Figure 3. Box plots of DC indicate stronger spatial shifts of iWSAs in patients with poor Circle of Willis configuration.

The boxes contain all single-subject DC values between the first and third quartiles, the line inside marks the median, and the whiskers reach from minimum to maximum (not including outliers). In patients with poor Circle of Willis configuration (left), iWSA overlap between scan 1 and 2 differed more between hemispheres (p=0.007) while patients with sound configuration of the CoW did not present significant side differences (p=0.938).

Figure 4. Paired Scatter Plots comparing single-subject coefficients of variance within the high CoV mask.

Data points represent mean coefficients of variance (CoV) of each subject for each hemisphere, lines connect the mean CoV between hemispheres. The red dashed line displays the group average. Stable CoV was found between hemispheres for both groups and scans. A two-sided t-test revealed no significant differences between hemispheres (controls in scan 1/2: p=0.86/0.94; patients in scan 1/2: p=0.92/0.92).

Figure 4. Coronal MinIP images of 2D RF (a), optimized method (b), and conventional SPACE (c). The optimization method achieves more thorough signal suppression compare to traditional 2D RF and suffer from fewer aliasing artifacts (white arrow). Due to the increase in spatial resolution, the optimized method shows more branches that are invisible to conventional SPACE images (yellow arrows).

Figure 5. The comparisons of IV-SPACE and conventional SPACE in the coronal plane. The enlarged images were shown on the right side. The lumen and orifice of an LSA are clearly depicted in the IV-SPACE image, whereas the lumen and orifice of the vessel were blurred in conventional SPACE. The depiction of the LSA vessel wall of IV-SPACE is also better than conventional SPACE.

Figure 1: MRI Protocol and derived parameters. Grey matter (GM) masks were derived from structural imaging, cerebral blood flow (CBF) by pseudo-continuous arterial spin labeling (pCASL),¹⁷ relative cerebral blood volume (rCBV) by dynamic susceptibility contrast (DSC)¹⁶ and combined with T₂* and T₂ yielding relative oxygen extraction fraction (rOEF)¹⁷. Based on the DSC time-series, voxels with high coefficient of variation (CoV, see Eq.1) were segmented (orange)¹² and applied to hemodynamic parameters (green) after exclusion of ventricles and large vessels¹⁵ (red).

Figure 4: Hemodynamic characteristics in high-CoV voxels. CBF (A), rCBV (B) and rOEF (C) are compared in grey matter (GM; yellow) vs. high-CoV voxels (CoV; red) in healthy controls (HC; green) and ICAS patients (orange). Dots show mean parameter values, black lines connect the same subject’s mean values and red dashed lines represent group average values. CBF was systematically lower due to background suppression¹³ and rOEF elevated due to T₂ bias²⁶. In high-CoV voxels of both groups, all parameters showed significant effects. While CBF (A) and rCBV (B) were higher, rOEF was lower (C).

Figure 1. The flowchart of GRASP-Pro reconstruction. In an intermediate step, standard GRASP reconstruction is performed on the low-resolution portion of the sorted radial k-space for estimating temporal basis (U). In the second step, GRASP-Pro is then performed on the full-resolution radial k-space incorporating the pre-learned basis (UK with only the first dominant K basis components) to reconstruct VK, the coefficients to represent the fully-resolution image-series under UK. The final images m is then computed as UKVK.

Figure 3. MRA MIP images at 5 representative phases with 10 and 20 spokes per frame, respectively. MIP images at the same TIs were shown for the two reconstructions for better comparison.

Figure 1: (left) In-silico model of the arterial cerebrovascular space, showing the extracted velocity and pressure field of the main vessels of the circle of Willis; (right) Example output from the separate in-vivo 4D Flow MRI analysis, showing time-integrated pathlines within vessel regions superimposed on axial magnitude images for two individuals, each at dx = 1.1 and 0.3 mm³.

