Figure 1. Representative black-blood thrombus imaging (BTI) images from four patients who were categorized as presenting with (a & b) iso- and (c & d) hyper- intense thrombus signals, respectively. The iso-intense thrombi (yellow arrows) had comparable signal intensity to the adjacent muscle while the hyper-intense thrombi (red arrows) had higher signal intensity than the adjacent muscle. When both iso- and hyper-intense thrombi were detected in a patient, the patient was then assigned to the group that corresponded to the dominant thrombus signal (b & d).

Table 2

Figure 1: Ferumoxytol-enhanced MRA is safe and radiation-free alternative form of imaging for pregnant patients to exclude pulmonary embolus. Shown are ferumoxytol-enhanced 3D MRA images of 31 year old pregnant patient in 33 gestational week acquired in free breathing, in sagittal (a), axial (b) and coronal (c) view.

Figure 2: Ferumoxytol-enhanced MRA enable visualization of pulmonary thrombus. Image acquisition is performed in free breathing. Extended blood pool contrast allows for images that can be acquired in an extended temporal window and repeatedly, without need of additional contrast agent administration. Shown are 3D MRA images of 22 year old pregnant patient in 7 gestational week, in coronal (a), axial (b) and sagittal (c) view.

Figure 1. Overview of the different reconstructions compared in this work. All reconstructions rely on the same SIMBA clustering of the free-running data. A) Standard SIMBA consists of a non-uniform fast Fourier transform of the data in the most populated cluster, B) XD-SIMBA consists of sorting the four most populated clusters based on the SSIM prior to performing an XD-GRASP reconstruction along the cluster dimension, C) XD-SIMBA followed by patch-based low-rank denoising and D) standard SIMBA followed by the same denoising algorithm applied to the four most populated clusters.

Figure 2. Representative image quality from one pediatric and one adult patient. A) The image characteristics differed between the various image reconstruction methods. Methods that share information between clusters generate less noisy images than the simple SIMBA reconstruction. B) In this adult patient, the aortic valve is depicted more clearly in the XD-SIMBA-LR images compared to the other reconstruction types as seen in the zoomed section. Finally, the cross-sections of the LAD and LCX coronary arteries appear more conspicuous in the XD-SIMBA-LR approach.

Figure 1. Overview of the 3-step auto-focusing framework for reconstructing cardiac and respiratory motion-compensated 3D whole-heart CMRA images from free-running data.

Figure 3. Reformatted images of the right coronary artery (RCA). Reformatted images of the RCA obtained with Soap-Bubble are shown for three representative volunteers for the reference 5D reconstruction (A) and for the proposed fNAV-based reconstruction (B). Arrows indicate RCA branches (yellow) or segments (green: beginning of the segment, red: end of the segment) for which the fNAV-based reconstruction approach allowed for improved detail visibility.

Figure 1 A. the proximal coronary artery has clear and smooth boundary (4 points)；B. the proximal coronary artery can be shown with moderately obscure boundary (2 points).

Table 1. Number of coronary segments with different image quality scores

Magnetization transfer with an inversion pulse is used in odd heartbeats (A), magnetization transfer solely is exploited in even heartbeats (B). A short TI is used for fat saturation in odd heartbeats, spectral presaturation (Fat Sat) is used in even heartbeats. Data acquisition is performed using a 3D Cartesian trajectory with spiral profile order . A low‐resolution 2D iNAV is acquired in each heartbeat. The bright‐blood MTC‐IR BOOST and MTC BOOST volumes are non‐rigidly motion corrected and combined in a PSIR‐like reconstruction to generate a complementary black‐blood volume

Bright blood images for a patient with congenital aortic stenosis, post Ross procedure with moderate regurgitation along the pulmonary homograft. Luminal signal loss in the pulmonary arteries due to turbulent flow and in the pulmonary veins and left atrium due to off resonance effects is observed with the clinical 3D sequence. MTC-BOOST demonstrates uniform signal in all cardiac and vascular structures.

Figure 2. Example images generated by the proposed MT-MACS from a 38-year-old healthy subject. Based on the dual-echo acquisition scheme, MT-MACS can achieve water/fat separation and provide water-only and fat-only images. For water-only images, three out of 300 image contrasts are selected for bright-blood, dark-blood and gray-blood imaging, respectively. For each image contrast, corresponding cardiac phase-resolved cine series can also be generated to assess the cardiac function. Coronal and transverse views of the fat images, as well as cine fat series, are also displayed.

Figure 1. Pulse sequence diagram for MT-MACS and corresponding k-space sampling pattern for auxiliary data. T2-prepared inversion recovery (T2IR) magnetization preparations are applied at constant intervals followed by dual-echo FLASH readouts. RF pulse flip angles following each T2IR are 3⁰ for the first 300 segments, and 1⁰ for next 200 segments to allow for greater magnetization recovery. Auxiliary data are interleaved with imaging data every 6 readouts and are collected at the 0⁰ radial spoke of the center partition.

