MRI investigation of orientation-dependent changes in microstructure and function in a mouse model of mild traumatic brain injury

MRI investigation of orientation-dependent changes in microstructure and function in a mouse model of mild traumatic brain injury | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article MRI investigation of orientation-dependent changes in microstructure and function in a mouse model of mild traumatic brain injury Amr Eed, Jake Hamilton, Xiaoyun Xu, Nicole Geremia, Vania F. Prado, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6985478/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 01 Dec, 2025 Read the published version in Acta Neuropathologica Communications → Version 1 posted 12 You are reading this latest preprint version Abstract While neuroimaging studies have revealed notable white matter damage following mild traumatic brain injury (mTBI), the specific tracts and brain regions affected vary widely across studies. Here, we explored whether the spatial orientation of white matter tracts influences susceptibility to mTBI, predicting that tracts oriented orthogonal to the axis of rotation of the head during impact (within the plane of rotation) would exhibit the most damage. Using a model of repeated rotational mTBI in mice, we acquired advanced diffusion MRI (diffusional kurtosis imaging using oscillating gradient encoding) and resting-state functional MRI (fMRI) data at baseline and 1-week post-injury. Consistent with our prediction, while both diffusivity and diffusional kurtosis decreased in the white matter of injured mice, only diffusional kurtosis revealed microstructural changes confined to tracts oriented orthogonal to the right-left axis of rotation. In addition, both region and subregion analyses showed FC deficits between regions connected via tracts running orthogonal to the rotation axis. The orientation-dependent changes in imaging metrics were validated by histopathological analyses. Females showed greater microstructural changes than males using dMRI following injury, while no sex differences were detected by fMRI. Interestingly, the region-specific and subregion-specific FC analyses showed overlapping but non-identical changes in FC suggesting the utility of using both coarse and fine levels of brain parcellation for FC analyses in mTBI. These findings suggest that mTBI imaging studies may benefit from the consideration that damage after mTBI will predominate in tracts that are oriented orthogonal to the axis of rotation produced by the impact and that diffusivity and diffusional kurtosis as well as region and subregion-specific fMRI analyses can detect these changes. Diffusion MRI Functional MRI Mild traumatic brain injury Concussion Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 1. Introduction Traumatic brain injury (TBI) is the leading cause of death and disability among all trauma-related injuries with an estimated annual incidence of 50–60 million [ 55 , 74 ]. Approximately 90% of TBIs are classified as mild TBI (mTBI) or concussion [ 12 ]. While mTBI typically results from both linear and rotational accelerations of the brain, rotational acceleration has been identified as particularly significant due to its ability to generate shear and strain forces on brain tissue [ 12 , 62 ]. White matter (WM) tracts are particularly susceptible to these rotational forces due to their highly anisotropic structure, explaining why diffuse axonal injury is a hallmark of concussion [ 6 , 26 ]. The magnitude and orientation of rotational forces produced during an impact cause distinct strain dynamics throughout the brain in mTBI. Studies modeling brain injuries have shown that it is not the magnitude of strain, but the component of strain oriented along axons (tract-oriented strain) that drives axonal damage [ 15 , 35 , 86 ]. Support for the clinical relevance of these findings is provided by Knutsen et al.[ 47 ] who used non-injurious head accelerations in human volunteers, to show that strain components orthogonal to the axis of rotation account for the majority of brain deformation, with negligible strain along the axis of rotation. While histological studies have supported this tract orientation-dependence of damage [ 16 , 27 ], there remains a notable gap in research investigating the relationship between the orientation of rotational forces produced by an mTBI and the subsequent distribution of pathological changes in vivo . Diffusion tensor imaging (DTI), a commonly used diffusion MRI (dMRI) technique, has been used extensively in attempts to identify pathological changes in WM tracts following mTBI in humans and animals [ 41 , 53 ]. These studies failed to reveal consistent results in terms of which WM tracts show dMRI changes after injury. For example, some studies have reported dMRI changes in the corpus callosum [ 56 , 87 , 89 ], while others highlight changes in the superior longitudinal fasciculus[ 38 , 57 ] and internal capsule [ 21 , 38 ]. It is likely that that these discrepancies arise from the heterogeneity in the orientation of rotational forces being studied. Thus, dMRI studies examining mTBI might be improved by analyzing the data in the context of the axis of rotation of a given injury. Additionally, DTI is limited by its assumption of Gaussian diffusion in complex microenvironments where changes such as axonal damage, myelin damage, and inflammatory processes may be occurring simultaneously as in the case of mTBI [ 41 , 42 ]. Diffusional kurtosis imaging (DKI) allows better characterization of complex microenvironments by capturing non-Gaussian diffusion characteristics to provide a measure of tissue complexity[ 44 , 45 ] and may thus improve detection of microstructural changes following mTBI. Furthermore, frequency-dependent dMRI using oscillating gradient encoding allows characterization of microstructural changes at different spatial scales [ 77 ], giving insight into the size of structures contributing to dMRI changes. While frequency-dependent dMRI has been applied in various pathologies [ 3 , 5 , 9 , 97 ], its application to mTBI remains unexplored. While dMRI provides insight into WM microstructural changes brought on by mTBI, resting-state fMRI (rs-fMRI) has the potential to reveal the functional consequences of WM injury. rs-fMRI has been extensively used to study mTBI in both humans [ 23 , 59 ] and animals [ 19 , 76 , 80 , 89 , 93 , 99 ]. Similar to DTI studies, these rs-fMRI studies report a range of findings including increases and decreases in functional connectivity in different brain regions including: the somatosensory and somatomotor cortex [ 37 , 40 , 89 ], the thalamus and the hippocampus [ 19 , 37 , 58 , 64 , 71 , 76 , 89 , 93 ]. As in the case of the discrepancies in the results of dMRI studies in mTBI, we predict that the lack of agreement in rs-fMRI studies in mTBI lies in the failure to analyze the rs-fMRI data to focus on changes in connectivity between brain regions connected by tracts that are oriented orthogonal to the axis of rotation produced by the injury. We have recently reported on a novel method to generate mTBIs with high rotational forces to study concussion in mice [ 88 ]. In the present study, we predicted that pathological changes in the WM tracts would be more easily detected by DKI compared to DTI metrics, that WM tracts oriented orthogonal to the axis of rotation produced by this mTBI would be preferentially injured, and that brain areas connected by tracts running orthogonal to the injury would show functional disconnection by rs-fMRI. We herein report that this model of mTBI produces the greatest pathological changes in tracts that are oriented orthogonal to the axis of rotation of the mTBI. Furthermore, this orientation-dependence of pathological changes was evident in DKI but not DTI metrics. In agreement with the DKI results the rs-fMRI shows that the mTBI results in a loss of functional connectivity between structures oriented in the anterior-posterior axis. 2. Materials and methods Subjects All animal procedures were approved by the Western University Animal Care Committee and were consistent with guidelines established by the Canadian Council on Animal Care. In vivo MRI data was collected from genetically modified mice homozygous for the APOE3 allele carrying humanized wildtype MAPT [ 75 ] and APP [ 8 ] (JAX stock #030898) under the control of the associated mouse promoter. Animals aged 6 months were randomly assigned to either sham (n = 15, 7 males 8 females) or mTBI groups (n = 16, 8 males and females). Repetitive mTBI Procedure The injury procedure has been reported in detail in Teasell et al. [ 88 ]. Briefly, animals were anesthetized by intra-peritoneum injection with Ketamine 80mg/kg and Xylazine 10mg/kg; then placed on a custom-made acrylic box, topped with a piece of pre-pierced clear plastic and positioned under a traumatic brain injury device (TBI 0310, Precision Systems and Instrumentation, LLC). The custom-made, 5mm-diameter, pliant, silicone tip was placed aligning with the bregma. The device was programmed to deliver an impact at an intended depth 8.0 mm at a 3.5 m/s velocity with a 500-millisecond dwell time. Following the impact, the mouse broke through the plastic and underwent a 180° sagittal rotation (rotation about the right-left axis) and landed on the cushioned bottom of the box. The animal was then transferred back to its cage and placed under a heating lamp until regaining consciousness. The sham animals received the corresponding doses of ketamine/xylazine, but without the impact. These procedures were repeated 3 times, with a 24-hour interval. All experimental animals regained consciousness within 5–10 minutes post corresponding procedures, with no obvious motor deficit. No skull fracture, hematoma, apnea or death associated with sham or mTBI mice was observed. MRI Acquisition Before scanning sessions, anesthesia was induced by placing mice in an induction chamber with 4% isoflurane in 100% oxygen with a flow rate of 1 L/min. Throughout the scanning session, isoflurane was maintained at 1.8% in 100% oxygen with a flow rate of 1 L/min through a custom-built nose cone. In vivo MR scanning sessions were performed on a 9.4 T Bruker Neo small animal scanner (Agilent, Palo Alto, CA, USA) equipped with a 6 cm gradient coil insert of 1 T/m strength, Bruker Avance III HD console with software package of Paravision-360.3.3 (Bruker BioSpin Corp, Billerica, MA), and a single-loop surface coil of 2x1 cm 2 . All subjects were scanned at baseline at 6 months of age and then again 1-week post-last impact. Anatomical, diffusion, and functional data were acquired for all scanning sessions in a scan time of 2 hours 12 minutes. The dMRI protocol included a pulsed gradient spin echo (PGSE) sequence (i.e., 0 Hz) with gradient duration of 9.4 ms (diffusion time = 12.3 ms) and oscillating gradient (OGSE) sequences with frequencies of 60 and 120 Hz (corresponding effective diffusion times of 2.3 and 1.2 ms [ 67 ]). The 0, 60, and 120 Hz acquisitions correspond to average molecular displacements of 8.6, 3.7, and 2.7 µm, respectively, according to the Einstein-Smoluchowski relation for hindered diffusion [ 65 ]. The 60 Hz acquisition implements frequency-tuned bipolar waveforms to reduce the TE of the acquisition [ 13 ]. For all frequencies, data consisted of 2 b = 0 volumes and 2 b-value shells of 1,000 and 2,500 s/mm 2 each with an efficient 10-direction scheme [ 36 ]. The dMRI data was acquired in one integrated scan using single-shot echo planar imaging (EPI) with 80% of k-space being sampled in the phase encode direction and parameters: TE/TR = 35.5/15000 ms, FOV = 19.2 x 14.4 x 15 mm 3 , in-plane resolution 200 x 200 µm 2 , slice thickness 500 µm, 4 repetitions, scan time of 66 minutes. T2*-weighted images for fMRI were acquired in 30 coronal slices using a gradient-echo EPI (GE-EPI) sequence covering the whole brain with the following parameters: TE/TR of 12/1500 ms, flip angle of 60°, FOV of 19.2 × 9.6 mm 2 , matrix size (MS) of 64 x 32, and slice thickness of 0.5 mm to produce a voxel size of 300 × 300 × 500 µm 3 . Four runs of 600 volumes were acquired consecutively for a total of 60 minutes. Two runs were acquired with a blip up and the other two were acquired with blip down to allow for distortion correction (details below). An additional T2-weighted anatomical image was acquired in the same space using a turbo rapid acquisition with relaxation enhancement (TurboRARE) sequence with the following parameters: TE/TR of 30/5500 ms, 8 averages, FOV of 19.2 × 9.6 mm 2 , MS of 128 × 64, a slice thickness of 0.5 mm to yield a voxel size of 150 × 150 × 500 µm 3 . The acquisition duration was 352 seconds. Diffusion MRI Preprocessing Complex-valued repetitions underwent partial Fourier reconstruction using POCS [ 34 ], frequency and signal drift correction, and phase alignment, similar to previous work [ 73 ]. The DESIGNER pipeline [ 17 ] was used to perform MP-PCA tensor denoising [ 25 , 69 , 91 ] with Rician bias correction [ 48 ], followed by Gibbs ringing correction for partial Fourier acquisitions [ 46 , 51 ]. Fitting Diffusivity (MD: mean diffusivity, RD: radial diffusivity, AD: axial diffusivity) and kurtosis (MKT: mean kurtosis tensor, RK: radial kurtosis, AK: axial kurtosis) maps were computed at each frequency using axisymmetric modelling and spatial regularization, as outlined in Hamilton et al. [ 36 ]. Regularization weighting was heuristically chosen to minimize the visual appearance of noise while retaining original image contrast and was held constant across subjects. Quality Assurance The signal-to-noise floor ratio (SNR) of all scans was measured as the signal mean in an ROI placed in the cortex across b = 2,500 s/mm 2 acquisitions before denoising, divided by the signal mean in an ROI outside the brain. One male in the mTBI cohort and one female in the sham cohort did not pass our dMRI quality assurance threshold due to low SNR (< 2.5, 4.5 median for all subjects) of baseline scans and qualitative noise contamination in diffusion-weighted volumes and parameter maps and thus were excluded from subsequent dMRI analysis. Tractography For all subjects baseline scans, deterministic whole brain tractography was performed using ExploreDTI [ 52 ] with parameters: Fiber length range 3–12 mm, angle threshold of 50 degrees, step size 0.1 mm, fractional anisotropy (FA) threshold 0.15. Data from all PGSE and OGSE acquisitions was used as an input for tractography to provide a more robust result as the principal eigenvector should not vary appreciably across the explored frequencies [ 36 ]. For each subject, the post-injury scan was registered to baseline scan using affine and symmetric diffeomorphic transforms with ANTs software [ 7 ]. This was done as opposed to performing tractography for each imaging session to reduce noise and ensure dMRI metric changes from possible tract degeneration were captured. In each tract, the mean principal eigenvector was used to separate tracts into groups according to their orientation relative to the axis of rotation (right-left axis). The central angle between each tract and the rotation axis was determined by computing the dot product of the mean principal eigenvector with the right-left axis. Tract groups of 0–50, 50–60, 60–70, 70–80, and 80–90º were chosen based on the spherical sector occupied by each bin and to keep the number of tracts in each group approximately equal. For each tract group, a tract density map was computed which indicated voxels with tracts passing through them. This tract density map was binarized and used to compute the mean of dMRI metrics at each frequency within tract bins for each subject and timepoint. Voxels with MD > 1.4 µm 2 /ms were excluded from binarized tract density maps to mitigate partial volume effects with cerebrospinal fluid. Functional MRI Preprocessing The analysis was done using multiple tools from FSL 6.0.7 [ 43 ], AFNI [ 18 ], ANTs [ 7 ], ITKSnap [ 100 ], and nilearn ( https://github.com/nilearn/nilearn ). All the processing workflows were implemented using the Nipype Python library [ 28 ]. The T2-weighted images were used to construct a study-based template and the skull was removed from the template manually using ITKSnap. The anatomical images from each subject were bias field corrected, registered to skull-stripped template and the inverse transformations applied to the template mask were used to automatically remove the skull from individual subjects. The extracted brains were then registered to study-based template brain. The functional images were corrected for motion using rigid-body registration of 6 degrees of freedom. The images were corrected for field distortions using FSL topup toolbox [ 82 ] taking advantage of the opposite phase-encoding acquisition. The middle volume of each subject was registered using rigid body transformations to the study-template brain. The inverse transformations applied to the template mask and the resulting masks were used to extract the brain from individual runs. Functional to anatomical co-registration transformations were calculated using affine registration. The functional images were then high-pass filtered using a 0.01 Hz to remove the low-frequency noise and to keep the signal associated with resting-state networks (> 0.01 Hz) and motion parameters were regressed out of the data. The cleaned images were then smoothed using an isotropic Gaussian kernel of 0.4 mm 3 . The transformations from individual anatomical images to the study-based template and from the functional to the anatomical images were combined and applied to the smoothed functional images to bring them into the study-based template space. Images in study-based template were used for subsequent connectivity analyses. The study-based template was nonlinearly registered to the anatomical image of the Allen Brain Atlas common coordinate framework version 3 (ABA-CCFv3) [ 94 ]. The inverse transformations were used to bring the annotation images from Allen Brain Atlas to the study-based template for functional connectivity (FC) atlas-based analysis. Annotation atlases creation ABA-CCFv3 annotation atlas was used to construct: a region atlas of 22 ROIs (Fig. 4a) (11 per hemisphere) and a more refined subregions atlas with 82 ROIS (SI Appendix, Fig. S4a) (41 per hemisphere) by combining adjacent structures. Atlas-based FC For each preprocessed run, the average timeseries within each label of the annotation atlas was computed, Pearson’s correlation coefficient between each pair of labels was calculated using nilearn and later converted to z -score using Fisher’s transformation. The resulting correlation matrices were averaged across runs and later used for statistical inferences. Independent component analysis (ICA) For group-level ICA analysis, the preprocessed 4D images of all subjects were concatenated and group ICA analysis was run using MELODIC tool in FSL [ 10 ]. The concatenated data was decomposed into 10, 15, 20, 25, 30, 40, and 50 dimensions. The 20 dimensions decomposition gave the best representation of the resting-state networks [ 29 , 101 ]. We excluded 1 component that overlapped with major ventricles and the remaining 19 components were used in further analysis. We then ran the Dual Regression (DR) analysis and used the 2nd stage maps in assessing the within-component connectivity [ 66 ]. To assess between-component connectivity strengths, the ICA components were thresholded by 3 and used as a probabilistic atlas. For visualization purposes, the study-based template was non-linearly registered to the ABA-CCFv3 anatomical atlas and the transformations were used to bring the ICA networks and DR statistical maps into the atlas space. Tract-based FC Viral tracer data from Allen Brain’s Institute Mouse connectivity data [ 68 ] was used to extract tracts identified in tracts orientation bins from dMRI analysis. Each experiment was chosen such that the tract has the highest injection volume among the white matter fiber tracts. The data for each experiment was downloaded, converted to NIfTI format, and the log10 transformed data was thresholded to 10 − 3.5 for better false positive control [ 68 ]. The source and target structures were determined based on structures with the highest injection volumes known to be interconnected by that tract (SI Appendix, Table S1 ). For instance, the corticospinal tract (CST) is known to connect the motor cortex to more posterior regions and finally to the spinal cord. A representative experiment was chosen with the primary motor cortex as the primary injection location and the pons was chosen as a target as it has the highest injection volume among structures known to be connected to the motor cortex through the CST. ABA CCFv3 annotation atlas was used to obtain binary masks of the source and target regions and these masks were multiplied by the binarized injection volume map to obtain regions between which correlation strength can be calculated. The timeseries within the source and target regions were averaged and correlation coefficient was calculated and transformed to Fisher’s z-score using AFNI’s 3dNetCorr. For targets that are ipsilateral to the injection volume, the source and target on both hemispheres were averaged, while for those tracts that traverse the brain midline only lateral regions were used. Histology Immunohistochemistry and Silver Staining At one week after the last impact, 12 mice (6 injured and 6 shams) were anesthetized with ketamine/Xylazine (2:1), then underwent trans-cardiac perfusion with ice-old saline, followed by 4% Paraformaldehyde in PBS. Brains were post-fixed overnight in 4% paraformaldehyde, then placed in 15% sucrose (Cat# S5-3, Fisher Scientific) in PBS for 6–12 hours and then 30% sucrose in PBS overnight and embedded in optimal cutting temperature (OCT) medium. For silver staining, fifteen floating coronal cryostat sections (50 µm) were collected at the level of the corpus callosum (bregma + 1/-1mm) and optic tracts (bregma − 1/-2mm), respectively. The staining was performed using the FD NeuroSilver Kit II (FD NeuroTechnologies, Ellicott City, MD) according to the manufacturer’s instructions. For immunohistochemistry, the rest of the brain was cryosectioned at 16 µm and collected serially onto Superfrost Plus slides. Cryosections were rinsed in 0.1M phosphate-buffer saline and blocked in 5% goat serum with 0.1% Triton X-100 for 2h at room temperature. Then the sections were incubated with primary antibodies: anti-GFAP antibody (Cat# G3893; Sigma, Germany) and anti-Iba1 antibody (Cat# 019-19741, WAKO Japan) at 4°C overnight, followed by 1h incubation with fluorescent secondary antibodies: donkey anti-mouse IgG, conjugated with Alexa Fluor 488 (1:1000; Invitrogen, Rockford IL, USA) and goat anti-rabbit IgG, conjugated with Alexa Fluro 594 (1:1000; Invitrogen). Digital images were captured using a Leica MICA or Leica Stellaris 5 confocal microscope. Electron microscopy Four injured mice and 2 sham mice were processed for EM, as described previously [ 98 ]. Briefly, mice were perfused with 0.1M phosphate buffer, followed by 4% paraformaldehyde supplemented with 2% glutaraldehyde. Vibratome sections (100um) were collected that included corpus callosum regions and internal capsule regions, respectively (ROIs). The specimens were then post-fixated with 1% Osmium Tetroxide in 0.1M Cacodylate buffer for 1 hr followed by overnight en-bloc staining with 1% Uranyl Acetate (at T = 4 ⁰C). Brain tissues were first rinsed with double-distilled water and then dehydrated in an ascending series of ethanol solutions and embedded in Spurr’s resin between two Aclar films at 60°C for 2–3 days for polymerization. After polymerization, ROIs were identified and isolated using a stereomicroscope [ 54 ]. A Leica UCT ultramicrotome with a Diatome diamond knife was used to obtain ultrathin serial sections (60–70 nm) from the ROIs. Serial sections were collected onto pioloform-coated copper slot grids. Electron microscopy examination was carried using the TEM at the Biotron facility at Western University. Statistics Diffusion MRI Sex differences were evaluated within sham and mTBI groups using one-way ANOVA models on metric changes post-injury relative to baseline for each group across each metric and OGSE frequency. After separating data by sex, a one-way ANOVA was used to examine group effects (sham vs. mTBI) for each metric and frequency. To control for multiple comparisons, false discovery rate (FDR) correction using the Benjamini-Hochberg procedure was applied to p-values obtained from ANOVA analyses of sex and injury effects. For metrics showing significant effects of injury, Sidak post hoc tests were conducted to identify specific differences between mTBI and sham cohorts within each tract bin. All statistical analysis was performed using R Statistical Software version 4.1.2 and GraphPad Prism version 10.2.0. Functional MRI For atlas-based and ICA-based FC statistical analysis, the average connectivity matrix from each subject was combined into a 4D matrix for each group and various comparisons were conducted using permutation testing as implemented in FSL permutation analysis of linear models (PALM) tool [ 95 ]. Sex differences were tested using unpaired permutation testing conducted separately within each group and condition (males vs females before and after in mTBI and sham groups). Left vs right hemispheres were compared using paired permutation testing within each group and condition. Similarly, changes before and after the injury or the sham procedure were tested using within each group and condition. For all the permutation tests, null distributions were generated using 10,000 permutations and exchangeability blocks were defined at the subject-level to account for the paired nature of the data in cases of paired comparisons. All results were corrected for multiple comparisons using FDR. The voxel-wise statistical analysis for the 2nd stage maps of the DR were compared using paired permutation testing with 10,000 permutations, threshold-free cluster enhancement (TFCE), and family-wise error rate (FWER) correction across voxels, components, and contrasts. For tractography-based FC, the connectivity strength between the source and target structures were compared using paired Student’s t-test for each tract separately. 