Figure 2: Linear regression plots for different spatial resolutions, comparing estimated relative pressure output ( by RB, UB, and vWERP) against reference CFD relative pressure ($$$\Delta p$$$). Colors indicate estimates over individual vascular sections shown in the diagram to the left.

Figure 1: Data processing workflow: Image data and ROI masks are taken as inputs. Vessels are first enhanced and then segmented. After generating the skeleton, a vessel distance map can be computed by applying the Euclidian distance transform. Local vessel density and VDM estimates are computed from the segmentation and distance maps respectively.

MIP – Maximum Intensity Projection; ROI – Region of Interest; VDM – Vessel Distance Mapping

Figure 5: Arterial and venous distance maps of the left hippocampus (central slice shown). By co-registering all subjects to the study template space a group VDM average could be computed. For comparison, all segmentations have been averaged in the template space (shown as MIP) to show the overlap of the vascular patterns.

MIP – Maximum Intensity Projection; VDM – Vessel Distance Mapping

A diagram of acute LVO characteristics and the corresponding typical HRMRI images. HRMRI distinguishes the occluded segments (filled arrowhead) from the slow-flow areas (empty arrowhead). A. ICAS-LVO treated with rescue stenting. Grade 1 enhancement is present on HRMRI, which could reflect higher atherosclerosis plaque percentages compared with those of clots. B. An ICAS-LVO not treated with a stent procedure. Grade 0 enhancement is present on HRMRI, reflecting a higher percentage of clot formation. C. An embolic occlusion treated with a stent retriever.

A representative case of occluded segment lengths measured with HRMRI in acute MCA occlusion 7 h after the onset of stroke. The occluded segment lengths measured with angiography (A) and HRMRI (Pre- [B1] and postcontrast [B2] imaging) were 19.9 and 16.2 millimeters, respectively. Slow-flow segments were present at the distal and proximal ends of the occluded segment (arrowhead). (C) The time-of-flight image revealed obstruction of the MCA but did not show the occlusion site specifically. (D1) a gross photograph of a clot. (D2) an H&E-stained section of the clot.

(A) All mice were imaged under 1.3-1.8% isoflurane using 4 optimized sequences whose parameter settings have been listed here, with total scan time of ~55 min/animal. (B) An atlas-based imaging data analysis pipeline was used to register mouse brain anatomical and pathological information to the Allen Brain Reference Atlas while K-means algorithm was used to cluster and quantify different lesion types.

(A) SWI detected hypointense lesions overlooked by T2W imaging. As shown in the pie chart, where the pie radius reflects the total lesion volume per mouse, allelic heterogeneity was shown to influence disease severity and regional distribution of brain regions. Notably, lesions in cortical regions were only observed in the G1038S cohort. (B) The volume of SWI-sensitive lesions was significantly higher in mutant groups (G1038S and G1344D) compared to WT (as expected), and differ between Col4a1 mutant strains.

Fig4. Representative schematic of Na⁺/K⁺ ATPase displays the biological explanation for utilizing ²³Na MRI to assess cerebral ischemia onset and recovery with treatment. Fractional changes in lesion volume and signal via ²³Na CSI provide a sensitive metric to assess treatments.

Fig1. T₂ relaxation images and corresponding rate decays demonstrate SPIO uptake in hMSC EV via sonication and compared to EV exposed to SPIO but not sonication. Lower layers corresponding to supernatant, with this layer in the un-sonicated sample displaying significant diffusion of SPIO into surrounding gel.

Intracellular volume fraction (ICVF) and orientation dispersion index (ODI) for the ipsilateral external capsule and striatum. Representative NODDI maps are provided for a specimen from the hMSC group. * with bracket = significant difference between treatment groups, + = significant difference between treatment group & naïve. #, ¥ and ● = significant difference within group from D3, 7 and 21, respectively. Significance level set at p<0.05.