Figure 1. Transversal MIP images of the brain showing the appearance of resolution-matched qTOF, qQISS, standard Cartesian TOF, and 3D phase contrast (PC) MRA. Note the excellent correlation of arterial anatomy of qTOF and qQISS with respect to standard TOF MRA, and the superior SNR with respect to 3D PC MRA. qTOF and qQISS obtained using root-mean-square combination of TE₁ and TE₃. Insets show that flow misregistration artifact seen with Cartesian TOF MRA (arrows) is reduced with qTOF and qQISS.

Figure 2. Transversal MIP images showing mean cross-sectional velocity maps corresponding to the qTOF and qQISS angiograms shown in Figure 1. Note the progressive reduction of blood flow velocity from the larger to smaller arterial branches.

Figure 3: A representative case with heterogeneous intramural hematoma (A, heterogeneity = 0.22) and a case with homogeneous intramural hematoma (B, heterogeneity = 1.00) on three-dimensional vessel wall MRI images.

Figure 2: Odds ratio and 95% confidence interval of the acute ischemic stroke group versus the non-stroke group based on univariate and multivariate logistic regression for each feature.

Figure 1. The schema diagram of BRIDGE with pulse gating.

BRIDGE consisting of 3D MSG-EPI with T2prep and IR pulses (A).

The estimated signal transitions of different tissue types in the thoracic aorta (B) show that data acquisition after a short inversion time (TI) yields bright blood (MRA) with suppressed background tissues such as muscle, and data acquisition after a longer TI yields dark blood images with suppressed vascular signals and enhanced vulnerable plaques.

Figure 4. Optimal image processing.

All images in each phase can be assessed in multiple planes with MPR processing. Bright blood is used for assessment of vascular anatomy and vascular lumen. Dark blood such as gray or black blood is used for assessment of vessel wall and plaque characterization and the extent of vulnerable plaques.

In MIP process, creation of vulnerable plaques-weighted MRA image allows for simultaneous assessment of vascular anatomy and plaque location. This image can be used as a surgical support image such as CAS.

Figure 1. Coronal limited maximum intensity projection images for the different total image quality scores: 2 (A)；3 (B)；4 (C)；5 (D).

Figure 2．Coronal limited maximum intensity projection images for the different main, right, and left pulmonary arterial trunkscores：2 (A)；3 (B)；4 (C).

Figure 1. Coronal image showing native aorta and graft region, as well as corresponding picture. The figure also shows locations where cross-sectional phase-encoding images were acquired.

Figure 2. (a) Magnitude images showing a cross-sectional area in the graft region (arrows), which grows with time during systole. (b) Velocity-encoded phase-contrast images showing a cross-sectional area in the graft region (arrows), where flow increases (blacker region) during early systole.

Figure 1) Turbulence (orange arrows) visualized with 4D flow in a glass phantom (water, 25°C, 1.3ml/s, VENC=40 cm/s)

Figure 2) Turbulence (orange arrows) visualized with injected ink in a PVA phantom by a camera (time frame top to bottom, 30fps, 320x240, 33% glycerol solution, 25°C, 0.85ml/s).

Figure 1. Strain curves in different rabbit models. Circumferential strain curves throughout the cardiac cycle in different segments of mid-ventricular short-axis slices. Strain is shown positive as peripheral gating was used with peak systole during first timeframes. Note decreased strain values with increased degree of coarctation. Strain values are larger than those in untreated CoA (scale is different for different panels).

Figure 2. Velocity curves in different rabbit models. Mean velocity throughout the cardiac cycle in ascending aorta (AAo) and proximal descending aorta (DAo) in different rabbit models of CoA. The ratio between maximum velocity in AAo to that in DAo was slightly >1 in control and treated CoA compared to untreated CoA, where the ratio was always >2 in the latter.

3D rendering of the QUTE-CE image capturing the abdominal vascular anatomy shown in anterior and posterior view.

3D rendering of the vasculature of the liver (A) and the kidneys (B) cropped from the same QUTE-CE image shown in Figure 2.

Results of the blinded image quality scores (A, B) and quantitative change in CNR (C) for each CS-SENSE factor. RCA visibility and the number of visualized features was not significantly affected by CS-SENSE acceleration, while CNR was significantly decreased using a CS-SENSE factor of 6.5 (p<0.05).

Representative MPRs of the right coronary artery in 5 patients with SENSE and CS-SENSE factors of 2.5 – 6.5 with progressively shorter scan times and degraded image quality.

Figure 1. Compared the TOF-MRA and TOF-RADIXON of three healthy volunteers in the source image. Left column, indicates signal loss in the vascular lumen (blue arrow). Mid column, breathing fat motion-artifact shows that the artifacts overlap the blood vessels (orange arrow). Right column, indicates loss of vascular signal due to respiratory motion-artifact (red arrow).