3. Results Representative Maps To enable analyses of dMRI and rs-fMRI data with respect to tract orientation we began by validating representative structural, functional, and diffusion MR images from mice at baseline (Fig. 1a-c). MD and MKT both show qualitative frequency-dependence with increased diffusivity and decreased kurtosis, as expected [ 5 , 36 ]. To examine the effect of tract orientation relative to the axis of rotation produced by the injury, tracts generated from tractography were separated into 5 bins based on their orientation relative the right-left axis of rotation (0º-50º, 50º-60º, 60º-70º, 70º-80º, 80º-90º) (Fig. 1d). In each bin, tract density maps were computed and used to evaluate dMRI metric changes post-injury (Fig. 1e). Sex differences in mTBI mice detected by dMRI Mice in the mTBI group (n = 8 males and females) underwent 3 mTBIs (one a day for 3 days) as described [ 88 ]. Mice in the sham group (n = 7 males, 8 females) underwent 3 sham procedures (anesthesia without injury). Mice were imaged at baseline and 1-week post-injury and dMRI metrics computed for each tract bin relative to the axis of rotation as delineated in Fig. 1. Data from 1 male mTBI and 1 female sham mouse were excluded from further dMRI analysis due to low quality data which did not pass our quality assurance threshold (detailed in Methods). To detect possible sex differences, all dMRI metrics were compared between male and female mTBI mice and between male and female shams. In mTBI mice a significant effect of sex was detected for MD, RD, AD, and MKT at all frequencies, and RK at 60 and 120 Hz ( p < 0.05) (Fig. 2). Sex differences in metric changes were relatively consistent across acquisition frequencies and tract orientations, with mTBI females having larger decreases in diffusivity and kurtosis metrics. In the sham mice, a significant effect of sex was detected for MKT and AK at 60 and 120 Hz ( p < 0.05), with males showing increased kurtosis relative to females following the sham procedure (Fig. 2 insets). Diffusional kurtosis detects tract orientation dependence of damage following mTBI in females Given the sex differences detected, we analyzed males and females separately in both sham and mTBI groups to evaluate dMRI changes due to injury. For kurtosis metrics (Fig. 3a) in females, a significant group effect (injured versus sham) was found at 60 Hz for MKT ( F = 6.145, p < 0.05) and RK ( F = 7.896, p < 0.05). Importantly, MKT and RK reveal a tract orientation dependence with larger decreases in the mTBI mice in tracts oriented orthogonal to the left-to-right axis of rotation generated by the injury, with significant post-hoc findings in the 80–90º tract bin for MKT and RK ( p < 0.05) at 60 Hz. Diffusivity metrics (Fig. 3b) showed significant group effects in MD and RD at all frequencies and AD at 0 Hz ( p < 0.05). However, diffusivity metrics did not show the same tract orientation dependence as the kurtosis metrics, with larger diffusivity decreases in the mTBI cohort being consistent across all tract orientations. No significant effects of group were found for any dMRI metric in males (SI Appendix, Fig. S1 ). Parcellating the brain into regions and subregions annotation atlases To minimize registration errors, reduce artifact due to motion, and reduce the number of statistical comparisons required, the ABA-CCFv3 annotation atlas was down-sampled to create a region-specific atlas (Fig. 4a) consisting of 22 anatomical regions (11 per hemisphere). A finer subregion-specific atlas (SI Appendix, Fig. S4a) consisting of 82 subregions (41 per hemisphere) was generated to enable a more detailed FC analysis. The region-specific atlas was created by combining together adjacent structures. For example, the gustatory, the anterior cingulate, and the somatomotor areas of the isocortex (to name a few subregions) were considered a single structure in the region-specific atlas and designated collectively as the isocortex. In the creation of the subregion-specific atlas, we preserved one level of hierarchy such that the isocortex was divided into its component parts. No sex differences in mTBI mice detected by fMRI To test for sex differences in rs-fMRI, we compared the FC between males and females in sham and mTBI groups at baseline and again post-injury or post sham procedure. FC comparisons between males and females failed to reveal any differences that survived multiple comparisons correction for the region-specific or the subregion-specific atlases (SI Appendix, Fig. S2 & S3). Thus, the FC analyses were performed on 4 groups of mice with sexes combined: sham at baseline, shams at 1 week post sham procedure, mTBI mice at baseline and mTBI mice 1 week post-injury. mTBI induces wide decrease in FC across the brain The rs-fMRI data was analyzed to uncover differences in FC within the sham group (comparing FC at baseline to FC at 1 week post-sham procedure), within the mTBI group (comparing FC at baseline to FC at 1 week post-mTBI) and between the sham and mTBI groups (comparing FC at baseline to FC at 1 week post-sham procedure or mTBI). After correcting for multiple comparisons, statistically significant reductions in FC were observed in the mTBI group 1 week after the injury compared to FC in both sham groups and to the FC in the mTBI baseline group. As expected, there were no statistically significant differences in FC between the sham groups at baseline and at 1 week post-sham procedure and between the shams at baseline and the mTBI mice at baseline (Fig. 4b and 4c and SI Appendix, Sig. S4a and 4b). Region-specific changes in FC One week after the injury, the mTBI group showed a widespread decrease in FC compared to their baseline values (Fig. 4d). The largest reductions in FC were between regions in the frontal, temporal and posterior regions of the brain that would rely on long-range connections that run from frontal and temporal regions to the posterior regions (Fig. 4e). The FC between anterior regions such as the olfactory region and posterior regions (the hindbrain and cerebellum) showed the most significant decreases in FC (Fig. 4e) with a more than 50% decrease in FC. FC between the isocortex and more posterior regions (the midbrain, hindbrain, and cerebellum) were also decreased after mTBI (Fig. 4d). Temporal regions such as the hippocampal formation showed decreases in FC with the isocortex, hindbrain and cerebellum. Subregion-specific changes in FC FC analyses were also carried out using the subregion-specific atlas to evaluate if region-specific FC changes could be attributed to FC changes between component subregions and to determine if a finer FC subregions analysis could uncover changes in FC hidden by the courser region-specific analyses. This subregion-specific analysis revealed that the decrease in FC between the isocortex and more posterior regions could be specifically attributed to a decrease in FC between the frontal pole, the somatosensory/somatomotor cortex, the anterior cingulate cortex, the agranular insular cortex, and temporal association areas with regions of the midbrain, hindbrain and cerebellum (SI Appendix, Fig. S4). The subregion analyses further identified that decreases in the FC of the Ammon’s horn (CA) and subiculum subregions of the hippocampus accounted for the decreased FC between the hippocampal formation and hindbrain/cerebellum shown in the region-specific analysis (SI Appendix, Fig. S4). In some cases, the region-specific analyses detected changes in FC that were not found between the component subregions. For example FC changes between left olfactory areas and left cerebellum, right isocortex and left midbrain, and right hippocampus and left midbrain were not present in the subregion-specific analyses. On the other hand the FC using the subregion-specific analysis revealed a significant decrease in FC between the amygdala and the vermal regions of the cerebellum that was not appreciated in the FC using the region-specific atlas in which all cerebellum substructures were aggregated into one cerebellar region (Fig. 4a). Statistically significant changes in FC between subregions are represented by connecting lines in the circular plot in SI Appendix, Fig. S4e with the thickness of the lines reflecting the percentage of decrease in FC after the mTBI. FC changes after mTBI show right-left asymmetries Comparing changes in FC in mTBI mice 1 week after injury to their baseline values revealed right-left asymmetries that were more evident in the region-specific atlas than in the subregion-specific atlas (Fig. 4e and SI Appendix, Fig. S4e). The right hemisphere showed decreases in FC between the isocortex and its targets in the hippocampus, hindbrain and cerebellum and between the hippocampus and its targets in the isocortex, hindbrain and cerebellum. In the left hemisphere, only left olfactory regions showed decreases in FC with the hindbrain, and the cerebellum. Analyses of FC in the right hemisphere of sham mice also identified decreases in FC between the hippocampus and the isocortex, indicating that these FC asymmetries were reflective of normal FC asymmetries in mice and are unrelated to the mTBI (SI Appendix, Fig. S5 & S6). Identification of resting-state networks The data from all subjects in both groups was aggregated together and decomposed into 20 resting-state networks (RSN), one component was disregarded as noise due to peak activation overlaps with major ventricles. The most common networks previously reported in the literature were identified in our data (SI Appendix, Fig. S7). We identified cortical networks representing the olfactory areas, somatomotor regions, anterior regions of the default mode network (DMN) including the prelimbic areas, the anterior cingulate area, and the orbital area. Subcortical networks were also identified such as the amygdala, the hippocampus, the striatum and the hypothalamus. Some networks showed laterality such as somatomotor, amygdala, striatum, hypothalamus, midbrain, and hindbrain networks. mTBI causes FC changes within resting-state networks DR analysis showed two affected networks with diminished connectivity in the mTBI group after the injury. The first was the network consisting of regions of the ventral striatum, pallidum, and hypothalamus that showed clusters of decreased FC in the taenia tecta region of the olfactory areas (Fig. 5d, top). The second network affected by mTBI was the right somatomotor network that showed clusters of lower FC in the caudate putamen region of the striatum (Fig. 5d, bottom). Three resting-state networks showed an increase in FC in the control group following the sham procedure including, the hypothalamus network, the right amygdala network, and the right somatomotor network (SI Appendix, Fig. S8) that showed decreased FC in the mTBI group. mTBI causes FC changes between resting-state networks Top-bottom correlation between the somatomotor network and more ventral regions in the olfactory and ventral striatum networks and between the vermal region of the cerebellar network and the midbrain network showed significant decreases in FC following mTBI (Fig. 5c). In comparison to the sham group (Fig. 5a), FC between the hippocampal network and midbrain and the cerebellar networks showed hypoconnectivity following mTBI (Fig. 5b and c). Decreases in FC between regions directly connected by WM tracts orthogonal to right-left axis of rotation To further test our prediction that brain areas connected by WM tracts that are oriented orthogonal to the right-left axis of rotation of the mTBI, we calculated the FC between regions that are known to be anatomically connected through WM tracts that run in the rostral-caudal direction as determined by viral tracing studies [ 68 ]. In contrast to atlas-based or network-based FC which cannot differentiate between direct and indirect connections, this approach guarantees that the FC of the regions analyzed are directly, anatomically connected. The FC served by the CST was the most-affected tract among all those tested including the cingulum bundle, the optic tract, and the cerebellar peduncle (SI Appendix, Fig. S9). FC between the motor cortex and the hindbrain that is served by the CST was decreased by 17% following mTBI (Fig. 6a and c). In contrast, the FC of regions connected by fibers that run parallel to the right-left axis including the genu of the corpus callosum (Fig. 6b and d), the stria terminalis, and the arbor vitae (SI Appendix, Fig. S10) were unchanged. Histological Findings The imaging studies above suggested that pathological changes due to mTBI predominate in WM tracts oriented orthogonal to the right-left axis of rotation of the injury. To validate this finding we conducted silver staining, immunohistochemistry, and electron microscopy in sham and mTBI mice in WM tracts that oriented 0º-50º (the body of the corpus callosum) and 80º-90º (the cingulum bundle and the optic tract for silver staining and immunohistochemistry, internal capsule for electron microscopy) to the axis of rotation produced by the mTBI (Fig. 7). Silver staining was used to assess axonal pathology at 1 week post mTBI (Fig. 7a-f). Very few silver-stained fibers were observed in any WM tracts in any of the 6 sham mice. In mTBI mice at 1 week post-injury, only a few silver-stained fibers were noted in the body of the corpus callosum. However, WM tracts that run orthogonal to the right-left axis of rotation produced by the mTBI, such as the sagittal fibers of the cingulum bundle and of the optic tract, showed prominent and widespread silver-stained fibers that were easily detected in all mTBI brains. Immunohistochemistry was used to evaluate these same WM tracts for evidence of localized inflammation and glial reactivity. Representative immunostaining to identify astrocytes (using GFAP immunoreactivity) and microglia (using Iba-1 immunoreactivity) is shown in Fig. 7g-l. As expected, shams demonstrated some GFAP + astrocytes in all tracts with a small number of Iba1 + microglia. The microglial morphology in shams was predominantly that of resting microglia with many ramified branched processes. In the mTBI mice GFAP and Iba1 staining was more prevalent and the microglial morphology was more rounded with fewer processes suggesting a more activated phenotype. Electron microscopic examination of several brain regions identified based on abnormal silver staining revealed extensive ultrastructure changes in the corpus callosum and internal capsule in the mouse brains with mTBI. Multiple stages of degenerating axons were observed (Fig. 7n): axon demyelination; excessive myelination and axon-myelin dissociation (vacuoles formed between myelin sheath and axon). The dystrophic axons were scattered among many normal-appearing axons. These abnormalities were absent in sham samples (Fig. 7m). An example of a reactive astrocyte is shown in Fig. 7o that demonstrates an astrocyte protrusion was adjacent to a degenerative neurite that contained cytoskeletal fragments, a typical glial response to axonal injury. 4. Discussion In this work we investigated the use of dMRI and fMRI to detect microstructural and FC changes in mice 1 week following repeated mTBI. Specifically, we evaluated how the orientation of the axis of rotation produced by the mTBI influences the ability of dMRI and fMRI to detect pathological changes in the brain. Several findings were noted and are discussed below including: (1) that sex differences following mTBI in mice could be detected by dMRI but not by fMRI, (2) that while both diffusivity and kurtosis metrics could distinguish mTBI mice from shams, kurtosis was more sensitive to tract-specific injury and (3) that, after mTBI, diffusional kurtosis and fMRI detect pathological microstructural changes and associated reduced functional connectivity between brain regions connected by tracts oriented orthogonal to the axis of injury produced by the mTBI. Sex differences following mTBI detected with dMRI but not fMRI Sex differences in dMRI metric changes post-injury relative to baseline were assessed in sham and mTBI groups. In sham mice, sex differences were only detected in MKT and AK at 60 and 120 Hz, with no changes in diffusivity metrics. However, in the mTBI mice, diffusivity and kurtosis metrics in females were significantly reduced compared to males. Compared to male mTBI mice, female mTBI mice exhibited greater reductions in MD, RD, AD, and MKT across all acquisition frequencies and RK at 60 and 120 Hz. These dMRI findings are consistent with a greater vulnerability to mTBI-induced damage in female mice. In human dMRI studies of mTBI, sex differences are inconsistent across studies, with some studies reporting greater changes in males [ 24 , 96 ] and others in females [ 83 ]. These discrepancies may be influenced by variability in injury severity between the sexes in these studies. Similarly, preclinical animal studies yield mixed results [ 33 ], although histological studies often report greater WM damage in females relative to males [ 84 , 90 ]. A proposed biological mechanism for these sex-dependent differences suggests that females possess a higher proportion of small caliber axons that are more vulnerable to rotational injury [ 84 ], due to a sparser microtubule network [ 22 ]. Failure to detect sex-differences using rs-fMRI may be due to insufficient power to detect FC differences between mTBI males and females or due to less of a decrease in FC in mTBI females than their WM injury suggests. Compensatory changes in axonal firing rates, synaptic plasticity or other processes commonly seen in concussion [ 14 , 57 ] could also reduce the sex-dependent impact of the microstructural injury on FC. It is also possible that FC is driven by the larger caliber axons in nerve tracts, reducing its sensitivity to damage in the higher fraction of smaller caliber axons found in females. While the correspondence between structural and functional connectivity is broadly accepted [ 20 , 30 , 81 ], rs-fMRI is more vulnerable to motion artifacts and FC can be contaminated by spurious correlations arising from vascular and respiratory artifacts [ 39 ] that can obscure subtle sex differences. Additionally, sensitivity to injury-induced changes in the vascular tree may have obscured sex-specific connectivity changes detectable by fMRI that would not have interfered with the dMRI analyses. Tract orientation dependence of damage detected with diffusional kurtosis The mTBI females showed a significant effect of injury for MD and RD at all frequencies, AD at 0 Hz, and MKT and RK at 60 Hz. The reductions in MKT and RK showed a tract orientation dependence with larger decreases found in tracts oriented more orthogonal to the rotation axis, with both MKT and RK showing significant decreases following mTBI at 60 Hz in the 80–90º bin. By comparison the reductions in diffusivity metrics did not exhibit the same orientation dependence as MKT and RK. Our histological findings show widespread mild inflammation and diffuse axonal injury in tracts oriented orthogonal to the rotation axis, with lesser amounts in tracts parallel to this axis, in addition to myelin disruption in mTBI mice. We propose that mTBI-induces tract-specific damage that is largely restricted to tracts oriented orthogonal to the axis of rotation of the injury and an inflammatory response characterized by swelling and gliosis that is more widespread. We hypothesize that the decreases in diffusivity metrics after mTBI are primarily driven by inflammatory changes that would be expected to be less spatially restricted than tract-specific damage and decrease diffusivity [ 42 ]. This would explain the finding of diffusivity reductions across all tract orientations after mTBI. In contrast, changes in kurtosis metrics may be driven primarily by axon-specific microstructural damage explaining why decreases in kurtosis are restricted to tracts in the 80–90º bin. The assertion that changes in kurtosis may be due to axonal alterations after mTBI is supported by the finding that increased axon diameter and membrane permeability lead to decreased kurtosis [ 5 ]. After an mTBI there may be a disproportionate loss of small caliber axons due to their greater vulnerability to mechanical deformation from rotational forces [ 84 ]. The disproportionate loss of small caliber axons after mTBI would lead to a post-injury increase in average axon diameter in a damaged tract and thus, a decrease in diffusional kurtosis. Alternatively, studies have shown that membrane permeability increases following mTBI leading to increased transmembrane water exchange and decreased kurtosis [ 49 , 72 , 85 ]. Our histological and ultrastructural studies show tract-specific axonal damage, and myelin abnormalities that would be expected to increase transmembrane water exchange and thereby decrease diffusional kurtosis. While it is difficult to disentangle these changes from the current study, it is likely that several ultrastructural abnormalities contribute to the observed changes in diffusional kurtosis. As female mTBI mice showed reductions in both diffusivity and kurtosis metrics but males did not we conclude, as have others [ 84 , 103 ], that females may experience more extensive axonal damage and greater inflammation than males following mTBI. Frequency-dependence of changes detected by dMRI Our findings revealed significant group differences in orthogonal tracts for both MKT and RK metrics at 60 Hz. Although data at 0 Hz exhibited similar trends, the results did not reach statistical significance (MKT: F = 5.038, p = 0.050). In contrast, group differences appear less pronounced at 120 Hz. Previous studies have reported differential sensitivity between OGSE and PGSE acquisition in various models, including rodent models of ischemia [ 3 , 97 ], demyelination/inflammation [ 4 , 5 ], as well as in human ischemic stroke [ 9 ]. While microstructural changes at smaller spatial scales (i.e., axon level) would be expected to be more prominent at higher OGSE frequencies, our results suggest that a low non-zero frequency may be optimal to detect changes following mTBI when multiple microstructural changes are occurring simultaneously. Additionally, differences in waveform design—specifically, the use of frequency-tuned waveforms at 60 Hz versus conventional OGSE waveforms at 120 Hz—may yield different sensitivities to microstructural features. Region-specific and subregion-specific FC analyses reveals complementary advantages We found a significant widespread decrease in FC in mTBI mice following the injury, that was not observed in the sham group. We analyzed FC using both a coarse, region-specific atlas and a fine sub-region specific atlas. Analyzing FC between coarsely parcellated brain regions revealed decreases in long-range FC between anterior and posterior regions. As FC between anterior and posterior brain regions must rely either on direct or indirect anatomical connections that are oriented in the anterior-posterior direction, the decrease in FC is consistent with our DKI data showing that tract-specific injury predominates in tracts oriented orthogonal to the right-left axis of rotation produced by the injury. Subregion-specific analyses were used to identify subregions responsible for the reductions in FC between brain regions and in at least one case also identified a decrease in FC between subregions (the amygdala and the vermal regions of the cerebellum) that were not identified in coarser regional analyses. The identification of FC changes between subregions but not between the larger regions that they lie in may be due to those FC changes being obscured by FC measures between the other component subregions making detection difficult. There were also some FC changes detectable by region-specific analyses that were not detected by sub-region analyses. This may have been because the FC changes by the region-specific analyses were due to the cumulative effect of multiple small changes in sub-region FC that failed individually to produce a significant change in FC. The reduced complexity of brain region-specific FC analyses allowed us to