Mean, Axial and Radial Diffusivity (MD, AD, RD) and Fractional Isotropy for the ipsilateral external capsule and striatum. Representative DTI maps are provided for a specimen from the hMSC group. * with bracket = significant difference between treatment groups, + = significant difference between treatment group & naïve. #, ¥ and ● = significant difference within group from D3, 7 and 21, respectively. Significance level set at p<0.05.

Figure 1. The best diagnostic biomarker of CASH 3–12 months prior to DCEQP imaging. (A) Weighted β_i coefficients of covariates (all p<0.05) in the perfusion biomarker of CASH 3–12 months prior; error bars are SE. (B) The biomarker distinguished CASH 3–12 months prior (AUC=0.86, p<0.0001) better than entropy (AUC=0.62, p=0.003) or skewness (AUC=0.62, p=0.003) alone. Abbreviations: area under the curve (AUC), cavernous angioma with symptomatic hemorrhage (CASH), dynamic contrast-enhanced quantitative perfusion (DCEQP); standard error (SE).

Figure 2. The best prognostic biomarker of bleeding/growth within a year after DCEQP imaging. (A) Weighted β_i coefficients of covariates (all p<0.05) in the prognostic biomarker; error bars are SE. One outlier (studentized residual>3) was excluded. (B) The biomarker (AUC=0.80, p<0.0001) predicted bleeding/growth better than mean permeability (AUC=0.65, p=0.019) or low-perfusion cluster mean (AUC=0.67, p=0.007) alone. Abbreviations: area under the curve (AUC), dynamic contrast-enhanced quantitative perfusion (DCEQP); standard error (SE).

Fig.3. Comparison of acute cerebral infarction patterns between young and elderly patients. The left and right pies above showed the percentage of different distribution patterns of acute cerebral infarction lesions in young and elderly patients, respectively.

Fig.5. Bar charts of vascular territory patterns of acute cerebral infarction lesions between young and elderly patients. Young patients had significantly higher percentage of lesions in both anterior and posterior circulations than elderly patients.

Figure 1. A detailed structure of the proposed 3D multi-scale ResNet.

Figure 4. Some exemplar results of the proposed 3D multi-scale ResNet. The drawn orange and blue circles refer to the truly detected lacune and non-lacune lesions, respectively.

Figure 1 | Brain regions show CT differences between BS and HCs. The thinner cortices are observed mainly in the bilateral frontal and temporal lobes. CT, cortical thickness; BS, basal ganglia stroke; HCs, healthy controls; L, left; R, right.

Figure 3 | Within-group global network measures, and between-group differences of these measures. Clustering coefficient (A,B),Transitivity(C,D) and Global Efficiency (E,F)of the BS and HC networks. The red star indicates the difference between the two groups (B,D,F). BS, basal ganglia stroke; HCs, healthy controls.

Figure 2. A. 3D visualization of CCM lesions in baseline (pink) and follow-up (green) scans for 2 patients. The lesions that are enlarged in the follow-up scan are highlighted with the blue circle while new lesions are denoted by an arrow. B. The enlargement of the lesion (blue circle) in the follow-up scan compared to the baseline is evident in the consecutive slices from a patient.

Figure 1. Pipeline for detecting the different sized lesions. (Top) Pipeline for small lesion detection. (Bottom) Pipeline for larger lesion detection. The potential lesion candidates are fed to the GUI where the user labels the lesions.

The straight sinus on both SWI and TOF

diameters

(a)Stroke images and labeling region; (b) GT, Prediction and Overlap in each mask.

The DSCs between the GT and the prediction in different ADC thresholds and different stroke sizes.

Figure 1. A. A cortical microinfarct visible on 7T T1-weighted MP2RAGE MRI scan. The CMI is not visible on B. 7T FLAIR MRI scan, C. 3T T1-weighted MRI scan and D. 3T FLAIR MRI scan.

Table 2. 7T and 3T MRI sequences parameters.

