, dorsal attention, limbic, default mode and visual networks, relative to the phenotype with few deficits. The predominant mood/behavioural deficit phenotype had lower volumes in dorsal attention, limbic and visual networks, compared to the phenotype with few deficits. Contrary to expectation, we did not detect differences in network-specific volumes between the phenotypes with mixed deficits versus few deficits. We identified four clinical phenotypes and their neuroanatomic correlates in a well-characterized cohort of individuals with chronic TBI. Phenotypes defined by symptom clusters, as opposed to global functioning, could inform clinical trial stratification. Individuals with predominant cognitive and mood/behavioural deficits had reduced cortical volumes in specific cortical networks, providing insights into sensitive, though not specific, candidate imaging biomarkers of clinical symptom phenotypes after chronic TBI and potential targets for intervention.
Year:
2025
Year:
2025
Year:
2025
Year:
2025
Year:
2025
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Collaborators (22)
Patrik Michel
Associate Professor
Lausanne University Hospital
Daniel Perl
Professor
Uniformed Services University of the Health Sciences F Edward Hébert School of Medicine
Aaron Warren
Harvard University
Nicole R Zürcher
Assistant Professor
-
Jonathan Victor
Professor
Weill Cornell Medical College
Maximilian Friedrich
Attending Neurologist | Research Group Leader
University Hospital Ulm
Jacob Hooker
Massachusetts General Hospital
Robert Constable
Professor
Yale University School of Medicine
Satrajit Ghosh
Senior Research Scientist
Massachusetts Institute of Technology
Lawrence Hirsch
Professor
Yale University
Christine L Mac Donald
University of Washington
Henry Tregidgo
University College London
Maarten G. Lansberg
Professor
Stanford University
W. Taylor Kimberly
-
Juan Eugenio Iglesias
Associate Professor
Athinoula A. Martinos Center for Biomedical Imaging
Amber Nolan
Associate Professor
University of Washington
Christos Ganos
Director
-
Christophe Destrieux
Pr
Université de Tours
Yelena Bodien
Assistant Professor
Vanderbilt University Medical Center
Laura Lewis
Professor
Northeastern University
Jeanne M. Hoffman
Professor
University of Washington
David R Schreier
Massachusetts General Hospital
Make your next move.
Prepare your application
Review drafts before you send them.