appreciate the spatial relationships between brain regions experiencing FC changes after mTBI such as L Olfactory-L cerebellum, R cortex-L midbrain, and R hippocampus-L midbrain because it entailed fewer region-to-region FC comparisons (231 vs 3321). The subregions specific analyses had the advantage of identifying specific subregions within the larger brain structures. Because our understanding of animal and human behavior is based on our understanding of the functions of subregions of the brain, the subregion-specific FC analysis may be useful when trying to predict the types of behavioral changes expected after mTBI. Predicting behavioral changes post-mTBI based on FC losses between brain regions would be difficult as those relatively larger regions of the brain subserve multiple functions. Right -left asymmetries in FC analyses Right-left asymmetries were found in the FC analyses. In particular, decreases in FC between the isocortex and its targets in the hippocampus, hindbrain and cerebellum were predominantly right-sided. This may reflect right-left asymmetries in FC that naturally exist in mice that were revealed by injury. Many right-left FC asymmetries were found in FC of shams, consistent with the known naturally asymmetry in the mouse brain that has been reported in the literature [ 32 , 79 ]. Data-driven FC analyses In addition to the atlas-based analyses discussed above, we also carried out a data driven approach called ICA to decompose the brain into spatially independent networks[ 10 ] and were able to identify the most-commonly reported brain networks [ 29 , 101 ]. Between-component analysis validated our observations of a loss of FC at 1 week after injury in networks containing the somatomotor region, olfactory regions, hippocampus, midbrain, and cerebellum with a notable degree of asymmetry as previously elaborated in the atlas-based analysis. As expected, the networks with reduced FC had significant contributions from brain regions in the anterior and posterior parts of the brain. Interestingly shams showed an increase in FC in the right somatomotor network 1-week post-sham procedure suggesting that the decreased FC in this network in mTBI mice is more significant than might otherwise be appreciated. Studies using the engineered rotational acceleration CHIMERA model of mTBI[ 61 ] that is similar to the model used in the present manuscript report similar decreases in FC at 7 days post-injury [ 76 , 89 ]. Changes in FC following TBI are well-documented in both human (for a review see, [ 23 , 59 ]) and animal literature [ 19 , 76 , 80 , 89 , 93 , 99 ], however, the direction and size of these FC changes varies widely between studies. Human literature reveals that changes in FC following head injury is very time-dependent. At 24–72 hours following a head injury FC increases as a response to the injury [ 11 , 57 , 63 , 78 , 104 ]. The acute increase of synaptic glutamate following a brain injury might explain the increase in FC as homeostatic plasticity restores the excitatory/inhibitory balance (reviewed in [ 31 ]). The acute, post-injury increase in FC is followed by a decrease in FC starting 4–5 days after the injury and can last up to a few months [ 11 , 60 , 70 , 92 , 102 , 104 ]. At several months after the injury FC may increase once again as the injured brain remodels circuits to compensate for lost functions [ 1 , 11 , 57 ]. Comparing direct FC across WM fibers that run orthogonal to plane of injury The region and subregion-specific FC analyses could not distinguish between direct (monosynaptic) and indirect (polysynaptic) connections. To directly evaluate FC changes served by tracts running orthogonal to the right-left axis of rotation, we evaluated FC changes between brain regions shown to be directly connected by viral tract-tracing studies. We found FC between regions connected by the CST to be the most affected by mTBI. Reductions in FC in other tracts (OT and cingulum bundle) oriented orthogonal to the right-left axis of rotation were less robust and not statistically significant. This might be explained by the greater length of the CST that would render it more vulnerable to damage. It is also possible that the use of smaller source and target regions to calculate FC for the other tracts made detecting statistically significant FC changes more difficult. Conclusions: Implications for concussion imaging studies Several conclusions may be made from this study: First, this study shows that after rotational mTBI, tracts orthogonal to the rotation axis experience the most damage in terms of microstructural changes detected with dMRI, which leads to loss of FC between regions connected (either directly or indirectly) by these tracts. It is notable that animal models of mTBI that include a rotational component about a right-left axis, as in the model of mTBI reported here, show similar patterns of injury. Groups using the CHIMERA, or momentum exchange model of mTBI, that cause head rotations around a right-left axis, report injury focused in the optic tract[ 19 , 76 ] and external capsule[ 89 ] by FC and histological analyses [ 50 ]. On the other hand models of mTBI where the animals are struck on the side of the head producing an axis of rotation due to injury that is oriented in the anterior-posterior direction induces changes in the tracts that run right-left such as the corpus callosum as demonstrated by predominant FC changes in interhemispheric connectivity [ 2 , 40 , 64 , 71 ]. Thus, the inconsistency in identifying regions and tracts affected by mTBI in the human literature may stem from the heterogeneity in the axes of rotation produced by different injuries in humans leading to patterns of injury that will be as varied as the injuries themselves. In animal studies of mTBI where the injuries are more consistent it may be that greater care in identifying the axis of rotation of a particular injury and analyzing the imaging data with the recognition that the injured tracts will be those oriented orthogonal to the axis of injury will improve the reproducibility and reliability of results. Second, our results suggest that, while both diffusivity and kurtosis metrics could identify damage in mice after mTBI, kurtosis metrics were more sensitive to tract-specific damage. Thus, while DKI requires a more extensive acquisition protocol, both DTI and DKI are useful in the detection of pathological changes following mTBI. Third we show that the use of frequency-dependent dMRI further increased sensitivity to small structural changes from mTBI compared to conventionally used PGSE acquisitions. Fourth, our study revealed sex differences detectable by dMRI that were not detectable fMRI. In particular, female mTBI mice showed greater reduction in dMRI metrics than male mTBI mice. These results demonstrate the importance of analyzing mTBI data for sex differences and also indicates that fMRI may have less sensitivity to mTBI pathology than dMRI. Fifth, we found utility in carrying out the fMRI analysis in a region-specific and subregion-specific manner. The region-specific analysis was based on parcellating the mouse brain into 22 regions (11 per hemisphere) while the subregion-specific analysis was based on parcellating the mouse brain into 82 regions (41 per hemisphere). The region-specific analysis, being dependent on a coarser parcellation of the brain than the subregion analysis, is less vulnerable to variability in the data caused by registration errors, and movement artifacts than the subregion-specific. Moreover, averaging across bigger regions translates to higher SNR. The greater specificity of the subregion analysis however will allow future studies to predict behavioral outcomes from mTBI with greater accuracy than the region-specific analysis as our knowledge of brain function is based on anatomical subregions. It is hoped that these lessons learned will improve the reliability and reproducibility of imaging concussion in animals and guide our thinking about how to better image the concussed human brain. Declarations Authors Contributions: Conceptualization: A.E., J.H., X.X., N.G., C.A.B., R.S.M., A.B.; Data curation: A.E., J.H., X.X., N.G.; Formal analysis: A.E., J.H., X.X.; Funding acquisition: A.B.; V.F.P., M.A.M.P., C.A.B., R.S.M Resources: X.X., N.G., V.F.P., M.A.M.P., A.B.; Supervision: C.A.B., R.S.M., A.B.; Writing - original draft preparation: A.E., J.H., X.X.; Writing - review and editing: A.E., J.H., X.X., N.G., V.F.P., M.A.M.P., C.A.B., R.S.M., A.B. Ethics approval and consent to participate : Ethics and Consent to Participate declarations: not applicable. Consent for publication: Consent to Publish declaration: not applicable. Data Availability : Raw MRI data (NIfTI format) will be made publicly available through the Federated Research Data Repository (FRDR) prior to publication. All analysis code used in this study is available from the corresponding author upon reasonable request. Competing Interests: The authors have no relevant financial or non-financial interests to disclose. Funding: J. H. was supported by the Natural Sciences and Engineering Research Council of Canada: Canada Graduate Scholarships—Doctoral Program (NSERC-CGS D). C. A. B. was supported by Canada Research Chairs (950-231993). Data collection was supported by the Canada First Research Excellence Fund to BrainsCAN. Study design, data collection, analyses, interpretation and manuscript preparation was supported by the Canadian Institute of Health Research FDN 148453, the National Hockey League Players Association Challenge Fund, and the US Department of Defense under congress-directed medical research program (CDMRP), Peer Reviewed Alzheimer’s Research Program (PRARP) by award# W81XWH-20-1-0323. Acknowledgments: The MAPTKI mice were a kind gift from Dr. Takaomi C. Saido (RIKEN Brain Science Institute). References Abbas K, Shenk TE, Poole VN, Breedlove EL, Leverenz LJ, Nauman EA et al (2015) Alteration of Default Mode Network in High School Football Athletes Due to Repetitive Subconcussive Mild Traumatic Brain Injury: A Resting-State Functional Magnetic Resonance Imaging Study. Brain Connect 5:91–101. doi: 10.1089/brain.2014.0279 Adams C, Bazzigaluppi P, Beckett TL, Bishay J, Weisspapir I, Dorr A et al (2018) Neurogliovascular dysfunction in a model of repeated traumatic brain injury. Theranostics 8:4824–4836. doi: 10.7150/thno.24747 Aggarwal M, Burnsed J, Martin LJ, Northington FJ, Zhang J (2014) Imaging neurodegeneration in the mouse hippocampus after neonatal hypoxia-ischemia using oscillating gradient diffusion MRI. Magn Reson Med 72:829–840. doi: 10.1002/mrm.24956 Aggarwal M, Jones M V., Calabresi PA, Mori S, Zhang J (2012) Probing mouse brain microstructure using oscillating gradient diffusion MRI. Magn Reson Med 67:98–109. doi: 10.1002/mrm.22981 Aggarwal M, Smith MD, Calabresi PA (2020) Diffusion‐time dependence of diffusional kurtosis in the mouse brain. Magn Reson Med 84:1564–1578. doi: 10.1002/mrm.28189 Angelova P, Kehayov I, Davarski A, Kitov B (2021) Contemporary insight into diffuse axonal injury. Folia Med (Plovdiv) 63:163–170. doi: 10.3897/folmed.63.e53709 Avants BB, Tustison NJ, Song G, Cook PA, Klein A, Gee JC (2011) A reproducible evaluation of ANTs similarity metric performance in brain image registration. Neuroimage 54:2033–44. doi: 10.1016/j.neuroimage.2010.09.025 Baglietto-Vargas D, Forner S, Cai L, Martini AC, Trujillo-Estrada L, Swarup V et al (2021) Generation of a humanized Aβ expressing mouse demonstrating aspects of Alzheimer’s disease-like pathology. Nat Commun 12:2421. doi: 10.1038/s41467-021-22624-z Baron CA, Kate M, Gioia L, Butcher K, Emery D, Budde M et al (2015) Reduction of Diffusion-Weighted Imaging Contrast of Acute Ischemic Stroke at Short Diffusion Times. Stroke 46:2136–2141. doi: 10.1161/STROKEAHA.115.008815 Beckmann CF, DeLuca M, Devlin JT, Smith SM (2005) Investigations into resting-state connectivity using independent component analysis. Philosophical Transactions of the Royal Society B: Biological Sciences 360:1001–1013. doi: 10.1098/rstb.2005.1634 Bharath RD, Munivenkatappa A, Gohel S, Panda R, Saini J, Rajeswaran J et al (2015) Recovery of resting brain connectivity ensuing mild traumatic brain injury. Front Hum Neurosci 9. doi: 10.3389/fnhum.2015.00513 Blennow K, Brody DL, Kochanek PM, Levin H, McKee A, Ribbers GM et al (2016) Traumatic brain injuries. Nat Rev Dis Primers 2:16084. doi: 10.1038/nrdp.2016.84 Borsos KB, Tse DHY, Dubovan PI, Baron CA (2023) Tuned bipolar oscillating gradients for mapping frequency dispersion of diffusion kurtosis in the human brain. Magn Reson Med 89:756–766. doi: 10.1002/mrm.29473 Boshra R, Ruiter KI, Dhindsa K, Sonnadara R, Reilly JP, Connolly JF (2020) On the time-course of functional connectivity: theory of a dynamic progression of concussion effects. Brain Commun 2. doi: 10.1093/braincomms/fcaa063 Braun NJ, Liao D, Alford PW (2021) Orientation of neurites influences severity of mechanically induced tau pathology. Biophys J 120:3272–3282. doi: 10.1016/j.bpj.2021.07.011 Browne KD, Chen X-H, Meaney DF, Smith DH (2011) Mild Traumatic Brain Injury and Diffuse Axonal Injury in Swine. J Neurotrauma 28:1747–1755. doi: 10.1089/neu.2011.1913 Chen J, Ades-Aron B, Lee H-H, Mehrin S, Pang M, Novikov DS et al (2024) Optimization and validation of the DESIGNER preprocessing pipeline for clinical diffusion MRI in white matter aging. Imaging Neuroscience 2:1–17. doi: 10.1162/imag_a_00125 Cox RW (1996) AFNI: Software for Analysis and Visualization of Functional Magnetic Resonance Neuroimages. Computers and Biomedical Research 29:162–173. doi: 10.1006/cbmr.1996.0014 Criado-Marrero M, Ravi S, Bhaskar E, Barroso D, Pizzi MA, Williams L et al (2024) Age dictates brain functional connectivity and axonal integrity following repetitive mild traumatic brain injuries in mice. Neuroimage 298:120764. doi: 10.1016/j.neuroimage.2024.120764 Damoiseaux JS, Greicius MD (2009) Greater than the sum of its parts: a review of studies combining structural connectivity and resting-state functional connectivity. Brain Struct Funct 213:525–33. doi: 10.1007/s00429-009-0208-6 Dean PJA, Sato JR, Vieira G, McNamara A, Sterr A (2015) Long-term structural changes after mTBI and their relation to post-concussion symptoms. Brain Inj 29:1211–1218. doi: 10.3109/02699052.2015.1035334 Dollé J-P, Jaye A, Anderson SA, Ahmadzadeh H, Shenoy VB, Smith DH (2018) Newfound sex differences in axonal structure underlie differential outcomes from in vitro traumatic axonal injury. Exp Neurol 300:121–134. doi: 10.1016/j.expneurol.2017.11.001 Eierud C, Craddock RC, Fletcher S, Aulakh M, King-Casas B, Kuehl D et al (2014) Neuroimaging after mild traumatic brain injury: Review and meta-analysis. Neuroimage Clin 4:283–294. doi: 10.1016/j.nicl.2013.12.009 Fakhran S, Yaeger K, Collins M, Alhilali L (2014) Sex Differences in White Matter Abnormalities after Mild Traumatic Brain Injury: Localization and Correlation with Outcome. Radiology 272:815–823. doi: 10.1148/radiol.14132512 Gavish M, Donoho DL (2017) Optimal Shrinkage of Singular Values. IEEE Trans Inf Theory 63:2137–2152. doi: 10.1109/TIT.2017.2653801 Gennarelli TA, Thibault LE, Adams JH, Graham DI, Thompson CJ, Marcincin RP (1982) Diffuse axonal injury and traumatic coma in the primate. Ann Neurol 12:564–574. doi: 10.1002/ana.410120611 Gennarelli TA, Thibault LE, Tomei G, Wiser R, Graham D, Adams J (1987) Directional Dependence of Axonal Brain Injury due to Centroidal and Non-Centroidal Acceleration Gorgolewski K, Burns CD, Madison C, Clark D, Halchenko YO, Waskom ML et al (2011) Nipype: A Flexible, Lightweight and Extensible Neuroimaging Data Processing Framework in Python. Front Neuroinform 5. doi: 10.3389/fninf.2011.00013 Grandjean J, Canella C, Anckaerts C, Ayrancı G, Bougacha S, Bienert T et al (2020) Common functional networks in the mouse brain revealed by multi-centre resting-state fMRI analysis. Neuroimage 205:116278. doi: 10.1016/j.neuroimage.2019.116278 Greicius MD, Supekar K, Menon V, Dougherty RF (2009) Resting-state functional connectivity reflects structural connectivity in the default mode network. Cereb Cortex 19:72–8. doi: 10.1093/cercor/bhn059 Guerriero RM, Giza CC, Rotenberg A (2015) Glutamate and GABA Imbalance Following Traumatic Brain Injury. Curr Neurol Neurosci Rep 15:27. doi: 10.1007/s11910-015-0545-1 Güntürkün O, Ströckens F, Ocklenburg S (2020) Brain Lateralization: A Comparative Perspective. Physiol Rev 100:1019–1063. doi: 10.1152/physrev.00006.2019 Gupte RP, Brooks WM, Vukas RR, Pierce JD, Harris JL (2019) Sex Differences in Traumatic Brain Injury: What We Know and What We Should Know. J Neurotrauma 36:3063–3091. doi: 10.1089/neu.2018.6171 Haacke EM, Lindskogj ED, Lin W (1991) A fast, iterative, partial-fourier technique capable of local phase recovery. Journal of Magnetic Resonance (1969) 92:126–145. doi: 10.1016/0022-2364(91)90253-P Hajiaghamemar M, Margulies SS (2021) Multi-Scale White Matter Tract Embedded Brain Finite Element Model Predicts the Location of Traumatic Diffuse Axonal Injury. J Neurotrauma 38:144–157. doi: 10.1089/neu.2019.6791 Hamilton J, Xu K, Geremia N, Prado VF, Prado MAM, Brown A et al (2024) Robust frequency-dependent diffusional kurtosis computation using an efficient direction scheme, axisymmetric modelling, and spatial regularization. Imaging Neuroscience 2:1–22. doi: 10.1162/imag_a_00055 Harris NG, Verley DR, Gutman BA, Thompson PM, Yeh HJ, Brown JA (2016) Disconnection and hyper-connectivity underlie reorganization after TBI: A rodent functional connectomic analysis. Exp Neurol 277:124–138. doi: 10.1016/j.expneurol.2015.12.020 Hellewell SC, Nguyen VPB, Jayasena RN, Welton T, Grieve SM (2020) Characteristic patterns of white matter tract injury in sport-related concussion: An image based meta-analysis. Neuroimage Clin 26:102253. doi: 10.1016/j.nicl.2020.102253 Honey CJ, Sporns O, Cammoun L, Gigandet X, Thiran JP, Meuli R et al (2009) Predicting human resting-state functional connectivity from structural connectivity. Proc Natl Acad Sci U S A 106:2035–40. doi: 10.1073/pnas.0811168106 Huang S, Shen Q, Watts LT, Long JA, O’Boyle M, Nguyen T et al (2021) Resting-State Functional Magnetic Resonance Imaging of Interhemispheric Functional Connectivity in Experimental Traumatic Brain Injury. Neurotrauma Rep 2:526–540. doi: 10.1089/neur.2021.0023 Hutchinson EB, Schwerin SC, Avram A V., Juliano SL, Pierpaoli C (2018) Diffusion MRI and the detection of alterations following traumatic brain injury. J Neurosci Res 96:612–625. doi: 10.1002/jnr.24065 Jelescu IO, Fieremans E (2023) Sensitivity and specificity of diffusion MRI to neuroinflammatory processes. pp 31–50. doi: 10.1016/b978-0-323-91771-1.00010-1 Jenkinson M, Beckmann CF, Behrens TEJ, Woolrich MW, Smith SM (2012) FSL. Neuroimage 62:782–790. doi: 10.1016/j.neuroimage.2011.09.015 Jensen JH, Helpern JA (2010) MRI quantification of non-Gaussian water diffusion by kurtosis analysis. NMR Biomed 23:698–710. doi: 10.1002/nbm.1518 Jensen JH, Helpern JA, Ramani A, Lu H, Kaczynski K (2005) Diffusional kurtosis imaging: The quantification of non-gaussian water diffusion by means of magnetic resonance imaging. Magn Reson Med 53:1432–1440. doi: 10.1002/mrm.20508 Kellner E, Dhital B, Kiselev VG, Reisert M (2016) Gibbs‐ringing artifact removal based on local subvoxel‐shifts. Magn Reson Med 76:1574–1581. doi: 10.1002/mrm.26054 Knutsen AK, Gomez AD, Gangolli M, Wang W-T, Chan D, Lu Y-C et al (2020) In vivo estimates of axonal stretch and 3D brain deformation during mild head impact. Brain Multiphys 1:100015. doi: 10.1016/j.brain.2020.100015 Koay CG, Basser PJ (2006) Analytically exact correction scheme for signal extraction from noisy magnitude MR signals. Journal of Magnetic Resonance 179:317–322. doi: 10.1016/j.jmr.2006.01.016 Krieg JL, Leonard A V., Turner RJ, Corrigan F (2023) Identifying the Phenotypes of Diffuse Axonal Injury Following Traumatic Brain Injury. Brain Sci 13:1607. doi: 10.3390/brainsci13111607 Kulkarni P, Morrison TR, Cai X, Iriah S, Simon N, Sabrick J et al (2019) Neuroradiological Changes Following Single or Repetitive Mild TBI. Front Syst Neurosci 13:34. doi: 10.3389/fnsys.2019.00034 Lee H, Novikov DS, Fieremans E (2021) Removal of partial Fourier‐induced Gibbs (RPG) ringing artifacts in MRI. Magn Reson Med 86:2733–2750. doi: 10.1002/mrm.28830 Leemans A, Jeurissen D, Sijbers J, Jones DK (2009) ExploreDTI: a graphical toolbox for processing, analyzing, and visualizing diffusion MR data. In: 17th Annual Meeting of Intl Soc Mag Reson Med. Hawaii, p 3537 Lindsey HM, Hodges CB, Greer KM, Wilde EA, Merkley TL (2023) Diffusion-Weighted Imaging in Mild Traumatic Brain Injury: A Systematic Review of the Literature. Neuropsychol Rev 33:42–121. doi: 10.1007/s11065-021-09485-5 Liu X-B, Schumann CM (2014) Optimization of electron microscopy for human brains with long-term fixation and fixed-frozen sections. Acta Neuropathol Commun 2:42. doi: 10.1186/2051-5960-2-42 Maas AIR, Menon DK, Manley GT, Abrams M, Åkerlund C, Andelic N et al (2022) Traumatic brain injury: progress and challenges in prevention, clinical care, and research. Lancet Neurol 21:1004–1060. doi: 10.1016/S1474-4422(22)00309-X Macdonald C, Dikranian K, Song S, Bayly P, Holtzman D, Brody D (2007) Detection of traumatic axonal injury with diffusion tensor imaging in a mouse model of traumatic brain injury. Exp Neurol 205:116–131. doi: 10.1016/j.expneurol.2007.01.035 Manning KY, Schranz A, Bartha R, Dekaban GA, Barreira C, Brown A et al (2017) Multiparametric MRI changes persist beyond recovery in concussed adolescent hockey players. Neurology 89:2157–2166. doi: 10.1212/WNL.0000000000004669 Markicevic M, Mandino F, Toyonaga T, Cai Z, Fesharaki-Zadeh A, Shen X et al (2024) Repetitive Mild Closed-Head Injury Induced Synapse Loss and Increased Local BOLD-fMRI Signal Homogeneity. J Neurotrauma 41:2528–2544. doi: 10.1089/neu.2024.0095 Mayer AR, Bellgowan PSF, Hanlon FM (2015) Functional magnetic resonance imaging of mild traumatic brain injury. Neurosci Biobehav Rev 49:8–18. doi: 10.1016/j.neubiorev.2014.11.016 Mayer AR, Mannell M V., Ling J, Gasparovic C, Yeo RA (2011) Functional connectivity in mild traumatic brain injury. Hum Brain Mapp 32:1825–1835. doi: 10.1002/hbm.21151 McNamara EH, Grillakis AA, Tucker LB, McCabe JT (2020) The closed-head impact model of engineered rotational acceleration (CHIMERA) as an application for traumatic brain injury pre-clinical research: A status report. Exp Neurol 333:113409. doi: 10.1016/j.expneurol.2020.113409 Meaney DF, Smith DH (2011) Biomechanics of Concussion. Clin Sports Med 30:19–31. doi: 10.1016/j.csm.2010.08.009 Meier TB, Giraldo-Chica M, España LY, Mayer AR, Harezlak J, Nencka AS et al (2020) Resting-State fMRI Metrics in Acute Sport-Related Concussion and Their Association with Clinical Recovery: A Study from the NCAA-DOD CARE Consortium. J Neurotrauma 37:152–162. doi: 10.1089/neu.2019.6471 Mohamed AZ, Cumming P, Nasrallah FA (2021) Traumatic brain injury augurs ill for prolonged deficits in the brain’s structural and functional integrity following controlled cortical impact injury. Sci Rep 11:21559. doi: 10.1038/s41598-021-00660-5 Morozov S, Sergunova K, Petraikin A, Akhmad E, Kivasev S, Semenov D et al (2020) Diffusion processes modeling in magnetic resonance imaging. Insights Imaging 11:60. doi: 10.1186/s13244-020-00863-w Nickerson LD, Smith SM, Öngür D, Beckmann CF (2017) Using Dual Regression to Investigate Network Shape and Amplitude in Functional Connectivity Analyses. Front Neurosci 11. doi: 10.3389/fnins.2017.00115 Novikov DS, Kiselev VG (2011) Surface-to-volume ratio with oscillating gradients. Journal of Magnetic Resonance 210:141–145. doi: 10.1016/j.jmr.2011.02.011 Oh SW, Harris JA, Ng L, Winslow B, Cain N, Mihalas S et al (2014) A mesoscale connectome of the mouse brain. Nature 508:207–214. doi: 10.1038/nature13186 Olesen JL, Ianus A, Østergaard L, Shemesh N, Jespersen SN (2023) Tensor denoising of multidimensional MRI data. Magn Reson Med 89:1160–1172. doi: 10.1002/mrm.29478 Palacios EM, Yuh EL, Chang Y-S, Yue JK, Schnyer DM, Okonkwo DO et al (2017) Resting-State Functional Connectivity Alterations Associated with Six-Month Outcomes in Mild Traumatic Brain Injury. J Neurotrauma 34:1546–1557. doi: 10.1089/neu.2016.4752 Parent M, Li Y, Santhakumar V, Hyder F, Sanganahalli BG, Kannurpatti SS (2019) Alterations of Parenchymal Microstructure, Neuronal Connectivity, and Cerebrovascular Resistance at Adolescence after Mild-to-Moderate Traumatic Brain Injury in Early Development. J Neurotrauma 36:601–608. doi: 10.1089/neu.2018.5741 Pettus EH, Christman CW, Giebel ML, Polvishock JT (1994) Traumatically Induced Altered Membrane Permeability: Its Relationship to Traumatically Induced Reactive Axonal Change. J Neurotrauma 11:507–522. doi: 10.1089/neu.1994.11.507 Rahman N, Xu K, Budde MD, Brown A, Baron CA (2023) A longitudinal microstructural MRI dataset in healthy C57Bl/6 mice at 9.4 Tesla. Sci Data 10:94. doi: 10.1038/s41597-023-01942-5 Rubiano AM, Carney N, Chesnut R, Puyana JC (2015) Global neurotrauma research challenges and opportunities. Nature 527:S193–S197. doi: 10.1038/nature16035 Saito T, Mihira N, Matsuba Y, Sasaguri H, Hashimoto S, Narasimhan S et al (2019) Humanization of the entire murine Mapt gene