Figure 1. A 45-year-old man who had undergone radiation therapy after surgical resection of frontal lobe glioma at 26 years old. (A) FLAIR imaging shows peri-lesion white matter lesions. (B) and (C) SWI shows multiple cerebral microbleeds locating widespread the brain. (D) MRA shows multiple intracranial arteries stenosis. (E) Post-contrast 3D T1W HR-MRI shows eccentric plaques with moderate focal enhancement (E, arrow) in right M1 segment of middle cerebral artery. (F) Post-contrast 3D T1W HR-MRI shows prominent circle enhancement in bilateral V4 segment of vertebral arteries.

Figure 2. A 47-year-old woman who had undergone radiation therapy for pituitary adenoma at 36 years old. Pre-contrast (A) and post-contrast (B) 3D T1W HR-MRI show concentric wall thickening (A, arrow) with prominent circle enhancement (B, arrow) in bilateral V4 segment of vertebral arteries on coronal images. Pre-contrast (C) and post-contrast (D) 3D T1W HR-MRI show track train sign-concentric wall thickening (C, arrow) with prominent homogeneous enhancement (D, arrow) in right V4 segment of vertebral artery on axial images.

Figure 2. Neuropathologic characteristics.

Figure 1. Demographic and clinical characteristics.

Figure 2. Representative images from a patient with ischemic stroke showing the ability of brain temperature to differentiate regions after ischemic stroke. (A) Tissue regions defined from MR images include infarct core (purple), ischemic penumbra (blue), ipsilateral (green), and contralateral (yellow) healthy tissue; (B) Simulated brain temperature map before occlusion; (C) Simulated brain temperature map after 100% occlusion; (D) Brain temperature difference map (post- minus pre-occlusion). The green boundaries are major continuous regions of ischemic penumbra.

Figure 3. Predicted brain temperature evolution from onset to 8 hours after right middle cerebral artery occlusion in four regions: ipsilateral healthy tissue (ipsi-healthy), contralateral tissue (contra-healthy), infarct core (infarct), and ischemic penumbra (penumbra), where steady state brain temperatures were 37.21 ˚C, 37.37 ˚C, 37.49 ˚C, and 37.70 ˚C, respectively. These in silico results are consistent with previous in vivo studies and suggest that brain temperature may be used as a predictive biomarker to identify hemodynamically distinct regions.

Figure 3: Top images indicate DWI and FLAIR slices fed into each network to segment the penumbra seen in the bottom images. The dual network prediction indicates the subtraction of FLAIR segmented infarct from DWI segmented ischemic tissue. The multi-input prediction indicates predicted penumbra using both DWI and FLAIR as network input and the red line is an outline of the ground truth penumbra for this slice. Dual and multi-input Dice coefficients for this slice are 0.83 and 0.78 respectively.

Table 1. Dice Coefficients, sensitivity, positive predictive values, infarct/ischemic tissue/penumbra volume differences, and Spearman correlation coefficients, with 95% confidence intervals, for each network’s segmentation performance of either ischemic tissue (DWI), infarct (FLAIR), or penumbra (Dual Network and Multi-Input).

Man, 73 years old, Paroxysmal weakness of both lower limbs for 5 months. Figure A ,PWI raw image; Figure B , MRA showed right middle cerebral artery occlusion; Figure C ,Tmax 4s was 238.04ml, Tmax 6s was 113.60ml, Tmax 8s was 29.66ml, Tmax 10s was 4.42ml ( HCR=47.72% ; HIR = 3.89%).

Comparison between models

Figure 4 represents subgroup comparisons of the proportion of absence of FVH between symptomatic occlusion and asymptomatic occlusion. TIA: transit ischemic attack.

Figure 1 represents insular and M1-M6 regions for Fluid-Attenuated Inversion Recovery vascular hyperintensity Alberta Stroke Program Early Computed Tomography Score (FVH-ASPECTS) assessment. The illustrative case is scored as 3.