Figure 2. Compared the TOF-MRA and TOF-RADIXON of three healthy volunteers in the MIP images with a background cut by post-processing. In the TOF-MRA, due to the fat in the chest wall, it overlaps the MIP image as a white haze compared to the TOF-RADIXON.

Figure1. The representative image of the origin and distribution of LFCA; ABOLFC, ascending branch of the LFCA; TBOLFC, transverse branch of the LFCA; DBOLFC, descending branch of the LFCA.

Figure2. the surface measurement signs of the perforator superficial point of the descending branch of the LFCA

Figure 4. Representative MIP images of (a) SPIR T1TFE, (b) bTFE DIXON without HISSCO, and (c) bTFE DIXON with HISSCO applied. Multiple slice swaps observed below the knee in the original bTFE DIXON (arrow) were all corrected by HISSCO. Note that the most distal swap in (b) happened only on the right side of the leg but nicely corrected in (c).

Figure 3. Comparison of the water images of (a) bTFE DIXON without HISSCO and (b) bTFE DIXON with HISSCO applied.

Figure 3. Example MIP images (upper rows) and their zoomed-in views (lower rows, from the red dashed boxes) with different resolutions from two healthy volunteers. The yellow arrowheads denote the improved depiction of small arterial branches, and the blue arrowheads denote the slight signal loss in the large artery on the image with very high resolution (512×512).

Figure 4. Box-plots of quality scores of FBI images with different resolutions: (A) subjective evaluation of large vessel depictions, (B) subjective evaluation of small vessel delineations, (C) objective evaluation of sharpness and (D) objective evaluation of CNR of artery-to-background. * denotes statistical significance between the high-resolution images and the standard-resolution images (256×256). (Wilcoxon signed-rank tests for A and B; paired t-tests for C and D. P<0.05).

Figure 1 – 69-year-old man with PAD. Corresponding standard 2D-QISS MRA (A), 3D tsSOS-QISS MRA (B), and CTA (C) are shown. Total occlusion in the right distal superficial femoral artery and popliteal artery (yellow brackets) and significant stenosis in the left superficial femoral artery (yellow open arrows) are visualized. Stair-step artifacts can be seen on 2D-QISS (A, green brackets), which are not present on the corresponding 3D tsSOS-QISS (B, green brackets).

Figure 5. Representative REPI 4D MRA of the foot (left) and the hand (right). REPI 4D MRA could visualize blood vessels from proximal to distal with high robustness. The optimal VENC for visualizing peripheral blood vessels was 10cm/s.

Figure 4. Peripheral vessels of the foot. REPI VENC=10cm/s, TFE VENC=10cm/s, 3D-PCA VENC=10cm/s, 2D-TOF used for comparison. REPI 4D-MRA significantly improves vascular structure in AREA2,3,4, compared with the other three methods. Visualized.

Figure 4. Representative full MIP (left) partial MIP (center) of REACT-MRA and MPRAGE (right) images from the same REACT-MD dataset. It is noteworthy that REACT-MD provided both MRA and MPRAGE images in one single scan and allowed for imaging over a large FOV at 3.0T.

Figure 2. A comparison of the procedural time scheme using the proposed approach to the conventional method in the practice. (a) conventional procedure included survey, planning and separate scans for REACT and MPRAGE images. REACT-MD can shorten the scan time for REACT and MPRAGE (b). Furthermore, fast planning-less approach (c) further reducing the total examination time.

Figure 1: Multi-contrast images (CTA: computed tomography angiography, UTE: ultra-short TE, T1-weighted, T2-weighted) demonstrating excellent contrast between PAD lesion components in popliteal (first and second row, red arrows) and peroneal artery (third row, red arrows). Circumferential calcifications, as well as occlusions, can be readily appreciated due to the intrinsic high spatial resolution achievable at the 7T scanner compared to the CTA. No CTA available, 2nd row. (Red stars: Occluded PTFE grafts)

Figure 2: 3D UTE microMRI (A) and corresponding microCT (B) image of the excised lesion demonstrating good agreement between hypointense area on the micro MRI and the circumferential calcification identified with the microCT. Pseudo-color image (C) created from the 7T images illustrate the differentiation power of the near in-vivo 7T to differentiate lesion components also in respect to the ex-vivo microCT.

Figure 2: Comparison of morphology of different femoral artery segments between males and females with plaque. CFA: common femoral artery; pSFA: proximal superficial femoral artery AC: adductor canal PA: popliteal artery.