provides a murine model of pathological human tau propagation. Journal of Biological Chemistry 294:12754–12765. doi: 10.1074/jbc.RA119.009487 Sakthivel R, Criado-Marrero M, Barroso D, Braga IM, Bolen M, Rubinovich U et al (2023) Fixed Time-Point Analysis Reveals Repetitive Mild Traumatic Brain Injury Effects on Resting State Functional Magnetic Resonance Imaging Connectivity and Neuro-Spatial Protein Profiles. J Neurotrauma 40:2037–2049. doi: 10.1089/neu.2022.0464 Schachter M, Does MD, Anderson AW, Gore JC (2000) Measurements of Restricted Diffusion Using an Oscillating Gradient Spin-Echo Sequence. Journal of Magnetic Resonance 147:232–237. doi: 10.1006/jmre.2000.2203 Shumskaya E, Andriessen TMJC, Norris DG, Vos PE (2012) Abnormal whole-brain functional networks in homogeneous acute mild traumatic brain injury. Neurology 79:175–182. doi: 10.1212/WNL.0b013e31825f04fb Silberfeld A, Roe JM, Ellegood J, Lerch JP, Qiu L, Kim Y et al (2025) Right-left Brain-Wide Asymmetry of Neuroanatomy in the Mouse Brain. Neuroimage 307:121017. doi: 10.1016/j.neuroimage.2025.121017 Sinke MRT, Otte WM, Meerwaldt AE, Franx BAA, Ali MHM, Rakib F et al (2021) Imaging Markers for the Characterization of Gray and White Matter Changes from Acute to Chronic Stages after Experimental Traumatic Brain Injury. J Neurotrauma 38:1642–1653. doi: 10.1089/neu.2020.7151 Skudlarski P, Jagannathan K, Calhoun VD, Hampson M, Skudlarska BA, Pearlson G (2008) Measuring brain connectivity: diffusion tensor imaging validates resting state temporal correlations. Neuroimage 43:554–61. doi: 10.1016/j.neuroimage.2008.07.063 Smith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TEJ, Johansen-Berg H et al (2004) Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage 23:S208–S219. doi: 10.1016/j.neuroimage.2004.07.051 Sollmann N, Echlin PS, Schultz V, Viher P V., Lyall AE, Tripodis Y et al (2018) Sex differences in white matter alterations following repetitive subconcussive head impacts in collegiate ice hockey players. Neuroimage Clin 17:642–649. doi: 10.1016/j.nicl.2017.11.020 Song H, Tomasevich A, Paolini A, Browne KD, Wofford KL, Kelley B et al (2024) Sex differences in the extent of acute axonal pathologies after experimental concussion. Acta Neuropathol 147:79. doi: 10.1007/s00401-024-02735-9 Stone JR, Okonkwo DO, Dialo AO, Rubin DG, Mutlu LK, Povlishock JT et al (2004) Impaired axonal transport and altered axolemmal permeability occur in distinct populations of damaged axons following traumatic brain injury. Exp Neurol 190:59–69. doi: 10.1016/j.expneurol.2004.05.022 Sullivan S, Eucker SA, Gabrieli D, Bradfield C, Coats B, Maltese MR et al (2015) White matter tract-oriented deformation predicts traumatic axonal brain injury and reveals rotational direction-specific vulnerabilities. Biomech Model Mechanobiol 14:877–896. doi: 10.1007/s10237-014-0643-z Tayebi M, Holdsworth SJ, Champagne AA, Cook DJ, Nielsen P, Lee T-R et al (2021) The role of diffusion tensor imaging in characterizing injury patterns on athletes with concussion and subconcussive injury: a systematic review. Brain Inj 35:621–644. doi: 10.1080/02699052.2021.1895313 Teasell EM, Potts E, Geremia N, Lu L, Xu X, Mao H et al (2025) A Clinically Relevant Mouse Model of Concussion Incorporating High Rotational Forces. Neurotrauma Rep 6:184–190. doi: 10.1089/neur.2024.0165 To XV, Nasrallah FA (2021) A roadmap of brain recovery in a mouse model of concussion: insights from neuroimaging. Acta Neuropathol Commun 9:2. doi: 10.1186/s40478-020-01098-y Velayudhan PS, Mak JJ, Gazdzinski LM, Wheeler AL (2022) Persistent white matter vulnerability in a mouse model of mild traumatic brain injury. BMC Neurosci 23:46. doi: 10.1186/s12868-022-00730-y Veraart J, Novikov DS, Christiaens D, Ades-aron B, Sijbers J, Fieremans E (2016) Denoising of diffusion MRI using random matrix theory. Neuroimage 142:394–406. doi: 10.1016/j.neuroimage.2016.08.016 Verley DR, Torolira D, Pulido B, Gutman B, Bragin A, Mayer A et al (2018) Remote Changes in Cortical Excitability after Experimental Traumatic Brain Injury and Functional Reorganization. J Neurotrauma 35:2448–2461. doi: 10.1089/neu.2017.5536 Vinh To X, Soni N, Medeiros R, Alateeq K, Nasrallah FA (2022) Traumatic brain injury alterations in the functional connectome are associated with neuroinflammation but not tau in a P30IL tauopathy mouse model. Brain Res 1789:147955. doi: 10.1016/j.brainres.2022.147955 Wang Q, Ding S-L, Li Y, Royall J, Feng D, Lesnar P et al (2020) The Allen Mouse Brain Common Coordinate Framework: A 3D Reference Atlas. Cell 181:936-953.e20. doi: 10.1016/j.cell.2020.04.007 Winkler AM, Ridgway GR, Webster MA, Smith SM, Nichols TE (2014) Permutation inference for the general linear model. Neuroimage 92:381–397. doi: 10.1016/j.neuroimage.2014.01.060 Wright DK, Symons GF, O’Brien WT, McDonald SJ, Zamani A, Major B et al (2021) Diffusion Imaging Reveals Sex Differences in the White Matter Following Sports-Related Concussion. Cerebral Cortex 31:4411–4419. doi: 10.1093/cercor/bhab095 Wu D, Li Q, Northington FJ, Zhang J (2018) Oscillating gradient diffusion kurtosis imaging of normal and injured mouse brains. NMR Biomed 31:e3917. doi: 10.1002/nbm.3917 Xu X, Cowan M, Beraldo F, Schranz A, McCunn P, Geremia N et al (2021) Repetitive mild traumatic brain injury in mice triggers a slowly developing cascade of long-term and persistent behavioral deficits and pathological changes. Acta Neuropathol Commun 9:60. doi: 10.1186/s40478-021-01161-2 Yang Z, Zhu T, Pompilus M, Fu Y, Zhu J, Arjona K et al (2021) Compensatory functional connectome changes in a rat model of traumatic brain injury. Brain Commun 3. doi: 10.1093/braincomms/fcab244 Yushkevich PA, Gao Y, Gerig G (2016) ITK-SNAP: An interactive tool for semi-automatic segmentation of multi-modality biomedical images. In: 2016 38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). IEEE, pp 3342–3345 Zerbi V, Grandjean J, Rudin M, Wenderoth N (2015) Mapping the mouse brain with rs-fMRI: An optimized pipeline for functional network identification. Neuroimage 123:11–21. doi: 10.1016/j.neuroimage.2015.07.090 Zhang J, Solar K, Safar K, Zamyadi R, Vandewouw MM, Da Costa L et al (2024) The structural, functional, and neurophysiological connectome of mild traumatic brain injury: a DTI, fMRI and MEG multimodal clustering and data fusion study. medRxiv. doi: 10.1101/2024.06.24.24309379 Zhao P, Zhu P, Zhang D, Yin B, Wang Y, Hussein NM et al (2022) Sex Differences in Cerebral Blood Flow and Serum Inflammatory Cytokines and Their Relationships in Mild Traumatic Brain Injury. Front Neurol 12. doi: 10.3389/fneur.2021.755152 Zhu DC, Covassin T, Nogle S, Doyle S, Russell D, Pearson RL et al (2015) A Potential Biomarker in Sports-Related Concussion: Brain Functional Connectivity Alteration of the Default-Mode Network Measured with Longitudinal Resting-State fMRI over Thirty Days. J Neurotrauma 32:327–341. doi: 10.1089/neu.2014.3413 Additional Declarations No competing interests reported. Supplementary Files Supplementarydata.docx Cite Share Download PDF Status: Published Journal Publication published 01 Dec, 2025 Read the published version in Acta Neuropathologica Communications → Version 1 posted Editorial decision: Revision requested 25 Aug, 2025 Reviews received at journal 26 Jul, 2025 Reviews received at journal 26 Jul, 2025 Reviews received at journal 24 Jul, 2025 Reviewers agreed at journal 13 Jul, 2025 Reviewers agreed at journal 12 Jul, 2025 Reviewers agreed at journal 10 Jul, 2025 Reviewers agreed at journal 10 Jul, 2025 Reviewers invited by journal 10 Jul, 2025 Editor assigned by journal 02 Jul, 2025 Submission checks completed at journal 02 Jul, 2025 First submitted to journal 26 Jun, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6985478","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":484714619,"identity":"6ab356d6-4c82-4e9a-918b-391b04d194fd","order_by":0,"name":"Amr Eed","email":"","orcid":"","institution":"Western University","correspondingAuthor":false,"prefix":"","firstName":"Amr","middleName":"","lastName":"Eed","suffix":""},{"id":484714620,"identity":"d7f67dea-7bea-45a6-9740-5ff0c0df68d0","order_by":1,"name":"Jake Hamilton","email":"","orcid":"","institution":"Western University","correspondingAuthor":false,"prefix":"","firstName":"Jake","middleName":"","lastName":"Hamilton","suffix":""},{"id":484714621,"identity":"2d4ad463-eaaf-4010-95ae-5dd1dddd28fc","order_by":2,"name":"Xiaoyun Xu","email":"","orcid":"","institution":"Western University","correspondingAuthor":false,"prefix":"","firstName":"Xiaoyun","middleName":"","lastName":"Xu","suffix":""},{"id":484714622,"identity":"554cede5-656c-4a50-b512-532959e226a7","order_by":3,"name":"Nicole Geremia","email":"","orcid":"","institution":"Western University","correspondingAuthor":false,"prefix":"","firstName":"Nicole","middleName":"","lastName":"Geremia","suffix":""},{"id":484714629,"identity":"872e16ad-65e1-4b31-bd67-08c55e2bc79a","order_by":4,"name":"Vania F. Prado","email":"","orcid":"","institution":"Western University","correspondingAuthor":false,"prefix":"","firstName":"Vania","middleName":"F.","lastName":"Prado","suffix":""},{"id":484714631,"identity":"a299af3b-93f8-47aa-9c78-ec662b45adab","order_by":5,"name":"Marco A.M. Prado","email":"","orcid":"","institution":"Western University","correspondingAuthor":false,"prefix":"","firstName":"Marco","middleName":"A.M.","lastName":"Prado","suffix":""},{"id":484714633,"identity":"7005d514-9211-4a1b-9875-b68fe9c1c00c","order_by":6,"name":"Corey A. Baron","email":"","orcid":"","institution":"Western University","correspondingAuthor":false,"prefix":"","firstName":"Corey","middleName":"A.","lastName":"Baron","suffix":""},{"id":484714635,"identity":"2f8a0e77-100e-43be-ac17-e98d579dc72e","order_by":7,"name":"Ravi S. Menon","email":"","orcid":"","institution":"Western University","correspondingAuthor":false,"prefix":"","firstName":"Ravi","middleName":"S.","lastName":"Menon","suffix":""},{"id":484714637,"identity":"e84cf8bd-462e-4921-88ba-bd90812bf074","order_by":8,"name":"Arthur Brown","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAvUlEQVRIiWNgGAWjYDACCTBpA2MQryUNwjhAgpbDJGjhn9378DHvjvP5/NLNzz5/+MMgz99AyJI7x42Nec/ctpw555jxjINtDIYzCFp1I41NOrfttoHBjQRjhoMNDAkEXScP0XIOqCX9M8OBPwwJ8oS0GEC0HABqyTFmOMDGkGBASIvhnWPMxn/bkg0k55wpZjjbJmG4kZAWudttjA9nttkZ8Eu3b2ao+GMjL0dICzogPg2MglEwCkbBKMADAKJEP8Pr1QJIAAAAAElFTkSuQmCC","orcid":"","institution":"Western University","correspondingAuthor":true,"prefix":"","firstName":"Arthur","middleName":"","lastName":"Brown","suffix":""}],"badges":[],"createdAt":"2025-06-26 16:53:20","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6985478/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6985478/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s40478-025-02183-w","type":"published","date":"2025-12-01T15:57:10+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":86673643,"identity":"ba9d7e2f-fc3e-4287-bf61-6029710d22ff","added_by":"auto","created_at":"2025-07-14 11:49:03","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":316168,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRepresentative structural, functional, and diffusion MRI from a mouse at baseline.\u003c/strong\u003e\u003cem\u003e\u003cstrong\u003e \u003c/strong\u003e\u003c/em\u003ea)\u003cem\u003e\u003cstrong\u003e \u003c/strong\u003e\u003c/em\u003eA structural T2-weighted scan, b) functional MRI maps, and c) diffusion MRI maps including MD and MKT at acquisition frequencies of 0 and 120 Hz, and RD, AD, RK, AK at 0 Hz. d) Tractograms with color indicating directionality of fibers, along with tracts within parallel (0-50º) and orthogonal (80-90º) tract bins (relative to the left-to-right axis of rotation). e)\u003cstrong\u003e \u003c/strong\u003eshows tract density maps indicating the number of tracts crossing each voxel overlaid on FA maps for corresponding tractograms in (d). Abbreviations: AD, axial diffusivity; AK, axial kurtosis; AP, anterior-posterior; CBT, cerebellum tracts; CC, corpus callosum; CG, cingulum; CST, corticospinal tract; FA, fractional anisotropy; LR, right-left; MD, mean diffusivity; MKT, mean kurtosis tensor; OPT, optic tract; RD, radial diffusivity; RK, radial kurtosis; SI, superior-inferior; ST, stria terminalis.\u003c/p\u003e","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-6985478/v1/75468c31b001e856564eb644.png"},{"id":86672653,"identity":"96a634fb-336c-445d-91e0-91ed78222b30","added_by":"auto","created_at":"2025-07-14 11:41:03","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":50961,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSex differences in dMRI metric changes detected within the mTBI cohort.\u003c/strong\u003e\u003cem\u003e \u003c/em\u003edMRI metric changes in mTBI mice across tract orientation bins, with sham data shown in corresponding insets. a)\u003cstrong\u003e \u003c/strong\u003eshows sex differences in MD across acquisition frequencies, where a significant main effect of sex was found in the mTBI group across all frequencies. b) shows sex differences in MKT across acquisition frequencies, where a significant main effect of sex was found in the mTBI group across all frequencies, and at 60 and 120 Hz in the sham group. c)\u003cstrong\u003e \u003c/strong\u003eshows sex differences in RD and AD at 0 Hz, where a significant main effect of sex was found in the mTBI group across all frequencies (60 and 120 Hz data not shown). d)\u003cstrong\u003e \u003c/strong\u003eshows sex differences in RK and AK at 60 and 120 Hz acquisition frequencies, where a significant main effect of sex was found for RK in the mTBI group at 60 and 120 Hz acquisition frequency, and for AK at 60 and 120 Hz in the sham group. The data is shown as change in each cohort post-injury relative to baseline, mean +/- SEM. § signifies a main effect of sex (\u003cem\u003ep \u003c/em\u003e\u0026lt; 0.05). Abbreviations: AD, axial diffusivity; AK, axial kurtosis; MD, mean diffusivity; MKT, mean kurtosis tensor; RD, radial diffusivity; RK, radial kurtosis; SEM, standard error of the mean.\u003c/p\u003e","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-6985478/v1/0b100a348c06b96beab1a1a7.png"},{"id":86672657,"identity":"4bbac7b8-036a-4c49-b129-0fbf1b87f389","added_by":"auto","created_at":"2025-07-14 11:41:03","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":99249,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDiffusional kurtosis reveals a tract orientation dependence following injury in females\u003c/strong\u003e\u003cem\u003e\u003cstrong\u003e. \u003c/strong\u003e\u003c/em\u003edMRI metric changes in sham and mTBI female mice across tract orientation bins. a) shows group differences in MKT and RK across acquisition frequencies, where a significant main effect of group was found for both metrics at 60 Hz along with significant post-hoc findings in the 80-90º tract bin.\u003cstrong\u003e \u003c/strong\u003eb) shows group differences in MD and RD across acquisition frequencies, where a significant main effect of group was found for both metrics at all acquisition frequencies. The data is shown as change in each cohort post-injury relative to baseline, mean +/- SEM. § signifies a main effect of group (\u003cem\u003ep \u003c/em\u003e\u0026lt; 0.05), while * signifies significant post-hoc findings (\u003cem\u003ep \u003c/em\u003e\u0026lt; 0.05). Abbreviations: MD, mean diffusivity; MKT, mean kurtosis tensor; RD, radial diffusivity; RK, radial kurtosis; SEM, standard error of the mean.\u003c/p\u003e","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-6985478/v1/6d4a9d342f73ac6a8071775f.png"},{"id":86673645,"identity":"ff245be5-1b48-48d1-bfa0-0d6723c4132a","added_by":"auto","created_at":"2025-07-14 11:49:03","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":205867,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003emTBI induces loss of FC one week after injury.\u003c/strong\u003e Hypothesis testing within each group before and after sham procedure or mTBI was done using permutation testing with 10,000 permutations and corrected for multiple comparisons using FDR. a) Atlas used to segment the brain. The regions atlas was adapted from ABA CCFv3 and consists of 11 ROIs per hemisphere. Structures on the right hemisphere are color-coded and show boundaries between different regions. The anatomical atlas is shown on the left hemisphere. b) Average connectivity matrix of the sham group (left) and mTBI group (right), at baseline (top) and after 1 week of the sham procedure or mTBI (bottom). Correlation strengths were calculated as Pearson’s correlation coefficient between each pair of regions and converted to \u003cem\u003ez\u003c/em\u003e-score using Fisher’s transformation. Correlations that showed a significant decrease from baseline, 1 week after injury are highlighted in dashed phosphorus squares. Boxplots of regions that showed significant decreases in FC after the sham procedure (c) and after mTBI (d). e) Circular plot showing edges with significant decrease from baseline, 1 week after injury. The thickness of the lines reflects the percentage of decrease with the thickest line represent biggest decrease from baseline correlation (63% decrease) and the thinnest line shows smallest decrease from baseline (25%). Abbreviations: ABA, Allen Brain Atlas; CB, cerebellum; CCFv3, common coordinate framework version 3; FC, functional connectivity; FDR, false-discovery rate; HB, hindbrain; HPF, hippocampal formation; mTBI, mild traumatic brain injury; MB, midbrain; OLF, olfactory regions.\u003c/p\u003e","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-6985478/v1/57e7debb77a2e4bc63e9049a.png"},{"id":86672660,"identity":"5b5408e2-0b12-4a52-b7c8-7c91e4dd9455","added_by":"auto","created_at":"2025-07-14 11:41:03","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":157995,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003emTBI reduces inter- and intra-network FC.\u003c/strong\u003e The brain was decomposed into 20 resting-state networks using ICA and the correlation strength between and within those networks was testing using permutation testing and 10,000 permutations. Boxplots of network pairs that showed significant decreases after the sham procedure (a) and after mTBI (b). Average connectivity matrix between each pair of the ICA networks in the sham group (left) and mTBI group (right), at baseline (top) and 1 week after the sham procedure or mTBI (bottom). Correlation strengths were calculated as Pearson’s correlation coefficient between each pair of regions and converted to \u003cem\u003ez\u003c/em\u003e-score using Fisher’s transformation. Correlations showed significant decrease from baseline, 1 week after injury are highlighted in phosphorus. c) Circular plot showing edges with significant decrease from baseline, 1 week after injury. The thickness of the lines reflects the percentage of decrease with the thickest line represent biggest decrease from baseline correlation (126% decrease) and the thinnest line shows smallest decrease from baseline (49%). d) Dual regression analysis results show significant decreases in correlation strength between networks (red-yellow) and clusters (blue-light blue) after injury in mTBI group. Between-components results were corrected using false-discovery rate, while dual regression results were corrected across voxels, networks, and contrasts using family-wise error correction. ICA networks are shown overlaid on an ABA CCFv3 anatomical atlas. Abbreviations: ABA, Allen Brain Atlas; CCFv3, common coordinate framework version 3; CP, caudate putamen; FC, functional connectivity; HY, hypothalamus; ICA, independent component analysis; MO, somatomotor areas; mTBI, mild traumatic brain; PAL, pallidum; STRv, striatum ventral region; TT, taenia tecta.\u003c/p\u003e","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-6985478/v1/1a91b4d250e858eeca7120b6.png"},{"id":86673648,"identity":"31f08fd8-5cf9-47ce-adc9-ec4f988c601a","added_by":"auto","created_at":"2025-07-14 11:49:03","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":214750,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTractography-based functional connectivity shows loss of connectivity across tracts that run orthogonal rather than parallel to the plane of rotation\u003c/strong\u003e. CST (a) and CC (b) were isolated using viral tracer data [68]. The left shows primary injection location in red and the fibers 3D rendering based on ABA CCFv3, the right shows signal density, and an orientation compass is in the middle. Box plots of connectivity strength between regions connected through CST (c) and CC (d). Paired Student’s t-test was performed within each group. The top view shows masks of the source and the target used to calculate FC overlaid on an ABA CCFv3 anatomical atlas. Abbreviations: ABA, Allen Brain Atlas; CCFv3, common coordinate framework version 3; FC, functional connectivity.\u003c/p\u003e","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-6985478/v1/ed56dd12452344b3940a33ca.png"},{"id":86672664,"identity":"bbcf8a58-463d-49f8-8a3e-2359959071eb","added_by":"auto","created_at":"2025-07-14 11:41:03","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":1179529,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSilver staining, immunohistochemistry, and electron microscopy of sham and mTBI animals. \u003c/strong\u003e(a-f) Photomicrographs of sliver-stained brain sections from sham (a-c) and mTBI mice (d-f). The black-stained fibers are evidence of axonal damage. Only rare examples of a silver-stained fiber could be found in any of the WM tracts of shams. In mTBI mice only a few scattered silver-stained fibers were found in the body of the corpus callosum where axons are oriented parallel to the right-left axis. In contrast, a large number of silver-stained fibers were found in the cingulum bundle (captured in these sections as they traverse the sagittal plane in the genu of the corpus callosum) and in the sagittal fibers of the optic tracts. Scale bar =100mm. Higher magnification images of silver staining are shown in corresponding insets. Representative photomicrographs showing GFAP+ astrocytes and Iba-1+ microglia in the anterior corpus callosum (g,j), the cingulum bundle (h,k) and optic tracts (i,l), in sham and mTBI brains. High magnification images showed hypertrophic ameboid microglia in k and l inset. Only resting microglia with ramified processed were seen in sham brains (h inset). Scale bar=10 mm for g, h, j and k; Scale bar=50 mm for i and l. (m-o) Electron microscopic images illustrate the ultrastructure changes in the corpus callosum and internal capsule regions following mTBI. (m) sham axons; (n) in mTBI brains, representative microphotograph exhibits multiple stages of degenerating axons: demyelinated axons (light pink); excessive myelination (red triangle) and axon-sheath dissociation (red star). Scale bar=400 nm. (o), astrocyte (a) with prolonged protrusion was observed near the degenerative neurite (yellow triangle). Scale bar=1 mm.\u003c/p\u003e","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-6985478/v1/2f7fd178933b7fa5129592a6.png"},{"id":97723777,"identity":"fe9a614c-e3cb-4225-ba8a-81a8ff384878","added_by":"auto","created_at":"2025-12-08 16:05:58","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3731851,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6985478/v1/5e052207-a3a4-4a7d-94af-f7877e87a7b0.pdf"},{"id":86672682,"identity":"dcf222da-f7ab-4f52-8448-704fe41a984e","added_by":"auto","created_at":"2025-07-14 11:41:05","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":68536517,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarydata.docx","url":"https://assets-eu.researchsquare.com/files/rs-6985478/v1/9d8c252375523176fb01fdf1.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"MRI investigation of orientation-dependent changes in microstructure and function in a mouse model of mild traumatic brain injury","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eTraumatic brain injury (TBI) is the leading cause of death and disability among all trauma-related injuries with an estimated annual incidence of 50\u0026ndash;60\u0026nbsp;million [\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e, \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e]. Approximately 90% of TBIs are classified as mild TBI (mTBI) or concussion [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. While mTBI typically results from both linear and rotational accelerations of the brain, rotational acceleration has been identified as particularly significant due to its ability to generate shear and strain forces on brain tissue [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e]. White matter (WM) tracts are particularly susceptible to these rotational forces due to their highly anisotropic structure, explaining why diffuse axonal injury is a hallmark of concussion [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eThe magnitude and orientation of rotational forces produced during an impact cause distinct strain dynamics throughout the brain in mTBI. Studies modeling brain injuries have shown that it is not the magnitude of strain, but the component of strain oriented along axons (tract-oriented strain) that drives axonal damage [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e86\u003c/span\u003e]. Support for the clinical relevance of these findings is provided by Knutsen et al.