Figure 4. It shows the difference of image quality of each vein with different acceleration factors. SV: septal vein, ICV: internal cerebral vein, STV: superior thalamostriate vein, BV: basal vein, DNV: dentate nucleus vein

Figure 3. It shows the CNR and SNR of the image with the different acceleration factors. SV: septal vein, ICV: internal cerebral vein, STV: superior thalamostriate vein, BV: basal vein, DNV: dentate nucleus vein

Figure 2. Descriptive figures of the change of collateral perfusion and total SVD score along with the interval from symptom onset to imaging (A). Collateral perfusion decreased with the increase of total SVD score (B). Collateral perfusion was defined as (CBF 2.5s minus CBF 1.5s) at lesion side minus (CBF 2.5s minus CBF 1.5s) at normal side. SVD, small vessel disease.

Figure 3. Cross-sectional correlates of collateral perfusion in symptomatic ICA/MCA severe stenosis or occlusion. ICA, internal carotid artery; MCA, middle cerebral artery; SVD, small vessel disease. a, interval from symptom onset to imaging (days); b, total SVD score. * p < 0.05.

Figure 3. An interaction plot from the least square means of the mixed model to predict the TC of different registrations types in control vs patients.

Figure 2. A boxplot of the tanimoto coefficient of the 4 registration types in patients and controls

Lesion prevalence maps for the tumor (panel A) and stroke group (panel B) are shown superimposed on the MNI brain in radiological view. The colors refer to the number of patients with a lesion at that voxel, with red indicating a higher number of patients. The maximum overlap is 48 and 17 for the tumor and stroke group, respectively. The MNI brain on the right indicates the location of the slices shown in the figure.

Lesion symptom results for the Boston Naming Task. Lesion overlap indicating those regions in which at least 3 patients had a lesion for the tumor group in blue (N=172), the stroke group in pink (N=103), and overlapping regions shown in the green outlined purple area (Panel A). Voxels significantly associated with performance on this task for the tumor group (red) and stroke group (blue) with the green outline indicating overlapping lesion coverage, as shown in the above panel (Panel B). Maps are shown on the MNI standard brain in radiological view.

Table 1: rsFC analysis among memory related structures for DP>HC(pFDR<0.01) at different frequency bands.

Fig1: Weaker rsFC among medial temporal lobe structures in DP than HC

This retrospective study included 14 cases of untreated brain abscess, 20 cases of acute ischemic stroke and 26 cases of acute hemorrhagic stroke. One representative DWI was selected and randomly divided into two groups. Six radiologists (7 to 20 years of experience) reviewed one group consists of 30 DW image before any guidance. The interpretation of another group was carried out after the explanation about the characteristics of low intensity on DWI.

Data are presented as the; mean (range). Each diagnostic radiologist spent 14.8 min. for the first reading and the accuracy was 70.3%. The reading time after the explanation about the low intensity rim sign was 7.3 min. The accuracy was increased to 85.8 % (p = 0.025).

Figure 3: Comparison of laterality index (L_index) of LSA morphological parameters between the plaque and non-plaque groups. The L_Index of LSA branches and total length in the plaque group were significantly higher (*, p<0.05) than those in the non-plaque. There was no statistical difference in L_Index of LSA stems.

Figure1: Plaque group, a 69-year-old female with right lenticulostriate infarction (B) and non-stenotic MCA (A). C: HR-VWI shows the plaque on the upper wall of the MCA-M1 segment (red arrow). LSA skeletons (G-I, from D-E) are shown. The numbers of LSA stems and branches and the total length of LSA on infarcted and healthy sides were 4/6/125 mm and 3/4/89 mm, respectively. The calculated laterality indexes were 0.14/0.30/0.17, respectively.

Bar graph of total changes in arterial flow, circumference, pulsatility index and AVM volume before and after SRS. Ipsilateral artery flow was significantly greater than contralateral artery flow pre-SRS and decreased significantly post-SRS. Ipsilateral artery circumference was greater than contralateral artery circumference before and after SRS. Ipsilateral artery pulsatility was lower than contralateral artery pulsatility prior to SRS and increased significantly after SRS. Volume significantly decreased after SRS.

33 year old female with left posterior temporal-occipital AVM treated with a single-stage of SRS (1600 cGy). While T2W imaging (1A, G) and MRA (1B) demonstrate gradual decrease of the nidus from before SRS to 2 years after SRS, 4D Flow (1C, D, F) demonstrates a 61.2% reduction in feeding artery blood flow and 76.5% reduction in draining vein blood flow by 2 months, and normalization of feeding artery blood flow by 2 years. In comparison, there is only 2.3% reduction in feeding artery circumference with a more substantial reduction in draining vein circumference (50.6%) by 2 months (1E).

Figure 1. Schematic drawing of the virtual patient generation pipeline. Solid boxes: input; dash-dotted box: output; dotted boxes: quasi-patient-specific mesh generation; dashed boxes: model parameter tuning.

Figure 2. (a) Coronal view of the affine registration from the mesh to the patient space, with overlap (green), patient only (yellow) and mesh only (red). (b) Bland-Altman plot of the brain volume corresponding to virtual ($$$V_{vp}$$$) and real patients ($$$V_{rp}$$$). (c) Average (ave) CBF values in virtual ($$$F_{vp}$$$) and real ($$$F_{rp}$$$) patients. (d) Minimum (min) and maximum (max) CBF values in real and virtual patients.

Figure 1. (a) Angiographic image and (b) a corresponding centerline representation with calculated vascular depth. (c) The average flow rate waveforms obtained in the most proximal (depth<50mm) as well as in the most distal aspects (depth>250mm) of the visible vasculature. (d) The initial guess, 3rd and 33rd iteration of the estimated arterial waveform obtained from the minimization process. (e) Histogram illustrating the average number of cross sections available as a function of depth in the vascular tree.

Scan parameters of T1 SE, T1 SPACE, MPRAGE, ToF, T2 TSE, and SWI protocols used for the vessel wall imaging measurements at 7T.

Anatomico-morphological characterization. 3D reconstruction of arterial ToF sequence at 3T MRI (left panel) and 7T MRI (right panel) in a case of a suspected UIA. 7T revealed the presence of an infundibulary branch mimicking UIA.

Figure 1: Analysis workflow. Preprocessing (noise, phase offset, velocity anti-aliasing corrections) and generation of dual-venc phase-contrast MR angiogram workflow not shown. A – TOF MRA DICOMS imported for segmentation. B – 3D segmentation of Circle of Willis from in-house analysis tool. TOF is registered with PC MRA to extract 4D flow information. C – Vessel branch centerline extraction for further analysis. D – Placement of analysis planes along example vessel every 1 mm. E – View of single plane ROI delineating vessel contours, velocity map, and velocity vector profile.

Figure 2: Asymmetry index analysis results for PI and RI in ICAD patients with symptomatic MCA vessels. PI and RI comparisons to controls showed statistical significance (p < 0.05).

Figure 1: Segmented arteries by iCafe from two Philips 3T TOF-MRA (S/I FOV 7.2cm) 2 weeks apart. The different colors indicate different arteries as denoted by the labels (Left (denoted by suffix L) and right (denoted by suffix R) sided ICA- internal carotid artery, M1/A2 – Middle cerebral artery M1, M2 branches, A1/A2 - Anterior cerebral artery A1, A2 branches), BA – Basilar artery, P1/P2 – posterior cerebral artery P1, P2 branches, PComm – posterior communicating artery). Quantitative artery length measurements can be derived based on the centerlines from these iCafe segmentations.