[\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e] who used non-injurious head accelerations in human volunteers, to show that strain components orthogonal to the axis of rotation account for the majority of brain deformation, with negligible strain along the axis of rotation. While histological studies have supported this tract orientation-dependence of damage [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e], there remains a notable gap in research investigating the relationship between the orientation of rotational forces produced by an mTBI and the subsequent distribution of pathological changes in \u003cem\u003evivo\u003c/em\u003e.\u003c/p\u003e\u003cp\u003eDiffusion tensor imaging (DTI), a commonly used diffusion MRI (dMRI) technique, has been used extensively in attempts to identify pathological changes in WM tracts following mTBI in humans and animals [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e]. These studies failed to reveal consistent results in terms of which WM tracts show dMRI changes after injury. For example, some studies have reported dMRI changes in the corpus callosum [\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e, \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e87\u003c/span\u003e, \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e89\u003c/span\u003e], while others highlight changes in the superior longitudinal fasciculus[\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e] and internal capsule [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. It is likely that that these discrepancies arise from the heterogeneity in the orientation of rotational forces being studied. Thus, dMRI studies examining mTBI might be improved by analyzing the data in the context of the axis of rotation of a given injury. Additionally, DTI is limited by its assumption of Gaussian diffusion in complex microenvironments where changes such as axonal damage, myelin damage, and inflammatory processes may be occurring simultaneously as in the case of mTBI [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. Diffusional kurtosis imaging (DKI) allows better characterization of complex microenvironments by capturing non-Gaussian diffusion characteristics to provide a measure of tissue complexity[\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e] and may thus improve detection of microstructural changes following mTBI. Furthermore, frequency-dependent dMRI using oscillating gradient encoding allows characterization of microstructural changes at different spatial scales [\u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e77\u003c/span\u003e], giving insight into the size of structures contributing to dMRI changes. While frequency-dependent dMRI has been applied in various pathologies [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e97\u003c/span\u003e], its application to mTBI remains unexplored.\u003c/p\u003e\u003cp\u003eWhile dMRI provides insight into WM microstructural changes brought on by mTBI, resting-state fMRI (rs-fMRI) has the potential to reveal the functional consequences of WM injury. rs-fMRI has been extensively used to study mTBI in both humans [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e] and animals [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e, \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e80\u003c/span\u003e, \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e89\u003c/span\u003e, \u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e93\u003c/span\u003e, \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e99\u003c/span\u003e]. Similar to DTI studies, these rs-fMRI studies report a range of findings including increases and decreases in functional connectivity in different brain regions including: the somatosensory and somatomotor cortex [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e, \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e89\u003c/span\u003e], the thalamus and the hippocampus [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e, \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e, \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e, \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e89\u003c/span\u003e, \u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e93\u003c/span\u003e]. As in the case of the discrepancies in the results of dMRI studies in mTBI, we predict that the lack of agreement in rs-fMRI studies in mTBI lies in the failure to analyze the rs-fMRI data to focus on changes in connectivity between brain regions connected by tracts that are oriented orthogonal to the axis of rotation produced by the injury.\u003c/p\u003e\u003cp\u003eWe have recently reported on a novel method to generate mTBIs with high rotational forces to study concussion in mice [\u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e88\u003c/span\u003e]. In the present study, we predicted that pathological changes in the WM tracts would be more easily detected by DKI compared to DTI metrics, that WM tracts oriented orthogonal to the axis of rotation produced by this mTBI would be preferentially injured, and that brain areas connected by tracts running orthogonal to the injury would show functional disconnection by rs-fMRI. We herein report that this model of mTBI produces the greatest pathological changes in tracts that are oriented orthogonal to the axis of rotation of the mTBI. Furthermore, this orientation-dependence of pathological changes was evident in DKI but not DTI metrics. In agreement with the DKI results the rs-fMRI shows that the mTBI results in a loss of functional connectivity between structures oriented in the anterior-posterior axis.\u003c/p\u003e"},{"header":"2. Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eSubjects\u003c/h2\u003e\u003cp\u003e All animal procedures were approved by the Western University Animal Care Committee and were consistent with guidelines established by the Canadian Council on Animal Care. \u003cem\u003eIn vivo\u003c/em\u003e MRI data was collected from genetically modified mice homozygous for the \u003cem\u003eAPOE3\u003c/em\u003e allele carrying humanized wildtype \u003cem\u003eMAPT\u003c/em\u003e[\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e] and \u003cem\u003eAPP\u003c/em\u003e[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] (JAX stock #030898) under the control of the associated mouse promoter. Animals aged 6 months were randomly assigned to either sham (n = 15, 7 males 8 females) or mTBI groups (n = 16, 8 males and females).\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eRepetitive mTBI Procedure\u003c/h3\u003e\n\u003cp\u003eThe injury procedure has been reported in detail in Teasell et al. [\u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e88\u003c/span\u003e]. Briefly, animals were anesthetized by intra-peritoneum injection with Ketamine 80mg/kg and Xylazine 10mg/kg; then placed on a custom-made acrylic box, topped with a piece of pre-pierced clear plastic and positioned under a traumatic brain injury device (TBI 0310, Precision Systems and Instrumentation, LLC). The custom-made, 5mm-diameter, pliant, silicone tip was placed aligning with the bregma. The device was programmed to deliver an impact at an intended depth 8.0 mm at a 3.5 m/s velocity with a 500-millisecond dwell time. Following the impact, the mouse broke through the plastic and underwent a 180° sagittal rotation (rotation about the right-left axis) and landed on the cushioned bottom of the box. The animal was then transferred back to its cage and placed under a heating lamp until regaining consciousness. The sham animals received the corresponding doses of ketamine/xylazine, but without the impact. These procedures were repeated 3 times, with a 24-hour interval. All experimental animals regained consciousness within 5–10 minutes post corresponding procedures, with no obvious motor deficit. No skull fracture, hematoma, apnea or death associated with sham or mTBI mice was observed.\u003c/p\u003e\n\u003ch3\u003eMRI Acquisition\u003c/h3\u003e\n\u003cp\u003eBefore scanning sessions, anesthesia was induced by placing mice in an induction chamber with 4% isoflurane in 100% oxygen with a flow rate of 1 L/min. Throughout the scanning session, isoflurane was maintained at 1.8% in 100% oxygen with a flow rate of 1 L/min through a custom-built nose cone. In \u003cem\u003evivo\u003c/em\u003e MR scanning sessions were performed on a 9.4 T Bruker Neo small animal scanner (Agilent, Palo Alto, CA, USA) equipped with a 6 cm gradient coil insert of 1 T/m strength, Bruker Avance III HD console with software package of Paravision-360.3.3 (Bruker BioSpin Corp, Billerica, MA), and a single-loop surface coil of 2x1 cm\u003csup\u003e2\u003c/sup\u003e. All subjects were scanned at baseline at 6 months of age and then again 1-week post-last impact. Anatomical, diffusion, and functional data were acquired for all scanning sessions in a scan time of 2 hours 12 minutes.\u003c/p\u003e\u003cp\u003eThe dMRI protocol included a pulsed gradient spin echo (PGSE) sequence (i.e., 0 Hz) with gradient duration of 9.4 ms (diffusion time = 12.3 ms) and oscillating gradient (OGSE) sequences with frequencies of 60 and 120 Hz (corresponding effective diffusion times of 2.3 and 1.2 ms [\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e]). The 0, 60, and 120 Hz acquisitions correspond to average molecular displacements of 8.6, 3.7, and 2.7 µm, respectively, according to the Einstein-Smoluchowski relation for hindered diffusion [\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e]. The 60 Hz acquisition implements frequency-tuned bipolar waveforms to reduce the TE of the acquisition [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. For all frequencies, data consisted of 2 b = 0 volumes and 2 b-value shells of 1,000 and 2,500 s/mm\u003csup\u003e2\u003c/sup\u003e each with an efficient 10-direction scheme [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. The dMRI data was acquired in one integrated scan using single-shot echo planar imaging (EPI) with 80% of k-space being sampled in the phase encode direction and parameters: TE/TR = 35.5/15000 ms, FOV = 19.2 x 14.4 x 15 mm\u003csup\u003e3\u003c/sup\u003e, in-plane resolution 200 x 200 µm\u003csup\u003e2\u003c/sup\u003e, slice thickness 500 µm, 4 repetitions, scan time of 66 minutes.\u003c/p\u003e\u003cp\u003eT2*-weighted images for fMRI were acquired in 30 coronal slices using a gradient-echo EPI (GE-EPI) sequence covering the whole brain with the following parameters: TE/TR of 12/1500 ms, flip angle of 60°, FOV of 19.2 × 9.6 mm\u003csup\u003e2\u003c/sup\u003e, matrix size (MS) of 64 x 32, and slice thickness of 0.5 mm to produce a voxel size of 300 × 300 × 500 µm\u003csup\u003e3\u003c/sup\u003e. Four runs of 600 volumes were acquired consecutively for a total of 60 minutes. Two runs were acquired with a blip up and the other two were acquired with blip down to allow for distortion correction (details below). An additional T2-weighted anatomical image was acquired in the same space using a turbo rapid acquisition with relaxation enhancement (TurboRARE) sequence with the following parameters: TE/TR of 30/5500 ms, 8 averages, FOV of 19.2 × 9.6 mm\u003csup\u003e2\u003c/sup\u003e, MS of 128 × 64, a slice thickness of 0.5 mm to yield a voxel size of 150 × 150 × 500 µm\u003csup\u003e3\u003c/sup\u003e. The acquisition duration was 352 seconds.\u003c/p\u003e\n\u003ch3\u003eDiffusion MRI\u003c/h3\u003e\n\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003ePreprocessing\u003c/h2\u003e\u003cp\u003eComplex-valued repetitions underwent partial Fourier reconstruction using POCS [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e], frequency and signal drift correction, and phase alignment, similar to previous work [\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e]. The DESIGNER pipeline [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] was used to perform MP-PCA tensor denoising [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e, \u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e91\u003c/span\u003e] with Rician bias correction [\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e], followed by Gibbs ringing correction for partial Fourier acquisitions [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e].\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003eFitting\u003c/h2\u003e\u003cp\u003eDiffusivity (MD: mean diffusivity, RD: radial diffusivity, AD: axial diffusivity) and kurtosis (MKT: mean kurtosis tensor, RK: radial kurtosis, AK: axial kurtosis) maps were computed at each frequency using axisymmetric modelling and spatial regularization, as outlined in Hamilton et al. [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. Regularization weighting was heuristically chosen to minimize the visual appearance of noise while retaining original image contrast and was held constant across subjects.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eQuality Assurance\u003c/h3\u003e\n\u003cp\u003eThe signal-to-noise floor ratio (SNR) of all scans was measured as the signal mean in an ROI placed in the cortex across b = 2,500 s/mm\u003csup\u003e2\u003c/sup\u003e acquisitions before denoising, divided by the signal mean in an ROI outside the brain. One male in the mTBI cohort and one female in the sham cohort did not pass our dMRI quality assurance threshold due to low SNR (\u0026lt; 2.5, 4.5 median for all subjects) of baseline scans and qualitative noise contamination in diffusion-weighted volumes and parameter maps and thus were excluded from subsequent dMRI analysis.\u003c/p\u003e\n\u003ch3\u003eTractography\u003c/h3\u003e\n\u003cp\u003eFor all subjects baseline scans, deterministic whole brain tractography was performed using ExploreDTI [\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e] with parameters: Fiber length range 3–12 mm, angle threshold of 50 degrees, step size 0.1 mm, fractional anisotropy (FA) threshold 0.15. Data from all PGSE and OGSE acquisitions was used as an input for tractography to provide a more robust result as the principal eigenvector should not vary appreciably across the explored frequencies [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. For each subject, the post-injury scan was registered to baseline scan using affine and symmetric diffeomorphic transforms with ANTs software [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. This was done as opposed to performing tractography for each imaging session to reduce noise and ensure dMRI metric changes from possible tract degeneration were captured.\u003c/p\u003e\u003cp\u003eIn each tract, the mean principal eigenvector was used to separate tracts into groups according to their orientation relative to the axis of rotation (right-left axis). The central angle between each tract and the rotation axis was determined by computing the dot product of the mean principal eigenvector with the right-left axis. Tract groups of 0–50, 50–60, 60–70, 70–80, and 80–90º were chosen based on the spherical sector occupied by each bin and to keep the number of tracts in each group approximately equal. For each tract group, a tract density map was computed which indicated voxels with tracts passing through them. This tract density map was binarized and used to compute the mean of dMRI metrics at each frequency within tract bins for each subject and timepoint. Voxels with MD \u0026gt; 1.4 µm\u003csup\u003e2\u003c/sup\u003e/ms were excluded from binarized tract density maps to mitigate partial volume effects with cerebrospinal fluid.\u003c/p\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003eFunctional MRI\u003c/h2\u003e\u003cdiv id=\"Sec12\" class=\"Section3\"\u003e\u003ch2\u003ePreprocessing\u003c/h2\u003e\u003cp\u003eThe analysis was done using multiple tools from FSL 6.0.7 [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e], AFNI [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e], ANTs [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], ITKSnap [\u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003e100\u003c/span\u003e], and nilearn (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/nilearn/nilearn\u003c/span\u003e\u003cspan address=\"https://github.com/nilearn/nilearn\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). All the processing workflows were implemented using the Nipype Python library [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. The T2-weighted images were used to construct a study-based template and the skull was removed from the template manually using ITKSnap. The anatomical images from each subject were bias field corrected, registered to skull-stripped template and the inverse transformations applied to the template mask were used to automatically remove the skull from individual subjects. The extracted brains were then registered to study-based template brain.\u003c/p\u003e\u003cp\u003eThe functional images were corrected for motion using rigid-body registration of 6 degrees of freedom. The images were corrected for field distortions using FSL topup toolbox [\u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e82\u003c/span\u003e] taking advantage of the opposite phase-encoding acquisition. The middle volume of each subject was registered using rigid body transformations to the study-template brain. The inverse transformations applied to the template mask and the resulting masks were used to extract the brain from individual runs. Functional to anatomical co-registration transformations were calculated using affine registration. The functional images were then high-pass filtered using a 0.01 Hz to remove the low-frequency noise and to keep the signal associated with resting-state networks (\u0026gt; 0.01 Hz) and motion parameters were regressed out of the data. The cleaned images were then smoothed using an isotropic Gaussian kernel of 0.4 mm\u003csup\u003e3\u003c/sup\u003e. The transformations from individual anatomical images to the study-based template and from the functional to the anatomical images were combined and applied to the smoothed functional images to bring them into the study-based template space. Images in study-based template were used for subsequent connectivity analyses. The study-based template was nonlinearly registered to the anatomical image of the Allen Brain Atlas common coordinate framework version 3 (ABA-CCFv3) [\u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e94\u003c/span\u003e]. The inverse transformations were used to bring the annotation images from Allen Brain Atlas to the study-based template for functional connectivity (FC) atlas-based analysis.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003eAnnotation atlases creation\u003c/h2\u003e\u003cp\u003eABA-CCFv3 annotation atlas was used to construct: a region atlas of 22 ROIs (Fig.\u0026nbsp;4a) (11 per hemisphere) and a more refined subregions atlas with 82 ROIS (SI Appendix, Fig. S4a) (41 per hemisphere) by combining adjacent structures.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003eAtlas-based FC\u003c/h2\u003e\u003cp\u003eFor each preprocessed run, the average timeseries within each label of the annotation atlas was computed, Pearson’s correlation coefficient between each pair of labels was calculated using nilearn and later converted to \u003cem\u003ez\u003c/em\u003e-score using Fisher’s transformation. The resulting correlation matrices were averaged across runs and later used for statistical inferences.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003eIndependent component analysis (ICA)\u003c/h2\u003e\u003cp\u003eFor group-level ICA analysis, the preprocessed 4D images of all subjects were concatenated and group ICA analysis was run using MELODIC tool in FSL [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. The concatenated data was decomposed into 10, 15, 20, 25, 30, 40, and 50 dimensions. The 20 dimensions decomposition gave the best representation of the resting-state networks [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e101\u003c/span\u003e]. We excluded 1 component that overlapped with major ventricles and the remaining 19 components were used in further analysis. We then ran the Dual Regression (DR) analysis and used the 2nd stage maps in assessing the within-component connectivity [\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e]. To assess between-component connectivity strengths, the ICA components were thresholded by 3 and used as a probabilistic atlas. For visualization purposes, the study-based template was non-linearly registered to the ABA-CCFv3 anatomical atlas and the transformations were used to bring the ICA networks and DR statistical maps into the atlas space.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\u003ch2\u003eTract-based FC\u003c/h2\u003e\u003cp\u003eViral tracer data from Allen Brain’s Institute Mouse connectivity data [\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e] was used to extract tracts identified in tracts orientation bins from dMRI analysis. Each experiment was chosen such that the tract has the highest injection volume among the white matter fiber tracts. The data for each experiment was downloaded, converted to NIfTI format, and the log10 transformed data was thresholded to 10\u003csup\u003e− 3.5\u003c/sup\u003e for better false positive control [\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e]. The source and target structures were determined based on structures with the highest injection volumes known to be interconnected by that tract (SI Appendix, Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). For instance, the corticospinal tract (CST) is known to connect the motor cortex to more posterior regions and finally to the spinal cord. A representative experiment was chosen with the primary motor cortex as the primary injection location and the pons was chosen as a target as it has the highest injection volume among structures known to be connected to the motor cortex through the CST.\u003c/p\u003e\u003cp\u003eABA CCFv3 annotation atlas was used to obtain binary masks of the source and target regions and these masks were multiplied by the binarized injection volume map to obtain regions between which correlation strength can be calculated. The timeseries within the source and target regions were averaged and correlation coefficient was calculated and transformed to Fisher’s z-score using AFNI’s 3dNetCorr. For targets that are ipsilateral to the injection volume, the source and target on both hemispheres were averaged, while for those tracts that traverse the brain midline only lateral regions were used.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\u003ch2\u003eHistology\u003c/h2\u003e\u003cdiv id=\"Sec18\" class=\"Section3\"\u003e\u003ch2\u003eImmunohistochemistry and Silver Staining\u003c/h2\u003e\u003cp\u003eAt one week after the last impact, 12 mice (6 injured and 6 shams) were anesthetized with ketamine/Xylazine (2:1), then underwent trans-cardiac perfusion with ice-old saline, followed by 4% Paraformaldehyde in PBS. Brains were post-fixed overnight in 4% paraformaldehyde, then placed in 15% sucrose (Cat# S5-3, Fisher Scientific) in PBS for 6–12 hours and then 30% sucrose in PBS overnight and embedded in optimal cutting temperature (OCT) medium.\u003c/p\u003e\u003cp\u003eFor silver staining, fifteen floating coronal cryostat sections (50 µm) were collected at the level of the corpus callosum (bregma + 1/-1mm) and optic tracts (bregma − 1/-2mm), respectively. The staining was performed using the FD NeuroSilver Kit II (FD NeuroTechnologies, Ellicott City, MD) according to the manufacturer’s instructions.\u003c/p\u003e\u003cp\u003eFor immunohistochemistry, the rest of the brain was cryosectioned at 16 µm and collected serially onto Superfrost Plus slides. Cryosections were rinsed in 0.1M phosphate-buffer saline and blocked in 5% goat serum with 0.1% Triton X-100 for 2h at room temperature. Then the sections were incubated with primary antibodies: anti-GFAP antibody (Cat# G3893; Sigma, Germany) and anti-Iba1 antibody (Cat# 019-19741, WAKO Japan) at 4°C overnight, followed by 1h incubation with fluorescent secondary antibodies: donkey anti-mouse IgG, conjugated with Alexa Fluor 488 (1:1000; Invitrogen, Rockford IL, USA) and goat anti-rabbit IgG, conjugated with Alexa Fluro 594 (1:1000; Invitrogen). Digital images were captured using a Leica MICA or Leica Stellaris 5 confocal microscope.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\u003ch2\u003eElectron microscopy\u003c/h2\u003e\u003cp\u003eFour injured mice and 2 sham mice were processed for EM, as described previously [\u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e98\u003c/span\u003e]. Briefly, mice were perfused with 0.1M phosphate buffer, followed by 4% paraformaldehyde supplemented with 2% glutaraldehyde. Vibratome sections (100um) were collected that included corpus callosum regions and internal capsule regions, respectively (ROIs). The specimens were then post-fixated with 1% Osmium Tetroxide in 0.1M Cacodylate buffer for 1 hr followed by overnight en-bloc staining with 1% Uranyl Acetate (at T = 4 ⁰C). Brain tissues were first rinsed with double-distilled water and then dehydrated in an ascending series of ethanol solutions and embedded in Spurr’s resin between two Aclar films at 60°C for 2–3 days for polymerization. After polymerization, ROIs were identified and isolated using a stereomicroscope [\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e]. A Leica UCT ultramicrotome with a Diatome diamond knife was used to obtain ultrathin serial sections (60–70 nm) from the ROIs. Serial sections were collected onto pioloform-coated copper slot grids. Electron microscopy examination was carried using the TEM at the Biotron facility at Western University.