Table 2. Comparison of scan-rescan reproducibility based on different protocol parameters on Siemens 3T

Figure 1. Measurements of vessel plaque characteristics. In a young patient (31 years old ) with an infarction in the left hemisphere (A), left MCA stenosis (B, arrowhead) was observed, A eccentric plaque in vessel wall (C, arrowhead) on high-resolution MRI was measured. In a old patient (68 years old ) with an infarction in the left basal ganglia (D), left MCA stenosis (E, arrowhead) was observed, A diffuse plaque in vessel wall (F, arrowhead) on high-resolution MRI was measured.

The vessel wall characteristics of plaque were summarized in Table 1. The demographic data were summarized in Table 2. P < 0.05 indicates a significant difference.

DSA (A) and TOF MRA (B) show an aneurysm at the left SCA and PCA junction (blue circle). Followup DSA (C) after basilar/left PCA flow-diverting stent placement demonstrates complete occlusion. MRA demonstrates questionable flow-related enhancement not easily discriminated from T1 shine through from organizing clot (D). Pre (E) and postcontrast (F) VWI demonstrate preserved high-resolution black blood angiography across the stented PCA segment (dashed arrow). Organizing clot within the aneurysm sac (orange circle) is noted, likely fed through a vasa vasorum, without flow voids.

Comparison of patient characteristics and vessel wall imaging findings between subjects with and without complete aneurysm occlusion of followup angiogram after flow diverter placement.

Figure 2. Illustration of tensile test setup (a), The stress-strain data (blue) was fitted with a linear function (black) to obtain the elastic module (material-2 at a = 0°) as a result of 0.50 MPa.

Figure 4. Elastic modulus of samples with respech to the real artery. Elastic modulus of all samples (b). As shown in the graph and the table to the graph elastic modulus values are in the range except material-2/Sample-1/45⁰; material-2/Sample 1 and 3/90⁰.

Figure 2：Hyper signal could be seen in the right corona radiata,so the MCA plaque is a symptomatic plaque(A).

MRA showed the mild stenosis of the right middle cerebral artery.(B)

The maximum wall thickness as measured using the MRI-PlaqueView software at the leision slice of the right middle artery was 0.81 mm.(C)

The plaque distribution range measured using RadiAnt DICOM Viewer software at the leison slice of the right middle artery was 116°.(D)

Figure 3：Hyper signal could be seen in the left cerebral peduncle, so the plaque was a symptomatic plaque(A).

The plaque distribution range at the leision slice of BA was 136°.(B)

The maximum wall thickness and the minimum wall thickness for the lesion slice were 2.07 mm and 0.41 mmrespectively (C and D)

Figure.1 A, MRA. B, Enhancement on aneurysm (large arrow) and pituitary infundibulum (small arrow). C, Corresponding image of aneurysm on MR-VW postgadolinium T1WI (left), TOF (inset, middle) and 4D Flow MRI (inset, right). D, Normal vectors of vessel wall. E, Joint visualization of MR-VW postgadolinium T1WI and vessel model. F, Distribution of enhancement values on low enhanced area (black) and high enhanced area (polychrome). G, WSS distribution. H, OSI distribution. The high enhanced area (large arrow) had lower WSS and higher OSI than low enhanced area (small arrow).

Figure.2 Scatter plot of distributions of WSS (A) and OSI (B) on low enhanced area (blue dots) and high enhanced area (red dots) of aneurysm located at the end of M1 of middle cerebral artery.

Figure 1: DSA, silent and TOF MRA images of a 14-year-old male patient with a left frontal CAVM. DSA anterior(a) and lateral (b) projection from the left internal carotid artery reveals AVM nidus (*) fed primarily by the left anterior and middle cerebral artery (arrowhead) and its drainer (arrow). The nidus (asterisk) feeders (arrowhead) and drainer (arrow) are better delineated on Silent MRA images (c,d) than the TOF MRA ones (e,f ).

Figure 2: DSA , silent and TOF MRA images of a 33-year-old female patient with a right basal ganglia CAVM. DSA anterior(a) and lateral (b) projection from the right internal carotid artery reveals AVM nidus (*) fed primarily by the right anterior and middle cerebral artery (arrowhead) and its deep venous drainer (arrow). The nidus (asterisk) and drainer (arrow) are better delineated on Silent MRA images (c,d) than the TOF MRA ones (e,f )

Figure 2. Representative case with BOD. A. Longitudinal plane of MCA with fusion image (PD+T1CE). B. Cross-sectional image where perforator arouses. Perforator (red arrow), plaque (white arrow) C. Plaque distribution was expressed as an angle. Perforator was located at margin of the plaque.

Figure 1.Diagram of patient selection.

Figure 1：(A): MRA demonstrated smooth wall and no obvious stenosis. (B): Head MRI showed abnormal signals in the left oculomotor nucleus. (C-E): 3D VW-MRI showed unevenly thickening and abnormal enhancement of basilar artery，bilateral posterior cerebral artery and middle cerebral artery. (F-G):Curved plannar reconstruction by 3D-VW-MRI showed the left and right middle cerebral artery had no apparent enhancement after 4 months therapy, while the vessel wall was getting thickness.

Figure 2：(A):MRA revealed multiple stenosis of both the anterior circulation and the posterior circulation. (B~H)3D VW-MRI showed smooth, concentric arterial wall thickening and enhancement of bilateral vertebral artery, the left internal carotid artery and middle cerebral artery.

Figure 1. Typical images of the right internal carotid artery(R-ICA). (A)(B)(D)HR-VWI shows an isointensity plaque on T1WI, T2WI and TOF sequences on the right internal carotid artery wall (arrow) without postgadolinium enhancement(E). (C)3D fusion image shows that maximum lipid-rich necrotic core percentage > 40%. (F)Pathological result shows that a large amount of lipid components in the plaque.

Figure 1. A schematic diagram of the selected areas(yellow) of part 1(left, Random irregular regions) and part 2(right, axial direction of special vessels).

Figure 2. The registration results from representative slices are shown.

Figure 2. Reconstruction of cervical vessels, from left to right: default, CS4-CS8.

Figure 1. Male, 68 years old. Use ROI to measure SI and SD on both sides. The measured left blood vessel SI value was 2485.45, muscle SI value was 523.2, and muscle SD value was 24.3. The measured right blood vessel SI value was 2152.71, muscle SI value 566.59, and muscle SD value 20.49. The largest and smallest diameters of internal carotid arteries on both sides were measured. The largest diameter measured on the right side is 7.13mm, and the smallest diameter is 6.25mm. The maximum diameter on the left is 6.76mm and the minimum diameter is 6.25mm.

A) Whole-brain CBF pre and post GAHT treatment in participant 1.

B) Whole-brain CBF pre and post GAHT treatment in participant 2.

C) Percent change in average whole-brain CBF. Note that percent change in CBF appears to be of the same magnitude as the change in Serum T shown in D.

D) Percent change in Serum T.

A) 3D projection of segmented vascular tree, with diameter values in participant 1. White arrows indicate regions where TOF-Voxels decreased after GAHT.

B) 3D projection of segmented vascular tree, with diameter values in participant 2. White arrows indicate regions where TOF-Voxels decreased after GAHT.

C) Percent change of TOF-Voxels. The decrease of TOF-Voxels is of the same magnitude as the decrease in CBF.

D) Change in arterial diameter in small arteries. Note that across all groups of small vessels there was a decrease in the number of voxels with in each diameter class.

Table 4. Univariate and multivariate logistic regression of clinical factors associated with the presence of plaque

Figure 1. (A) The carotid vessel wall images of a 62-year-old male patient with 6 months of hemodialysis. (B)The carotid vessel wall images of a 58-year-old male patient with 62 months of hemodialysis;