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e\u003ch2\u003eStatistics\u003c/h2\u003e\u003cdiv id=\"Sec21\" class=\"Section3\"\u003e\u003ch2\u003eDiffusion MRI\u003c/h2\u003e\u003cp\u003eSex differences were evaluated within sham and mTBI groups using one-way ANOVA models on metric changes post-injury relative to baseline for each group across each metric and OGSE frequency. After separating data by sex, a one-way ANOVA was used to examine group effects (sham vs. mTBI) for each metric and frequency. To control for multiple comparisons, false discovery rate (FDR) correction using the Benjamini-Hochberg procedure was applied to p-values obtained from ANOVA analyses of sex and injury effects. For metrics showing significant effects of injury, Sidak post hoc tests were conducted to identify specific differences between mTBI and sham cohorts within each tract bin. All statistical analysis was performed using R Statistical Software version 4.1.2 and GraphPad Prism version 10.2.0.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec22\" class=\"Section2\"\u003e\u003ch2\u003eFunctional MRI\u003c/h2\u003e\u003cp\u003eFor atlas-based and ICA-based FC statistical analysis, the average connectivity matrix from each subject was combined into a 4D matrix for each group and various comparisons were conducted using permutation testing as implemented in FSL permutation analysis of linear models (PALM) tool [\u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e95\u003c/span\u003e]. Sex differences were tested using unpaired permutation testing conducted separately within each group and condition (males vs females before and after in mTBI and sham groups). Left vs right hemispheres were compared using paired permutation testing within each group and condition. Similarly, changes before and after the injury or the sham procedure were tested using within each group and condition.\u003c/p\u003e\u003cp\u003eFor all the permutation tests, null distributions were generated using 10,000 permutations and exchangeability blocks were defined at the subject-level to account for the paired nature of the data in cases of paired comparisons. All results were corrected for multiple comparisons using FDR.\u003c/p\u003e\u003cp\u003eThe voxel-wise statistical analysis for the 2nd stage maps of the DR were compared using paired permutation testing with 10,000 permutations, threshold-free cluster enhancement (TFCE), and family-wise error rate (FWER) correction across voxels, components, and contrasts.\u003c/p\u003e\u003cp\u003eFor tractography-based FC, the connectivity strength between the source and target structures were compared using paired Student’s t-test for each tract separately.\u003c/p\u003e"},{"header":"3. Results","content":"\u003ch2\u003eRepresentative Maps\u003c/h2\u003e\u003cp\u003eTo enable analyses of dMRI and rs-fMRI data with respect to tract orientation we began by validating representative structural, functional, and diffusion MR images from mice at baseline (Fig.\u0026nbsp;1a-c). MD and MKT both show qualitative frequency-dependence with increased diffusivity and decreased kurtosis, as expected [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. To examine the effect of tract orientation relative to the axis of rotation produced by the injury, tracts generated from tractography were separated into 5 bins based on their orientation relative the right-left axis of rotation (0º-50º, 50º-60º, 60º-70º, 70º-80º, 80º-90º) (Fig.\u0026nbsp;1d). In each bin, tract density maps were computed and used to evaluate dMRI metric changes post-injury (Fig.\u0026nbsp;1e).\u003c/p\u003e\u003ch2\u003eSex differences in mTBI mice detected by dMRI\u003c/h2\u003e\u003cp\u003eMice in the mTBI group (n = 8 males and females) underwent 3 mTBIs (one a day for 3 days) as described [\u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e88\u003c/span\u003e]. Mice in the sham group (n = 7 males, 8 females) underwent 3 sham procedures (anesthesia without injury). Mice were imaged at baseline and 1-week post-injury and dMRI metrics computed for each tract bin relative to the axis of rotation as delineated in Fig.\u0026nbsp;1. Data from 1 male mTBI and 1 female sham mouse were excluded from further dMRI analysis due to low quality data which did not pass our quality assurance threshold (detailed in Methods). To detect possible sex differences, all dMRI metrics were compared between male and female mTBI mice and between male and female shams. In mTBI mice a significant effect of sex was detected for MD, RD, AD, and MKT at all frequencies, and RK at 60 and 120 Hz (\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05) (Fig.\u0026nbsp;2). Sex differences in metric changes were relatively consistent across acquisition frequencies and tract orientations, with mTBI females having larger decreases in diffusivity and kurtosis metrics. In the sham mice, a significant effect of sex was detected for MKT and AK at 60 and 120 Hz (\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05), with males showing increased kurtosis relative to females following the sham procedure (Fig.\u0026nbsp;2 insets).\u003c/p\u003e\u003ch2\u003eDiffusional kurtosis detects tract orientation dependence of damage following mTBI in females\u003c/h2\u003e\u003cp\u003eGiven the sex differences detected, we analyzed males and females separately in both sham and mTBI groups to evaluate dMRI changes due to injury. For kurtosis metrics (Fig.\u0026nbsp;3a) in females, a significant group effect (injured versus sham) was found at 60 Hz for MKT (\u003cem\u003eF\u003c/em\u003e = 6.145, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05) and RK (\u003cem\u003eF\u003c/em\u003e = 7.896, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05). Importantly, MKT and RK reveal a \u003cem\u003etract orientation dependence\u003c/em\u003e with larger decreases in the mTBI mice in tracts oriented orthogonal to the left-to-right axis of rotation generated by the injury, with significant post-hoc findings in the 80–90º tract bin for MKT and RK (\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05) at 60 Hz. Diffusivity metrics (Fig.\u0026nbsp;3b) showed significant group effects in MD and RD at all frequencies and AD at 0 Hz (\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05). However, diffusivity metrics did not show the same tract orientation dependence as the kurtosis metrics, with larger diffusivity decreases in the mTBI cohort being consistent across all tract orientations. No significant effects of group were found for any dMRI metric in males (SI Appendix, Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e\u003ch2\u003eParcellating the brain into regions and subregions annotation atlases\u003c/h2\u003e\u003cp\u003eTo minimize registration errors, reduce artifact due to motion, and reduce the number of statistical comparisons required, the ABA-CCFv3 annotation atlas was down-sampled to create a region-specific atlas (Fig.\u0026nbsp;4a) consisting of 22 anatomical regions (11 per hemisphere). A finer subregion-specific atlas (SI Appendix, Fig. S4a) consisting of 82 subregions (41 per hemisphere) was generated to enable a more detailed FC analysis. The region-specific atlas was created by combining together adjacent structures. For example, the gustatory, the anterior cingulate, and the somatomotor areas of the isocortex (to name a few subregions) were considered a single structure in the region-specific atlas and designated collectively as the isocortex. In the creation of the subregion-specific atlas, we preserved one level of hierarchy such that the isocortex was divided into its component parts.\u003c/p\u003e\u003ch2\u003eNo sex differences in mTBI mice detected by fMRI\u003c/h2\u003e\u003cp\u003eTo test for sex differences in rs-fMRI, we compared the FC between males and females in sham and mTBI groups at baseline and again post-injury or post sham procedure. FC comparisons between males and females failed to reveal any differences that survived multiple comparisons correction for the region-specific or the subregion-specific atlases (SI Appendix, Fig. S2 \u0026amp; S3). Thus, the FC analyses were performed on 4 groups of mice with sexes combined: sham at baseline, shams at 1 week post sham procedure, mTBI mice at baseline and mTBI mice 1 week post-injury.\u003c/p\u003e\u003ch2\u003emTBI induces wide decrease in FC across the brain\u003c/h2\u003e\u003cp\u003eThe rs-fMRI data was analyzed to uncover differences in FC within the sham group (comparing FC at baseline to FC at 1 week post-sham procedure), within the mTBI group (comparing FC at baseline to FC at 1 week post-mTBI) and between the sham and mTBI groups (comparing FC at baseline to FC at 1 week post-sham procedure or mTBI). After correcting for multiple comparisons, statistically significant reductions in FC were observed in the mTBI group 1 week after the injury compared to FC in both sham groups and to the FC in the mTBI baseline group. As expected, there were no statistically significant differences in FC between the sham groups at baseline and at 1 week post-sham procedure and between the shams at baseline and the mTBI mice at baseline (Fig.\u0026nbsp;4b and 4c and SI Appendix, Sig. S4a and 4b).\u003c/p\u003e\u003ch3\u003eRegion-specific changes in FC\u003c/h3\u003e\u003cp\u003eOne week after the injury, the mTBI group showed a widespread decrease in FC compared to their baseline values (Fig.\u0026nbsp;4d). The largest reductions in FC were between regions in the frontal, temporal and posterior regions of the brain that would rely on long-range connections that run from frontal and temporal regions to the posterior regions (Fig.\u0026nbsp;4e). The FC between anterior regions such as the olfactory region and posterior regions (the hindbrain and cerebellum) showed the most significant decreases in FC (Fig.\u0026nbsp;4e) with a more than 50% decrease in FC. FC between the isocortex and more posterior regions (the midbrain, hindbrain, and cerebellum) were also decreased after mTBI (Fig.\u0026nbsp;4d). Temporal regions such as the hippocampal formation showed decreases in FC with the isocortex, hindbrain and cerebellum.\u003c/p\u003e\u003ch2\u003eSubregion-specific changes in FC\u003c/h2\u003e\u003cp\u003eFC analyses were also carried out using the subregion-specific atlas to evaluate if region-specific FC changes could be attributed to FC changes between component subregions and to determine if a finer FC subregions analysis could uncover changes in FC hidden by the courser region-specific analyses. This subregion-specific analysis revealed that the decrease in FC between the isocortex and more posterior regions could be specifically attributed to a decrease in FC between the frontal pole, the somatosensory/somatomotor cortex, the anterior cingulate cortex, the agranular insular cortex, and temporal association areas with regions of the midbrain, hindbrain and cerebellum (SI Appendix, Fig. S4). The subregion analyses further identified that decreases in the FC of the Ammon’s horn (CA) and subiculum subregions of the hippocampus accounted for the decreased FC between the hippocampal formation and hindbrain/cerebellum shown in the region-specific analysis (SI Appendix, Fig. S4). In some cases, the region-specific analyses detected changes in FC that were not found between the component subregions. For example FC changes between left olfactory areas and left cerebellum, right isocortex and left midbrain, and right hippocampus and left midbrain were not present in the subregion-specific analyses. On the other hand the FC using the subregion-specific analysis revealed a significant decrease in FC between the amygdala and the vermal regions of the cerebellum that was not appreciated in the FC using the region-specific atlas in which all cerebellum substructures were aggregated into one cerebellar region (Fig.\u0026nbsp;4a). Statistically significant changes in FC between subregions are represented by connecting lines in the circular plot in SI Appendix, Fig. S4e with the thickness of the lines reflecting the percentage of decrease in FC after the mTBI.\u003c/p\u003e\u003ch2\u003eFC changes after mTBI show right-left asymmetries\u003c/h2\u003e\u003cp\u003eComparing changes in FC in mTBI mice 1 week after injury to their baseline values revealed right-left asymmetries that were more evident in the region-specific atlas than in the subregion-specific atlas (Fig.\u0026nbsp;4e and SI Appendix, Fig. S4e). The right hemisphere showed decreases in FC between the isocortex and its targets in the hippocampus, hindbrain and cerebellum and between the hippocampus and its targets in the isocortex, hindbrain and cerebellum. In the left hemisphere, only left olfactory regions showed decreases in FC with the hindbrain, and the cerebellum. Analyses of FC in the right hemisphere of sham mice also identified decreases in FC between the hippocampus and the isocortex, indicating that these FC asymmetries were reflective of normal FC asymmetries in mice and are unrelated to the mTBI (SI Appendix, Fig. S5 \u0026amp; S6).\u003c/p\u003e\u003ch2\u003eIdentification of resting-state networks\u003c/h2\u003e\u003cp\u003eThe data from all subjects in both groups was aggregated together and decomposed into 20 resting-state networks (RSN), one component was disregarded as noise due to peak activation overlaps with major ventricles. The most common networks previously reported in the literature were identified in our data (SI Appendix, Fig. S7). We identified cortical networks representing the olfactory areas, somatomotor regions, anterior regions of the default mode network (DMN) including the prelimbic areas, the anterior cingulate area, and the orbital area. Subcortical networks were also identified such as the amygdala, the hippocampus, the striatum and the hypothalamus. Some networks showed laterality such as somatomotor, amygdala, striatum, hypothalamus, midbrain, and hindbrain networks.\u003c/p\u003e\u003ch2\u003emTBI causes FC changes within resting-state networks\u003c/h2\u003e\u003cp\u003eDR analysis showed two affected networks with diminished connectivity in the mTBI group after the injury. The first was the network consisting of regions of the ventral striatum, pallidum, and hypothalamus that showed clusters of decreased FC in the taenia tecta region of the olfactory areas (Fig.\u0026nbsp;5d, top). The second network affected by mTBI was the right somatomotor network that showed clusters of lower FC in the caudate putamen region of the striatum (Fig.\u0026nbsp;5d, bottom). Three resting-state networks showed an increase in FC in the control group following the sham procedure including, the hypothalamus network, the right amygdala network, and the right somatomotor network (SI Appendix, Fig. S8) that showed decreased FC in the mTBI group.\u003c/p\u003e\u003ch3\u003emTBI causes FC changes between resting-state networks\u003c/h3\u003e\u003cp\u003eTop-bottom correlation between the somatomotor network and more ventral regions in the olfactory and ventral striatum networks and between the vermal region of the cerebellar network and the midbrain network showed significant decreases in FC following mTBI (Fig.\u0026nbsp;5c). In comparison to the sham group (Fig.\u0026nbsp;5a), FC between the hippocampal network and midbrain and the cerebellar networks showed hypoconnectivity following mTBI (Fig.\u0026nbsp;5b and c).\u003c/p\u003e\u003cp\u003e\u003cspan type=\"ItalicUnderline\" class=\"ItalicUnderline\" name=\"Emphasis\"\u003eDecreases in FC between regions directly connected by WM tracts orthogonal to right-left axis of rotation\u003c/span\u003e\u003c/p\u003e\u003cp\u003eTo further test our prediction that brain areas connected by WM tracts that are oriented orthogonal to the right-left axis of rotation of the mTBI, we calculated the FC between regions that are known to be anatomically connected through WM tracts that run in the rostral-caudal direction as determined by viral tracing studies [\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e]. In contrast to atlas-based or network-based FC which cannot differentiate between direct and indirect connections, this approach guarantees that the FC of the regions analyzed are directly, anatomically connected. The FC served by the CST was the most-affected tract among all those tested including the cingulum bundle, the optic tract, and the cerebellar peduncle (SI Appendix, Fig. S9). FC between the motor cortex and the hindbrain that is served by the CST was decreased by 17% following mTBI (Fig.\u0026nbsp;6a and c). In contrast, the FC of regions connected by fibers that run parallel to the right-left axis including the genu of the corpus callosum (Fig.\u0026nbsp;6b and d), the stria terminalis, and the arbor vitae (SI Appendix, Fig. S10) were unchanged.\u003c/p\u003e\u003ch3\u003eHistological Findings\u003c/h3\u003e\u003cp\u003eThe imaging studies above suggested that pathological changes due to mTBI predominate in WM tracts oriented orthogonal to the right-left axis of rotation of the injury. To validate this finding we conducted silver staining, immunohistochemistry, and electron microscopy in sham and mTBI mice in WM tracts that oriented 0º-50º (the body of the corpus callosum) and 80º-90º (the cingulum bundle and the optic tract for silver staining and immunohistochemistry, internal capsule for electron microscopy) to the axis of rotation produced by the mTBI (Fig.\u0026nbsp;7). Silver staining was used to assess axonal pathology at 1 week post mTBI (Fig.\u0026nbsp;7a-f). Very few silver-stained fibers were observed in any WM tracts in any of the 6 sham mice. In mTBI mice at 1 week post-injury, only a few silver-stained fibers were noted in the body of the corpus callosum. However, WM tracts that run orthogonal to the right-left axis of rotation produced by the mTBI, such as the sagittal fibers of the cingulum bundle and of the optic tract, showed prominent and widespread silver-stained fibers that were easily detected in all mTBI brains.\u003c/p\u003e\u003cp\u003eImmunohistochemistry was used to evaluate these same WM tracts for evidence of localized inflammation and glial reactivity. Representative immunostaining to identify astrocytes (using GFAP immunoreactivity) and microglia (using Iba-1 immunoreactivity) is shown in Fig.\u0026nbsp;7g-l. As expected, shams demonstrated some GFAP + astrocytes in all tracts with a small number of Iba1 + microglia. The microglial morphology in shams was predominantly that of resting microglia with many ramified branched processes. In the mTBI mice GFAP and Iba1 staining was more prevalent and the microglial morphology was more rounded with fewer processes suggesting a more activated phenotype.\u003c/p\u003e\u003cp\u003eElectron microscopic examination of several brain regions identified based on abnormal silver staining revealed extensive ultrastructure changes in the corpus callosum and internal capsule in the mouse brains with mTBI. Multiple stages of degenerating axons were observed (Fig.\u0026nbsp;7n): axon demyelination; excessive myelination and axon-myelin dissociation (vacuoles formed between myelin sheath and axon). The dystrophic axons were scattered among many normal-appearing axons. These abnormalities were absent in sham samples (Fig.\u0026nbsp;7m). An example of a reactive astrocyte is shown in Fig.\u0026nbsp;7o that demonstrates an astrocyte protrusion was adjacent to a degenerative neurite that contained cytoskeletal fragments, a typical glial response to axonal injury.\u003c/p\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eIn this work we investigated the use of dMRI and fMRI to detect microstructural and FC changes in mice 1 week following repeated mTBI. Specifically, we evaluated how the orientation of the axis of rotation produced by the mTBI influences the ability of dMRI and fMRI to detect pathological changes in the brain. Several findings were noted and are discussed below including: (1) that sex differences following mTBI in mice could be detected by dMRI but not by fMRI, (2) that while both diffusivity and kurtosis metrics could distinguish mTBI mice from shams, kurtosis was more sensitive to tract-specific injury and (3) that, after mTBI, diffusional kurtosis and fMRI detect pathological microstructural changes and associated reduced functional connectivity between brain regions connected by tracts oriented orthogonal to the axis of injury produced by the mTBI.\u003c/p\u003e\u003cdiv id=\"Sec38\" class=\"Section2\"\u003e\u003ch2\u003eSex differences following mTBI detected with dMRI but not fMRI\u003c/h2\u003e\u003cp\u003eSex differences in dMRI metric changes post-injury relative to baseline were assessed in sham and mTBI groups. In sham mice, sex differences were only detected in MKT and AK at 60 and 120 Hz, with no changes in diffusivity metrics. However, in the mTBI mice, diffusivity and kurtosis metrics in females were significantly reduced compared to males. Compared to male mTBI mice, female mTBI mice exhibited greater reductions in MD, RD, AD, and MKT across all acquisition frequencies and RK at 60 and 120 Hz. These dMRI findings are consistent with a greater vulnerability to mTBI-induced damage in female mice. In human dMRI studies of mTBI, sex differences are inconsistent across studies, with some studies reporting greater changes in males [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e96\u003c/span\u003e] and others in females [\u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e83\u003c/span\u003e]. These discrepancies may be influenced by variability in injury severity between the sexes in these studies. Similarly, preclinical animal studies yield mixed results [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e], although histological studies often report greater WM damage in females relative to males [\u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e84\u003c/span\u003e, \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e90\u003c/span\u003e]. A proposed biological mechanism for these sex-dependent differences suggests that females possess a higher proportion of small caliber axons that are more vulnerable to rotational injury [\u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e84\u003c/span\u003e], due to a sparser microtubule network [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Failure to detect sex-differences using rs-fMRI may be due to insufficient power to detect FC differences between mTBI males and females or due to less of a decrease in FC in mTBI females than their WM injury suggests. Compensatory changes in axonal firing rates, synaptic plasticity or other processes commonly seen in concussion [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e] could also reduce the sex-dependent impact of the microstructural injury on FC. It is also possible that FC is driven by the larger caliber axons in nerve tracts, reducing its sensitivity to damage in the higher fraction of smaller caliber axons found in females. While the correspondence between structural and functional connectivity is broadly accepted [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e, \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e81\u003c/span\u003e], rs-fMRI is more vulnerable to motion artifacts and FC can be contaminated by spurious correlations arising from vascular and respiratory artifacts [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e] that can obscure subtle sex differences. Additionally, sensitivity to injury-induced changes in the vascular tree may have obscured sex-specific connectivity changes detectable by fMRI that would not have interfered with the dMRI analyses.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec39\" class=\"Section2\"\u003e\u003ch2\u003eTract orientation dependence of damage detected with diffusional kurtosis\u003c/h2\u003e\u003cp\u003eThe mTBI females showed a significant effect of injury for MD and RD at all frequencies, AD at 0 Hz, and MKT and RK at 60 Hz. The reductions in MKT and RK showed a tract orientation dependence with larger decreases found in tracts oriented more orthogonal to the rotation axis, with both MKT and RK showing significant decreases following mTBI at 60 Hz in the 80\u0026ndash;90\u0026ordm; bin. By comparison the reductions in diffusivity metrics did not exhibit the same orientation dependence as MKT and RK.\u003c/p\u003e\u003cp\u003eOur histological findings show widespread mild inflammation and diffuse axonal injury in tracts oriented orthogonal to the rotation axis, with lesser amounts in tracts parallel to this axis, in addition to myelin disruption in mTBI mice. We propose that mTBI-induces tract-specific damage that is largely restricted to tracts oriented orthogonal to the axis of rotation of the injury and an inflammatory response characterized by swelling and gliosis that is more widespread. We hypothesize that the decreases in diffusivity metrics after mTBI are primarily driven by inflammatory changes that would be expected to be less spatially restricted than tract-specific damage and decrease diffusivity [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. This would explain the finding of diffusivity reductions across all tract orientations after mTBI. In contrast, changes in kurtosis metrics may be driven primarily by axon-specific microstructural damage explaining why decreases in kurtosis are restricted to tracts in the 80\u0026ndash;90\u0026ordm; bin. The assertion that changes in kurtosis may be due to axonal alterations after mTBI is supported by the finding that increased axon diameter and membrane permeability lead to decreased kurtosis [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. After an mTBI there may be a disproportionate loss of small caliber axons due to their greater vulnerability to mechanical deformation from rotational forces [\u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e84\u003c/span\u003e]. The disproportionate loss of small caliber axons after mTBI would lead to a post-injury increase in average axon diameter in a damaged tract and thus, a decrease in diffusional kurtosis. Alternatively, studies have shown that membrane permeability increases following mTBI leading to increased transmembrane water exchange and decreased kurtosis [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e, \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e, \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e85\u003c/span\u003e]. Our histological and ultrastructural studies show tract-specific axonal damage, and myelin abnormalities that would be expected to increase transmembrane water exchange and thereby decrease diffusional kurtosis. While it is difficult to disentangle these changes from the current study, it is likely that several ultrastructural abnormalities contribute to the observed changes in diffusional kurtosis. As female mTBI mice showed reductions in both diffusivity and kurtosis metrics but males did not we conclude, as have others [\u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e84\u003c/span\u003e, \u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e103\u003c/span\u003e], that females may experience more extensive axonal damage and greater inflammation than males following mTBI.\u003c/p\u003e\u003cdiv id=\"Sec40\" class=\"Section3\"\u003e\u003ch2\u003eFrequency-dependence of changes detected by dMRI\u003c/h2\u003e\u003cp\u003eOur findings revealed significant group differences in orthogonal tracts for both MKT and RK metrics at 60 Hz. Although data at 0 Hz exhibited similar trends, the results did not reach statistical significance (MKT: \u003cem\u003eF\u003c/em\u003e\u0026thinsp;=\u0026thinsp;5.038, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.050). In contrast, group differences appear less pronounced at 120 Hz. Previous studies have reported differential sensitivity between OGSE and PGSE acquisition in various models, including rodent models of ischemia [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e97\u003c/span\u003e], demyelination/inflammation [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e], as well as in human ischemic stroke [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. While microstructural changes at smaller spatial scales (i.e., axon level) would be expected to be more prominent at higher OGSE frequencies, our results suggest that a low non-zero frequency may be optimal to detect changes following mTBI when multiple microstructural changes are occurring simultaneously. Additionally, differences in waveform design\u0026mdash;specifically, the use of frequency-tuned waveforms at 60 Hz versus conventional OGSE waveforms at 120 Hz\u0026mdash;may yield different sensitivities to microstructural features.\u003c/p\u003e\u003cp\u003e\u003cspan type=\"ItalicUnderline\" class=\"ItalicUnderline\" name=\"Emphasis\"\u003eRegion-specific and subregion-specific FC analyses reveals complementary advantages\u003c/span\u003e\u003c/p\u003e\u003cp\u003eWe found a significant widespread decrease in FC in mTBI mice following the injury, that was not observed in the sham group. We analyzed FC using both a coarse, region-specific atlas and a fine sub-region specific atlas. Analyzing FC between coarsely parcellated brain regions revealed decreases in long-range FC between anterior and posterior regions. As FC between anterior and posterior brain regions must rely either on direct or indirect anatomical connections that are oriented in the anterior-posterior direction, the decrease in FC is consistent with our DKI data showing that tract-specific injury predominates in tracts oriented orthogonal to the right-left axis of rotation produced by the injury. Subregion-specific analyses were used to identify subregions responsible for the reductions in FC between brain regions and in at least one case also identified a decrease in FC between subregions (the amygdala and the vermal regions of the cerebellum) that were not identified in coarser regional analyses. The identification of FC changes between subregions but not between the larger regions that they lie in may be due to those FC changes being obscured by FC measures between the other component subregions making detection difficult. There were also some FC changes detectable by region-specific analyses that were not detected by sub-region analyses. This may have been because the FC changes by the region-specific analyses were due to the cumulative effect of multiple small changes in sub-region FC that failed individually to produce a significant change in FC. The reduced complexity of brain region-specific FC analyses allowed us to appreciate the spatial relationships between brain regions experiencing FC changes after mTBI such as L Olfactory-L cerebellum, R cortex-L midbrain, and R hippocampus-L midbrain because it entailed fewer region-to-region FC comparisons (231 vs 3321). The subregions specific analyses had the advantage of identifying specific subregions within the larger brain structures. Because our understanding of animal and human behavior is based on our understanding of the functions of subregions of the brain, the subregion-specific FC analysis may be useful when trying to predict the types of behavioral changes expected after mTBI. Predicting behavioral changes post-mTBI based on FC losses between brain regions would be difficult as those relatively larger regions of the brain subserve multiple functions.\u003c/p\u003e\u003cp\u003e\u003cspan type=\"ItalicUnderline\" class=\"ItalicUnderline\" name=\"Emphasis\"\u003eRight -left asymmetries in FC analyses\u003c/span\u003e\u003c/p\u003e\u003cp\u003eRight-left asymmetries were found in the FC analyses. In particular, decreases in FC between the isocortex and its targets in the hippocampus, hindbrain and cerebellum were predominantly right-sided. This may reflect right-left asymmetries in FC that naturally exist in mice that were revealed by injury. Many right-left FC asymmetries were found in FC of shams, consistent with the known naturally asymmetry in the mouse brain that has been reported in the literature [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e79\u003c/span\u003e].\u003c/p\u003e\u003cp\u003e\u003cspan type=\"ItalicUnderline\" class=\"ItalicUnderline\" name=\"Emphasis\"\u003eData-driven FC analyses\u003c/span\u003e\u003c/p\u003e\u003cp\u003eIn addition to the atlas-based analyses discussed above, we also carried out a data driven approach called ICA to decompose the brain into spatially independent networks[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e] and were able to identify the most-commonly reported brain networks [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e101\u003c/span\u003e]. Between-component analysis validated our observations of a loss of FC at 1 week after injury in networks containing the somatomotor region, olfactory regions, hippocampus, midbrain, and cerebellum with a notable degree of asymmetry as previously elaborated in the atlas-based analysis. As expected, the networks with reduced FC had significant contributions from brain regions in the anterior and posterior parts of the brain. Interestingly shams showed an increase in FC in the right somatomotor network 1-week post-sham procedure suggesting that the decreased FC in this network in mTBI mice is more significant than might otherwise be appreciated. Studies using the engineered rotational acceleration CHIMERA model of mTBI[\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e] that is similar to the model used in the present manuscript report similar decreases in FC at 7 days post-injury [\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e, \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e89\u003c/span\u003e]. Changes in FC following TBI are well-documented in both human (for a review see, [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e]) and animal literature [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e, \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e80\u003c/span\u003e, \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e89\u003c/span\u003e, \u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e93\u003c/span\u003e, \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e99\u003c/span\u003e], however, the direction and size of these FC changes varies widely between studies. Human literature reveals that changes in FC following head injury is very time-dependent. At 24\u0026ndash;72 hours following a head injury FC increases as a response to the injury [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e, \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e78\u003c/span\u003e, \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e104\u003c/span\u003e]. The acute increase of synaptic glutamate following a brain injury might explain the increase in FC as homeostatic plasticity restores the excitatory/inhibitory balance (reviewed in [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]). The acute, post-injury increase in FC is followed by a decrease in FC starting 4\u0026ndash;5 days after the injury and can last up to a few months [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e, \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e, \u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e92\u003c/span\u003e, \u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e102\u003c/span\u003e, \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e104\u003c/span\u003e]. At several months after the injury FC may increase once again as the injured brain remodels circuits to compensate for lost functions [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e].\u003c/p\u003e\u003cp\u003e\u003cspan type=\"ItalicUnderline\" class=\"ItalicUnderline\" name=\"Emphasis\"\u003eComparing direct FC across WM fibers that run orthogonal to plane of injury\u003c/span\u003e\u003c/p\u003e\u003cp\u003eThe region and subregion-specific FC analyses could not distinguish between direct (monosynaptic) and indirect (polysynaptic) connections. To directly evaluate FC changes served by tracts running orthogonal to the right-left axis of rotation, we evaluated FC changes between brain regions shown to be directly connected by viral tract-tracing studies. We found FC between regions connected by the CST to be the most affected by mTBI. Reductions in FC in other tracts (OT and cingulum bundle) oriented orthogonal to the right-left axis of rotation were less robust and not statistically significant. This might be explained by the greater length of the CST that would render it more vulnerable to damage. It is also possible that the use of smaller source and target regions to calculate FC for the other tracts made detecting statistically significant FC changes more difficult.\u003c/p\u003e\u003cp\u003e\u003cspan type=\"ItalicUnderline\" class=\"ItalicUnderline\" name=\"Emphasis\"\u003eConclusions: Implications for concussion imaging studies\u003c/span\u003e\u003c/p\u003e\u003cp\u003eSeveral conclusions may be made from this study: First, this study shows that after rotational mTBI, tracts orthogonal to the rotation axis experience the most damage in terms of microstructural changes detected with dMRI, which leads to loss of FC between regions connected (either directly or indirectly) by these tracts. It is notable that animal models of mTBI that include a rotational component about a right-left axis, as in the model of mTBI reported here, show similar patterns of injury. Groups using the CHIMERA, or momentum exchange model of mTBI, that cause head rotations around a right-left axis, report injury focused in the optic tract[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e] and external capsule[\u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e89\u003c/span\u003e] by FC and histological analyses [\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e]. On the other hand models of mTBI where the animals are struck on the side of the head producing an axis of rotation due to injury that is oriented in the anterior-posterior direction induces changes in the tracts that run right-left such as the corpus callosum as demonstrated by predominant FC changes in interhemispheric connectivity [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e, \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e]. Thus, the inconsistency in identifying regions and tracts affected by mTBI in the human literature may stem from the heterogeneity in the axes of rotation produced by different injuries in humans leading to patterns of injury that will be as varied as the injuries themselves. In animal studies of mTBI where the injuries are more consistent it may be that greater care in identifying the axis of rotation of a particular injury and analyzing the imaging data with the recognition that the injured tracts will be those oriented orthogonal to the axis of injury will improve the reproducibility and reliability of results. Second, our results suggest that, while both diffusivity and kurtosis metrics could identify damage in mice after mTBI, kurtosis metrics were more sensitive to tract-specific damage. Thus, while DKI requires a more extensive acquisition protocol, both DTI and DKI are useful in the detection of pathological changes following mTBI. Third we show that the use of frequency-dependent dMRI further increased sensitivity to small structural changes from mTBI compared to conventionally used PGSE acquisitions. Fourth, our study revealed sex differences detectable by dMRI that were not detectable fMRI. In particular, female mTBI mice showed greater reduction in dMRI metrics than male mTBI mice. These results demonstrate the importance of analyzing mTBI data for sex differences and also indicates that fMRI may have less sensitivity to mTBI pathology than dMRI. Fifth, we found utility in carrying out the fMRI analysis in a region-specific and subregion-specific manner. The region-specific analysis was based on parcellating the mouse brain into 22 regions (11 per hemisphere) while the subregion-specific analysis was based on parcellating the mouse brain into 82 regions (41 per hemisphere). The region-specific analysis, being dependent on a coarser parcellation of the brain than the subregion analysis, is less vulnerable to variability in the data caused by registration errors, and movement artifacts than the subregion-specific. Moreover, averaging across bigger regions translates to higher SNR. The greater specificity of the subregion analysis however will allow future studies to predict behavioral outcomes from mTBI with greater accuracy than the region-specific analysis as our knowledge of brain function is based on anatomical subregions. It is hoped that these lessons learned will improve the reliability and reproducibility of imaging concussion in animals and guide our thinking about how to better image the concussed human brain.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthors Contributions:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization: A.E., J.H., X.X., N.G., C.A.B., R.S.M., A.B.;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eData curation: A.E., J.H., X.X., N.G.;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFormal analysis: A.E., J.H., X.X.;\u003c/p\u003e\n\u003cp\u003eFunding acquisition: A.B.; V.F.P., M.A.M.P., C.A.B., R.S.M\u003c/p\u003e\n\u003cp\u003eResources: X.X., N.G., V.F.P., M.A.M.P., A.B.;\u003c/p\u003e\n\u003cp\u003eSupervision: C.A.B., R.S.M., A.B.;\u003c/p\u003e\n\u003cp\u003eWriting - original draft preparation: A.E., J.H., X.X.; \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWriting - review and editing: A.E., J.H., X.X., N.G., V.F.P., M.A.M.P., C.A.B., R.S.M., A.B.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cu\u003eEthics approval and consent to participate\u003c/u\u003e\u003c/em\u003e: \u0026nbsp; Ethics and Consent to Participate declarations: not applicable.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cu\u003eConsent for publication:\u003c/u\u003e\u003c/em\u003e\u0026nbsp; Consent to Publish declaration: not applicable.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cu\u003eData Availability\u003c/u\u003e\u003c/em\u003e\u003cem\u003e:\u0026nbsp;\u003c/em\u003eRaw MRI data (NIfTI format) will be made publicly available through the Federated Research Data Repository (FRDR) prior to publication. All analysis code used in this study is available from the corresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cu\u003eCompeting Interests:\u0026nbsp;\u003c/u\u003e\u003c/em\u003eThe authors have no relevant financial or non-financial interests to disclose.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cu\u003eFunding:\u003c/u\u003e\u003c/em\u003e\u0026nbsp; J. H. was supported by the Natural Sciences and Engineering Research Council of Canada: Canada Graduate Scholarships\u0026mdash;Doctoral Program (NSERC-CGS D). \u0026nbsp;C. A. B. was supported by Canada Research Chairs (950-231993). \u0026nbsp;Data collection was supported by the Canada First Research Excellence Fund to BrainsCAN. \u0026nbsp; Study design, data collection, analyses, interpretation and manuscript preparation was supported by the Canadian Institute of Health Research FDN 148453, the National Hockey League Players Association Challenge Fund, and the US Department of Defense under congress-directed medical research program (CDMRP), Peer Reviewed Alzheimer\u0026rsquo;s Research Program (PRARP) by award# W81XWH-20-1-0323.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe MAPTKI mice were a kind gift from Dr. Takaomi C. Saido (RIKEN Brain Science Institute).\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAbbas K, Shenk TE, Poole VN, Breedlove EL, Leverenz LJ, Nauman EA et al (2015) Alteration of Default Mode Network in High School Football Athletes Due to Repetitive Subconcussive Mild Traumatic Brain Injury: A Resting-State Functional Magnetic Resonance Imaging Study. Brain Connect 5:91\u0026ndash;101. doi: 10.1089/brain.2014.0279\u003c/li\u003e\n\u003cli\u003eAdams C, Bazzigaluppi P, Beckett TL, Bishay J, Weisspapir I, Dorr A et al (2018) Neurogliovascular dysfunction in a model of repeated traumatic brain injury. Theranostics 8:4824\u0026ndash;4836. doi: 10.7150/thno.24747\u003c/li\u003e\n\u003cli\u003eAggarwal M, Burnsed J, Martin LJ, Northington FJ, Zhang J (2014) Imaging neurodegeneration in the mouse hippocampus after neonatal hypoxia-ischemia using oscillating gradient diffusion MRI. Magn Reson Med 72:829\u0026ndash;840. doi: 10.1002/mrm.24956\u003c/li\u003e\n\u003cli\u003eAggarwal M, Jones M V., Calabresi PA, Mori S, Zhang J (2012) Probing mouse brain microstructure using oscillating gradient diffusion MRI. Magn Reson Med 67:98\u0026ndash;109. doi: 10.1002/mrm.22981\u003c/li\u003e\n\u003cli\u003eAggarwal M, Smith MD, Calabresi PA (2020) Diffusion‐time dependence of diffusional kurtosis in the mouse brain. Magn Reson Med 84:1564\u0026ndash;1578. doi: 10.1002/mrm.28189\u003c/li\u003e\n\u003cli\u003eAngelova P, Kehayov I, Davarski A, Kitov B (2021) Contemporary insight into diffuse axonal injury. Folia Med (Plovdiv) 63:163\u0026ndash;170. doi: 10.3897/folmed.63.e53709\u003c/li\u003e\n\u003cli\u003eAvants BB, Tustison NJ, Song G, Cook PA, Klein A, Gee JC (2011) A reproducible evaluation of ANTs similarity metric performance in brain image registration. Neuroimage 54:2033\u0026ndash;44. doi: 10.1016/j.neuroimage.2010.09.025\u003c/li\u003e\n\u003cli\u003eBaglietto-Vargas D, Forner S, Cai L, Martini AC, Trujillo-Estrada L, Swarup V et al (2021) Generation of a humanized A\u0026beta; expressing mouse demonstrating aspects of Alzheimer\u0026rsquo;s disease-like pathology. Nat Commun 12:2421. doi: 10.1038/s41467-021-22624-z\u003c/li\u003e\n\u003cli\u003eBaron CA, Kate M, Gioia L, Butcher K, Emery D, Budde M et al (2015) Reduction of Diffusion-Weighted Imaging Contrast of Acute Ischemic Stroke at Short Diffusion Times. Stroke 46:2136\u0026ndash;2141. doi: 10.1161/STROKEAHA.115.008815\u003c/li\u003e\n\u003cli\u003eBeckmann CF, DeLuca M, Devlin JT, Smith SM (2005) Investigations into resting-state connectivity using independent component analysis. Philosophical Transactions of the Royal Society B: Biological Sciences 360:1001\u0026ndash;1013. doi: 10.1098/rstb.2005.1634\u003c/li\u003e\n\u003cli\u003eBharath RD, Munivenkatappa A, Gohel S, Panda R, Saini J, Rajeswaran J et al (2015) Recovery of resting brain connectivity ensuing mild traumatic brain injury. Front Hum Neurosci 9. doi: 10.3389/fnhum.2015.00513\u003c/li\u003e\n\u003cli\u003eBlennow K, Brody DL, Kochanek PM, Levin H, McKee A, Ribbers GM et al (2016) Traumatic brain injuries. Nat Rev Dis Primers 2:16084. doi: 10.1038/nrdp.2016.84\u003c/li\u003e\n\u003cli\u003eBorsos KB, Tse DHY, Dubovan PI, Baron CA (2023) Tuned bipolar oscillating gradients for mapping frequency dispersion of diffusion kurtosis in the human brain. Magn Reson Med 89:756\u0026ndash;766. doi: 10.1002/mrm.29473\u003c/li\u003e\n\u003cli\u003eBoshra R, Ruiter KI, Dhindsa K, Sonnadara R, Reilly JP, Connolly JF (2020) On the time-course of functional connectivity: theory of a dynamic progression of concussion effects. Brain Commun 2. doi: 10.1093/braincomms/fcaa063\u003c/li\u003e\n\u003cli\u003eBraun NJ, Liao D, Alford PW (2021) Orientation of neurites influences severity of mechanically induced tau pathology. Biophys J 120:3272\u0026ndash;3282. doi: 10.1016/j.bpj.2021.07.011\u003c/li\u003e\n\u003cli\u003eBrowne KD, Chen X-H, Meaney DF, Smith DH (2011) Mild Traumatic Brain Injury and Diffuse Axonal Injury in Swine. J Neurotrauma 28:1747\u0026ndash;1755. doi: 10.1089/neu.2011.1913\u003c/li\u003e\n\u003cli\u003eChen J, Ades-Aron B, Lee H-H, Mehrin S, Pang M, Novikov DS et al (2024) Optimization and validation of the DESIGNER preprocessing pipeline for clinical diffusion MRI in white matter aging. Imaging Neuroscience 2:1\u0026ndash;17. doi: 10.1162/imag_a_00125\u003c/li\u003e\n\u003cli\u003eCox RW (1996) AFNI: Software for Analysis and Visualization of Functional Magnetic Resonance Neuroimages. Computers and Biomedical Research 29:162\u0026ndash;173. doi: 10.1006/cbmr.1996.0014\u003c/li\u003e\n\u003cli\u003eCriado-Marrero M, Ravi S, Bhaskar E, Barroso D, Pizzi MA, Williams L et al (2024) Age dictates brain functional connectivity and axonal integrity following repetitive mild traumatic brain injuries in mice. Neuroimage 298:120764. doi: 10.1016/j.neuroimage.2024.120764\u003c/li\u003e\n\u003cli\u003eDamoiseaux JS, Greicius MD (2009) Greater than the sum of its parts: a review of studies combining structural connectivity and resting-state functional connectivity. Brain Struct Funct 213:525\u0026ndash;33. doi: 10.1007/s00429-009-0208-6\u003c/li\u003e\n\u003cli\u003eDean PJA, Sato JR, Vieira G, McNamara A, Sterr A (2015) Long-term structural changes after mTBI and their relation to post-concussion symptoms. Brain Inj 29:1211\u0026ndash;1218. doi: 10.3109/02699052.2015.1035334\u003c/li\u003e\n\u003cli\u003eDoll\u0026eacute; J-P, Jaye A, Anderson SA, Ahmadzadeh H, Shenoy VB, Smith DH (2018) Newfound sex differences in axonal structure underlie differential outcomes from in vitro traumatic axonal injury. Exp Neurol 300:121\u0026ndash;134. doi: 10.1016/j.expneurol.2017.11.001\u003c/li\u003e\n\u003cli\u003eEierud C, Craddock RC, Fletcher S, Aulakh M, King-Casas B, Kuehl D et al (2014) Neuroimaging after mild traumatic brain injury: Review and meta-analysis. Neuroimage Clin 4:283\u0026ndash;294. doi: 10.1016/j.nicl.2013.12.009\u003c/li\u003e\n\u003cli\u003eFakhran S, Yaeger K, Collins M, Alhilali L (2014) Sex Differences in White Matter Abnormalities after Mild Traumatic Brain Injury: Localization and Correlation with Outcome. Radiology 272:815\u0026ndash;823. doi: 10.1148/radiol.14132512\u003c/li\u003e\n\u003cli\u003eGavish M, Donoho DL (2017) Optimal Shrinkage of Singular Values. IEEE Trans Inf Theory 63:2137\u0026ndash;2152. doi: 10.1109/TIT.2017.2653801\u003c/li\u003e\n\u003cli\u003eGennarelli TA, Thibault LE, Adams JH, Graham DI, Thompson CJ, Marcincin RP (1982) Diffuse axonal injury and traumatic coma in the primate. Ann Neurol 12:564\u0026ndash;574. doi: 10.1002/ana.410120611\u003c/li\u003e\n\u003cli\u003eGennarelli TA, Thibault LE, Tomei G, Wiser R, Graham D, Adams J (1987) Directional Dependence of Axonal Brain Injury due to Centroidal and Non-Centroidal Acceleration\u003c/li\u003e\n\u003cli\u003eGorgolewski K, Burns CD, Madison C, Clark D, Halchenko YO, Waskom ML et al (2011) Nipype: A Flexible, Lightweight and Extensible Neuroimaging Data Processing Framework in Python. Front Neuroinform 5. doi: 10.3389/fninf.2011.00013\u003c/li\u003e\n\u003cli\u003eGrandjean J, Canella C, Anckaerts C, Ayrancı G, Bougacha S, Bienert T et al (2020) Common functional networks in the mouse brain revealed by multi-centre resting-state fMRI analysis. Neuroimage 205:116278. doi: 10.1016/j.neuroimage.2019.116278\u003c/li\u003e\n\u003cli\u003eGreicius MD, Supekar K, Menon V, Dougherty RF (2009) Resting-state functional connectivity reflects structural connectivity in the default mode network. Cereb Cortex 19:72\u0026ndash;8. doi: 10.1093/cercor/bhn059\u003c/li\u003e\n\u003cli\u003eGuerriero RM, Giza CC, Rotenberg A (2015) Glutamate and GABA Imbalance Following Traumatic Brain Injury. Curr Neurol Neurosci Rep 15:27. doi: 10.1007/s11910-015-0545-1\u003c/li\u003e\n\u003cli\u003eG\u0026uuml;nt\u0026uuml;rk\u0026uuml;n O, Str\u0026ouml;ckens F, Ocklenburg S (2020) Brain Lateralization: A Comparative Perspective. Physiol Rev 100:1019\u0026ndash;1063. doi: 10.1152/physrev.00006.2019\u003c/li\u003e\n\u003cli\u003eGupte RP, Brooks WM, Vukas RR, Pierce JD, Harris JL (2019) Sex Differences in Traumatic Brain Injury: What We Know and What We Should Know. J Neurotrauma 36:3063\u0026ndash;3091. doi: 10.1089/neu.2018.6171\u003c/li\u003e\n\u003cli\u003eHaacke EM, Lindskogj ED, Lin W (1991) A fast, iterative, partial-fourier technique capable of local phase recovery. Journal of Magnetic Resonance (1969) 92:126\u0026ndash;145. doi: 10.1016/0022-2364(91)90253-P\u003c/li\u003e\n\u003cli\u003eHajiaghamemar M, Margulies SS (2021) Multi-Scale White Matter Tract Embedded Brain Finite Element Model Predicts the Location of Traumatic Diffuse Axonal Injury. J Neurotrauma 38:144\u0026ndash;157. doi: 10.1089/neu.2019.6791\u003c/li\u003e\n\u003cli\u003eHamilton J, Xu K, Geremia N, Prado VF, Prado MAM, Brown A et al (2024) Robust frequency-dependent diffusional kurtosis computation using an efficient direction scheme, axisymmetric modelling, and spatial regularization. Imaging Neuroscience 2:1\u0026ndash;22. doi: 10.1162/imag_a_00055\u003c/li\u003e\n\u003cli\u003eHarris NG, Verley DR, Gutman BA, Thompson PM, Yeh HJ, Brown JA (2016) Disconnection and hyper-connectivity underlie reorganization after TBI: A rodent functional connectomic analysis. Exp Neurol 277:124\u0026ndash;138. doi: 10.1016/j.expneurol.2015.12.020\u003c/li\u003e\n\u003cli\u003eHellewell SC, Nguyen VPB, Jayasena RN, Welton T, Grieve SM (2020) Characteristic patterns of white matter tract injury in sport-related concussion: An image based meta-analysis. Neuroimage Clin 26:102253. doi: 10.1016/j.nicl.2020.102253\u003c/li\u003e\n\u003cli\u003eHoney CJ, Sporns O, Cammoun L, Gigandet X, Thiran JP, Meuli R et al (2009) Predicting human resting-state functional connectivity from structural connectivity. Proc Natl Acad Sci U S A 106:2035\u0026ndash;40. doi: 10.1073/pnas.0811168106\u003c/li\u003e\n\u003cli\u003eHuang S, Shen Q, Watts LT, Long JA, O\u0026rsquo;Boyle M, Nguyen T et al (2021) Resting-State Functional Magnetic Resonance Imaging of Interhemispheric Functional Connectivity in Experimental Traumatic Brain Injury. Neurotrauma Rep 2:526\u0026ndash;540. doi: 10.1089/neur.2021.0023\u003c/li\u003e\n\u003cli\u003eHutchinson EB, Schwerin SC, Avram A V., Juliano SL, Pierpaoli C (2018) Diffusion MRI and the detection of alterations following traumatic brain injury. J Neurosci Res 96:612\u0026ndash;625. doi: 10.1002/jnr.24065\u003c/li\u003e\n\u003cli\u003eJelescu IO, Fieremans E (2023) Sensitivity and specificity of diffusion MRI to neuroinflammatory processes. pp 31\u0026ndash;50. doi: 10.1016/b978-0-323-91771-1.00010-1\u003c/li\u003e\n\u003cli\u003eJenkinson M, Beckmann CF, Behrens TEJ, Woolrich MW, Smith SM (2012) FSL. Neuroimage 62:782\u0026ndash;790. doi: 10.1016/j.neuroimage.2011.09.015\u003c/li\u003e\n\u003cli\u003eJensen JH, Helpern JA (2010) MRI quantification of non-Gaussian water diffusion by kurtosis analysis. NMR Biomed 23:698\u0026ndash;710. doi: 10.1002/nbm.1518\u003c/li\u003e\n\u003cli\u003eJensen JH, Helpern JA, Ramani A, Lu H, Kaczynski K (2005) Diffusional kurtosis imaging: The quantification of non-gaussian water diffusion by means of magnetic resonance imaging. Magn Reson Med 53:1432\u0026ndash;1440. doi: 10.1002/mrm.20508\u003c/li\u003e\n\u003cli\u003eKellner E, Dhital B, Kiselev VG, Reisert M (2016) Gibbs‐ringing artifact removal based on local subvoxel‐shifts. Magn Reson Med 76:1574\u0026ndash;1581. doi: 10.1002/mrm.26054\u003c/li\u003e\n\u003cli\u003eKnutsen AK, Gomez AD, Gangolli M, Wang W-T, Chan D, Lu Y-C et al (2020) In vivo estimates of axonal stretch and 3D brain deformation during mild head impact. Brain Multiphys 1:100015. doi: 10.1016/j.brain.2020.100015\u003c/li\u003e\n\u003cli\u003eKoay CG, Basser PJ (2006) Analytically exact correction scheme for signal extraction from noisy magnitude MR signals. Journal of Magnetic Resonance 179:317\u0026ndash;322. doi: 10.1016/j.jmr.2006.01.016\u003c/li\u003e\n\u003cli\u003eKrieg JL, Leonard A V., Turner RJ, Corrigan F (2023) Identifying the Phenotypes of Diffuse Axonal Injury Following Traumatic Brain Injury. Brain Sci 13:1607. doi: 10.3390/brainsci13111607\u003c/li\u003e\n\u003cli\u003eKulkarni P, Morrison TR, Cai X, Iriah S, Simon N, Sabrick J et al (2019) Neuroradiological Changes Following Single or Repetitive Mild TBI. Front Syst Neurosci 13:34. doi: 10.3389/fnsys.2019.00034\u003c/li\u003e\n\u003cli\u003eLee H, Novikov DS, Fieremans E (2021) Removal of partial Fourier‐induced Gibbs (RPG) ringing artifacts in MRI. Magn Reson Med 86:2733\u0026ndash;2750. doi: 10.1002/mrm.28830\u003c/li\u003e\n\u003cli\u003eLeemans A, Jeurissen D, Sijbers J, Jones DK (2009) ExploreDTI: a graphical toolbox for processing, analyzing, and visualizing diffusion MR data. In: 17th Annual Meeting of Intl Soc Mag Reson Med. Hawaii, p 3537\u003c/li\u003e\n\u003cli\u003eLindsey HM, Hodges CB, Greer KM, Wilde EA, Merkley TL (2023) Diffusion-Weighted Imaging in Mild Traumatic Brain Injury: A Systematic Review of the Literature. Neuropsychol Rev 33:42\u0026ndash;121. doi: 10.1007/s11065-021-09485-5\u003c/li\u003e\n\u003cli\u003eLiu X-B, Schumann CM (2014) Optimization of electron microscopy for human brains with long-term fixation and fixed-frozen sections. Acta Neuropathol Commun 2:42. doi: 10.1186/2051-5960-2-42\u003c/li\u003e\n\u003cli\u003eMaas AIR, Menon DK, Manley GT, Abrams M, \u0026Aring;kerlund C, Andelic N et al (2022) Traumatic brain injury: progress and challenges in prevention, clinical care, and research. Lancet Neurol 21:1004\u0026ndash;1060. doi: 10.1016/S1474-4422(22)00309-X\u003c/li\u003e\n\u003cli\u003eMacdonald C, Dikranian K, Song S, Bayly P, Holtzman D, Brody D (2007) Detection of traumatic axonal injury with diffusion tensor imaging in a mouse model of traumatic brain injury. Exp Neurol 205:116\u0026ndash;131. doi: 10.1016/j.expneurol.2007.01.035\u003c/li\u003e\n\u003cli\u003eManning KY, Schranz A, Bartha R, Dekaban GA, Barreira C, Brown A et al (2017) Multiparametric MRI changes persist beyond recovery in concussed adolescent hockey players. Neurology 89:2157\u0026ndash;2166. doi: 10.1212/WNL.0000000000004669\u003c/li\u003e\n\u003cli\u003eMarkicevic M, Mandino F, Toyonaga T, Cai Z, Fesharaki-Zadeh A, Shen X et al (2024) Repetitive Mild Closed-Head Injury Induced Synapse Loss and Increased Local BOLD-fMRI Signal Homogeneity. J Neurotrauma 41:2528\u0026ndash;2544. doi: 10.1089/neu.2024.0095\u003c/li\u003e\n\u003cli\u003eMayer AR, Bellgowan PSF, Hanlon FM (2015) Functional magnetic resonance imaging of mild traumatic brain injury. Neurosci Biobehav Rev 49:8\u0026ndash;18. doi: 10.1016/j.neubiorev.2014.11.016\u003c/li\u003e\n\u003cli\u003eMayer AR, Mannell M V., Ling J, Gasparovic C, Yeo RA (2011) Functional connectivity in mild traumatic brain injury. Hum Brain Mapp 32:1825\u0026ndash;1835. doi: 10.1002/hbm.21151\u003c/li\u003e\n\u003cli\u003eMcNamara EH, Grillakis AA, Tucker LB, McCabe JT (2020) The closed-head impact model of engineered rotational acceleration (CHIMERA) as an application for traumatic brain injury pre-clinical research: A status report. Exp Neurol 333:113409. doi: 10.1016/j.expneurol.2020.113409\u003c/li\u003e\n\u003cli\u003eMeaney DF, Smith DH (2011) Biomechanics of Concussion. Clin Sports Med 30:19\u0026ndash;31. doi: 10.1016/j.csm.2010.08.009\u003c/li\u003e\n\u003cli\u003eMeier TB, Giraldo-Chica M, Espa\u0026ntilde;a LY, Mayer AR, Harezlak J, Nencka AS et al (2020) Resting-State fMRI Metrics in Acute Sport-Related Concussion and Their Association with Clinical Recovery: A Study from the NCAA-DOD CARE Consortium. J Neurotrauma 37:152\u0026ndash;162. doi: 10.1089/neu.2019.6471\u003c/li\u003e\n\u003cli\u003eMohamed AZ, Cumming P, Nasrallah FA (2021) Traumatic brain injury augurs ill for prolonged deficits in the brain\u0026rsquo;s structural and functional integrity following controlled cortical impact injury. Sci Rep 11:21559. doi: 10.1038/s41598-021-00660-5\u003c/li\u003e\n\u003cli\u003eMorozov S, Sergunova K, Petraikin A, Akhmad E, Kivasev S, Semenov D et al (2020) Diffusion processes modeling in magnetic resonance imaging. Insights Imaging 11:60. doi: 10.1186/s13244-020-00863-w\u003c/li\u003e\n\u003cli\u003eNickerson LD, Smith SM, \u0026Ouml;ng\u0026uuml;r D, Beckmann CF (2017) Using Dual Regression to Investigate Network Shape and Amplitude in Functional Connectivity Analyses. Front Neurosci 11. doi: 10.3389/fnins.2017.00115\u003c/li\u003e\n\u003cli\u003eNovikov DS, Kiselev VG (2011) Surface-to-volume ratio with oscillating gradients. Journal of Magnetic Resonance 210:141\u0026ndash;145. doi: 10.1016/j.jmr.2011.02.011\u003c/li\u003e\n\u003cli\u003eOh SW, Harris JA, Ng L, Winslow B, Cain N, Mihalas S et al (2014) A mesoscale connectome of the mouse brain. Nature 508:207\u0026ndash;214. doi: 10.1038/nature13186\u003c/li\u003e\n\u003cli\u003eOlesen JL, Ianus A, \u0026Oslash;stergaard L, Shemesh N, Jespersen SN (2023) Tensor denoising of multidimensional MRI data. Magn Reson Med 89:1160\u0026ndash;1172. doi: 10.1002/mrm.29478\u003c/li\u003e\n\u003cli\u003ePalacios EM, Yuh EL, Chang Y-S, Yue JK, Schnyer DM, Okonkwo DO et al (2017) Resting-State Functional Connectivity Alterations Associated with Six-Month Outcomes in Mild Traumatic Brain Injury. J Neurotrauma 34:1546\u0026ndash;1557. doi: 10.1089/neu.2016.4752\u003c/li\u003e\n\u003cli\u003eParent M, Li Y, Santhakumar V, Hyder F, Sanganahalli BG, Kannurpatti SS (2019) Alterations of Parenchymal Microstructure, Neuronal Connectivity, and Cerebrovascular Resistance at Adolescence after Mild-to-Moderate Traumatic Brain Injury in Early Development. J Neurotrauma 36:601\u0026ndash;608. doi: 10.1089/neu.2018.5741\u003c/li\u003e\n\u003cli\u003ePettus EH, Christman CW, Giebel ML, Polvishock JT (1994) Traumatically Induced Altered Membrane Permeability: Its Relationship to Traumatically Induced Reactive Axonal Change. J Neurotrauma 11:507\u0026ndash;522. doi: 10.1089/neu.1994.11.507\u003c/li\u003e\n\u003cli\u003eRahman N, Xu K, Budde MD, Brown A, Baron CA (2023) A longitudinal microstructural MRI dataset in healthy C57Bl/6 mice at 9.4 Tesla. Sci Data 10:94. doi: 10.1038/s41597-023-01942-5\u003c/li\u003e\n\u003cli\u003eRubiano AM, Carney N, Chesnut R, Puyana JC (2015) Global neurotrauma research challenges and opportunities. Nature 527:S193\u0026ndash;S197. doi: 10.1038/nature16035\u003c/li\u003e\n\u003cli\u003eSaito T, Mihira N, Matsuba Y, Sasaguri H, Hashimoto S, Narasimhan S et al (2019) Humanization of the entire murine Mapt gene provides a murine model of pathological human tau propagation. Journal of Biological Chemistry 294:12754\u0026ndash;12765. doi: 10.1074/jbc.RA119.009487\u003c/li\u003e\n\u003cli\u003eSakthivel R, Criado-Marrero M, Barroso D, Braga IM, Bolen M, Rubinovich U et al (2023) Fixed Time-Point Analysis Reveals Repetitive Mild Traumatic Brain Injury Effects on Resting State Functional Magnetic Resonance Imaging Connectivity and Neuro-Spatial Protein Profiles. J Neurotrauma 40:2037\u0026ndash;2049. doi: 10.1089/neu.2022.0464\u003c/li\u003e\n\u003cli\u003eSchachter M, Does MD, Anderson AW, Gore JC (2000) Measurements of Restricted Diffusion Using an Oscillating Gradient Spin-Echo Sequence. Journal of Magnetic Resonance 147:232\u0026ndash;237. doi: 10.1006/jmre.2000.2203\u003c/li\u003e\n\u003cli\u003eShumskaya E, Andriessen TMJC, Norris DG, Vos PE (2012) Abnormal whole-brain functional networks in homogeneous acute mild traumatic brain injury. Neurology 79:175\u0026ndash;182. doi: 10.1212/WNL.0b013e31825f04fb\u003c/li\u003e\n\u003cli\u003eSilberfeld A, Roe JM, Ellegood J, Lerch JP, Qiu L, Kim Y et al (2025) Right-left Brain-Wide Asymmetry of Neuroanatomy in the Mouse Brain. Neuroimage 307:121017. doi: 10.1016/j.neuroimage.2025.121017\u003c/li\u003e\n\u003cli\u003eSinke MRT, Otte WM, Meerwaldt AE, Franx BAA, Ali MHM, Rakib F et al (2021) Imaging Markers for the Characterization of Gray and White Matter Changes from Acute to Chronic Stages after Experimental Traumatic Brain Injury. J Neurotrauma 38:1642\u0026ndash;1653. doi: 10.1089/neu.2020.7151\u003c/li\u003e\n\u003cli\u003eSkudlarski P, Jagannathan K, Calhoun VD, Hampson M, Skudlarska BA, Pearlson G (2008) Measuring brain connectivity: diffusion tensor imaging validates resting state temporal correlations. Neuroimage 43:554\u0026ndash;61. doi: 10.1016/j.neuroimage.2008.07.063\u003c/li\u003e\n\u003cli\u003eSmith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TEJ, Johansen-Berg H et al (2004) Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage 23:S208\u0026ndash;S219. doi: 10.1016/j.neuroimage.2004.07.051\u003c/li\u003e\n\u003cli\u003eSollmann N, Echlin PS, Schultz V, Viher P V., Lyall AE, Tripodis Y et al (2018) Sex differences in white matter alterations following repetitive subconcussive head impacts in collegiate ice hockey players. Neuroimage Clin 17:642\u0026ndash;649. doi: 10.1016/j.nicl.2017.11.020\u003c/li\u003e\n\u003cli\u003eSong H, Tomasevich A, Paolini A, Browne KD, Wofford KL, Kelley B et al (2024) Sex differences in the extent of acute axonal pathologies after experimental concussion. Acta Neuropathol 147:79. doi: 10.1007/s00401-024-02735-9\u003c/li\u003e\n\u003cli\u003eStone JR, Okonkwo DO, Dialo AO, Rubin DG, Mutlu LK, Povlishock JT et al (2004) Impaired axonal transport and altered axolemmal permeability occur in distinct populations of damaged axons following traumatic brain injury. Exp Neurol 190:59\u0026ndash;69. doi: 10.1016/j.expneurol.2004.05.022\u003c/li\u003e\n\u003cli\u003eSullivan S, Eucker SA, Gabrieli D, Bradfield C, Coats B, Maltese MR et al (2015) White matter tract-oriented deformation predicts traumatic axonal brain injury and reveals rotational direction-specific vulnerabilities. Biomech Model Mechanobiol 14:877\u0026ndash;896. doi: 10.1007/s10237-014-0643-z\u003c/li\u003e\n\u003cli\u003eTayebi M, Holdsworth SJ, Champagne AA, Cook DJ, Nielsen P, Lee T-R et al (2021) The role of diffusion tensor imaging in characterizing injury patterns on athletes with concussion and subconcussive injury: a systematic review. Brain Inj 35:621\u0026ndash;644. doi: 10.1080/02699052.2021.1895313\u003c/li\u003e\n\u003cli\u003eTeasell EM, Potts E, Geremia N, Lu L, Xu X, Mao H et al (2025) A Clinically Relevant Mouse Model of Concussion Incorporating High Rotational Forces. Neurotrauma Rep 6:184\u0026ndash;190. doi: 10.1089/neur.2024.0165\u003c/li\u003e\n\u003cli\u003eTo XV, Nasrallah FA (2021) A roadmap of brain recovery in a mouse model of concussion: insights from neuroimaging. Acta Neuropathol Commun 9:2. doi: 10.1186/s40478-020-01098-y\u003c/li\u003e\n\u003cli\u003eVelayudhan PS, Mak JJ, Gazdzinski LM, Wheeler AL (2022) Persistent white matter vulnerability in a mouse model of mild traumatic brain injury. BMC Neurosci 23:46. doi: 10.1186/s12868-022-00730-y\u003c/li\u003e\n\u003cli\u003eVeraart J, Novikov DS, Christiaens D, Ades-aron B, Sijbers J, Fieremans E (2016) Denoising of diffusion MRI using random matrix theory. Neuroimage 142:394\u0026ndash;406. doi: 10.1016/j.neuroimage.2016.08.016\u003c/li\u003e\n\u003cli\u003eVerley DR, Torolira D, Pulido B, Gutman B, Bragin A, Mayer A et al (2018) Remote Changes in Cortical Excitability after Experimental Traumatic Brain Injury and Functional Reorganization. J Neurotrauma 35:2448\u0026ndash;2461. doi: 10.1089/neu.2017.5536\u003c/li\u003e\n\u003cli\u003eVinh To X, Soni N, Medeiros R, Alateeq K, Nasrallah FA (2022) Traumatic brain injury alterations in the functional connectome are associated with neuroinflammation but not tau in a P30IL tauopathy mouse model. Brain Res 1789:147955. doi: 10.1016/j.brainres.2022.147955\u003c/li\u003e\n\u003cli\u003eWang Q, Ding S-L, Li Y, Royall J, Feng D, Lesnar P et al (2020) The Allen Mouse Brain Common Coordinate Framework: A 3D Reference Atlas. Cell 181:936-953.e20. doi: 10.1016/j.cell.2020.04.007\u003c/li\u003e\n\u003cli\u003eWinkler AM, Ridgway GR, Webster MA, Smith SM, Nichols TE (2014) Permutation inference for the general linear model. Neuroimage 92:381\u0026ndash;397. doi: 10.1016/j.neuroimage.2014.01.060\u003c/li\u003e\n\u003cli\u003eWright DK, Symons GF, O\u0026rsquo;Brien WT, McDonald SJ, Zamani A, Major B et al (2021) Diffusion Imaging Reveals Sex Differences in the White Matter Following Sports-Related Concussion. Cerebral Cortex 31:4411\u0026ndash;4419. doi: 10.1093/cercor/bhab095\u003c/li\u003e\n\u003cli\u003eWu D, Li Q, Northington FJ, Zhang J (2018) Oscillating gradient diffusion kurtosis imaging of normal and injured mouse brains. NMR Biomed 31:e3917. doi: 10.1002/nbm.3917\u003c/li\u003e\n\u003cli\u003eXu X, Cowan M, Beraldo F, Schranz A, McCunn P, Geremia N et al (2021) Repetitive mild traumatic brain injury in mice triggers a slowly developing cascade of long-term and persistent behavioral deficits and pathological changes. Acta Neuropathol Commun 9:60. doi: 10.1186/s40478-021-01161-2\u003c/li\u003e\n\u003cli\u003eYang Z, Zhu T, Pompilus M, Fu Y, Zhu J, Arjona K et al (2021) Compensatory functional connectome changes in a rat model of traumatic brain injury. Brain Commun 3. doi: 10.1093/braincomms/fcab244\u003c/li\u003e\n\u003cli\u003eYushkevich PA, Gao Y, Gerig G (2016) ITK-SNAP: An interactive tool for semi-automatic segmentation of multi-modality biomedical images. In: 2016 38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). IEEE, pp 3342\u0026ndash;3345\u003c/li\u003e\n\u003cli\u003eZerbi V, Grandjean J, Rudin M, Wenderoth N (2015) Mapping the mouse brain with rs-fMRI: An optimized pipeline for functional network identification. Neuroimage 123:11\u0026ndash;21. doi: 10.1016/j.neuroimage.2015.07.090\u003c/li\u003e\n\u003cli\u003eZhang J, Solar K, Safar K, Zamyadi R, Vandewouw MM, Da Costa L et al (2024) The structural, functional, and neurophysiological connectome of mild traumatic brain injury: a DTI, fMRI and MEG multimodal clustering and data fusion study. medRxiv. doi: 10.1101/2024.06.24.24309379\u003c/li\u003e\n\u003cli\u003eZhao P, Zhu P, Zhang D, Yin B, Wang Y, Hussein NM et al (2022) Sex Differences in Cerebral Blood Flow and Serum Inflammatory Cytokines and Their Relationships in Mild Traumatic Brain Injury. Front Neurol 12. doi: 10.3389/fneur.2021.755152\u003c/li\u003e\n\u003cli\u003eZhu DC, Covassin T, Nogle S, Doyle S, Russell D, Pearson RL et al (2015) A Potential Biomarker in Sports-Related Concussion: Brain Functional Connectivity Alteration of the Default-Mode Network Measured with Longitudinal Resting-State fMRI over Thirty Days. J Neurotrauma 32:327\u0026ndash;341. doi: 10.1089/neu.2014.3413\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"acta-neuropathologica-communications","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"anec","sideBox":"Learn more about [Acta Neuropathologica Communications](https://actaneurocomms.biomedcentral.com/)","snPcode":"40478","submissionUrl":"https://submission.springernature.com/new-submission/40478/3","title":"Acta Neuropathologica Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Diffusion MRI, Functional MRI, Mild traumatic brain injury, Concussion","lastPublishedDoi":"10.21203/rs.3.rs-6985478/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6985478/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eWhile neuroimaging studies have revealed notable white matter damage following mild traumatic brain injury (mTBI), the specific tracts and brain regions affected vary widely across studies. Here, we explored whether the spatial orientation of white matter tracts influences susceptibility to mTBI, predicting that tracts oriented orthogonal to the axis of rotation of the head during impact (within the plane of rotation) would exhibit the most damage. Using a model of repeated rotational mTBI in mice, we acquired advanced diffusion MRI (diffusional kurtosis imaging using oscillating gradient encoding) and resting-state functional MRI (fMRI) data at baseline and 1-week post-injury. Consistent with our prediction, while both diffusivity and diffusional kurtosis decreased in the white matter of injured mice, only diffusional kurtosis revealed microstructural changes confined to tracts oriented orthogonal to the right-left axis of rotation. In addition, both region and subregion analyses showed FC deficits between regions connected via tracts running orthogonal to the rotation axis. The orientation-dependent changes in imaging metrics were validated by histopathological analyses. Females showed greater microstructural changes than males using dMRI following injury, while no sex differences were detected by fMRI. Interestingly, the region-specific and subregion-specific FC analyses showed overlapping but non-identical changes in FC suggesting the utility of using both coarse and fine levels of brain parcellation for FC analyses in mTBI. These findings suggest that mTBI imaging studies may benefit from the consideration that damage after mTBI will predominate in tracts that are oriented orthogonal to the axis of rotation produced by the impact and that diffusivity and diffusional kurtosis as well as region and subregion-specific fMRI analyses can detect these changes.\u003c/p\u003e","manuscriptTitle":"MRI investigation of orientation-dependent changes in microstructure and function in a mouse model of mild traumatic brain injury","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-14 11:40:58","doi":"10.21203/rs.3.rs-6985478/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-08-25T13:56:02+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-27T03:54:57+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-26T19:50:11+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-24T07:40:59+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"17617540767194755121111930775792942918","date":"2025-07-14T00:09:42+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"149319315045831895660419205169033970953","date":"2025-07-13T03:30:31+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"105204855746872038458894404216022817711","date":"2025-07-11T01:24:29+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"124211431860950241532164962386424221334","date":"2025-07-10T19:18:39+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-07-10T14:15:46+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-07-02T04:50:43+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-07-02T04:50:38+00:00","index":"","fulltext":""},{"type":"submitted","content":"Acta Neuropathologica Communications","date":"2025-06-26T16:50:43+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"acta-neuropathologica-communications","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"anec","sideBox":"Learn more about [Acta Neuropathologica Communications](https://actaneurocomms.biomedcentral.com/)","snPcode":"40478","submissionUrl":"https://submission.springernature.com/new-submission/40478/3","title":"Acta Neuropathologica Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"f4c4a206-4bfa-4fee-8d42-251b9b7795c6","owner":[],"postedDate":"July 14th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-12-08T15:59:32+00:00","versionOfRecord":{"articleIdentity":"rs-6985478","link":"https://doi.org/10.1186/s40478-025-02183-w","journal":{"identity":"acta-neuropathologica-communications","isVorOnly":false,"title":"Acta Neuropathologica Communications"},"publishedOn":"2025-12-01 15:57:10","publishedOnDateReadable":"December 1st, 2025"},"versionCreatedAt":"2025-07-14 11:40:58","video":"","vorDoi":"10.1186/s40478-025-02183-w","vorDoiUrl":"https://doi.org/10.1186/s40478-025-02183-w","workflowStages":[]},"version":"v1","identity":"rs-6985478","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6985478","identity":"rs-6985478","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}