Figure 3. Comparison of morphology of different femoral artery segments between males and females with plaque with plaque

Figure 5: First gadoteridol injection (3 mL/s - left Figure 2) plots predicted (blue) vs observed (orange) which integrates R₁ and R₂* effects with excellent correlation between predicted and observed SI. The middle graph shows the same predictive model but excludes R₁ effects, with decreased accuracy between predicted and observed SI. The right graph shows the same predictive model without both R₁ and R₂* effects, which is the least accurate construct. This is consistent across all agents and injections.

Figure 2: Three consecutive injections of gadoteridol at 3 (left), 1 (middle), and 2 (right) mL/s with predicted (blue) versus observed (orange) SI line plots. Relative gain between observed and predictive SI was adjusted for the first injection baseline, and held constant for the subsequent two injections. Blood R₁ and R₂* effects were integrated into the model, with excellent correlation between predicted and observed SI across all three injections.

Figure 2: Comparison of different 3D-MERGE scans on a volunteer (coronal view). (a) single-echo 3D-MERGE with composite hard pulse iMSDE and SPIR fat suppression. (b) single-echo 3D-MERGE with adiabatic iMSDE and SPIR fat suppression. Water (c) and fat (d) images as well as fat fraction (e) and field (f) maps are decomposed from dual-echo 3D-MERGE with adiabatic iMSDE. Note the signal intensity uniformity and fat suppression difference in those red arrow pointed regions in (a)-(c).

Figure 3: Comparison of different 3D-MERGE scans on a volunteer (axial view). (a) single-echo 3D-MERGE with composite hard pulse iMSDE and SPIR fat suppression. (b) single-echo 3D-MERGE with adiabatic iMSDE and SPIR fat suppression. Water (c) and fat (d) images as well as fat fraction (e) and field (f) maps are decomposed from dual-echo 3D-MERGE with adiabatic iMSDE. Note the signal intensity uniformity and fat suppression difference in those red circle regions in (a)-(c).

Figure 2. The NC, IPH and LM were identified and quantified on matched histological section and MR imaging

Figure 4. The signal intensity threshold ratio of plaque components referred to reference tissue

Table 1. Imaging parameters of free-breathing MSDE-prepared 3D golden-angle radial stack-of-stars (3D Vane-MSDE) and 2D TSE MultiVane XD (MVXD-MSDE) used in this study. Conventional Cartesian 3D TSE with MSDE-preparation and motion compensation was performed for comparison (3D TSE-MSDE).

Figure 2. Free-breathing black-blood imaging with proposed 3D Vane-MSDE and MVXD-MSDE. Homogeneous blood suppression was achieved by MSDE preparation in both aorta (yellow arrows) and inferior vena cava (arrow heads). Images were selected at four different levels from top of the kidney to aortic bifurcation in a healthy volunteer. Conventional bright-blood 3D Vane and T2 MVXD images at the corresponding slice locations were shown for comparison.

Figure 2 – A. The cuff is proximal to the superficial femoral artery and vein. B. Artery lumen (A_t) is measured before cuff-occlusion and post-cuff release, to evaluate flow mediated dilation. C. Venous saturation (SvO₂) is measured continuously post-cuff release (acquisition is interrupted 3 times to measure A_t), using the phase difference between vein (v.) and artery (a.) (SvO_2b=pre-cuff SvO₂; T_w=washout time). D. Arterial blood flow velocity averaged across the artery lumen, plotted vs time. (V_P=peak velocity; HI=blood acceleration; T_P=time to peak; T_FF=hyperemia duration).

Figure 1 – Demographics, details on smoking/vaping history and exposure, and gender distribution in the three groups.

Fig.3 This 68-year-old woman had an anterior communicating artery aneurysm and was treated with surgical clipping. Maximum intensity projection (MIP) images obtained by TOF-MRA (A) and uTE-MRA (B). The volume rendering image reconstructed from computed tomography angiography (C).

Fig.4 This 46-year-old woman had an anterior communicating artery aneurysm and was treated with coil embolization. Axial images and maximum intensity projection (MIP) images obtained by TOF-MRA (A, B) and uTE-MRA (C, D). Neck remnant was confirmed in follow-up MRA images obtained 6 months after coil embolization (arrows). DSA images were obtained in anteroposterior(E) lateral(F) views. Neck remnant was also observed in the DSA images taken after MRA imaging (arrows).

Figure 4: Two Circle of Willis schematics demonstrating normal and abnormal blood flow directions. Schematic on the left shows all normal vessel flow directions (blue); schematic on the right shows abnormal flow direction in the left PComm (red).

Figure 1: Whole brain dMRA (temporal resolution: ~100 ms) scan obtained with iSNAP dynamic MRA. Shown above are 9 out of 21 frames in a scan.

Figure 1. Manual regions of interest (ROIs) were placed on the CBF map by using the aligned anatomical image as guidance. ROI was about 20-100mm². (A) bilateral superior frontal gyrus and posterior central gyrus.(B)bilateral superior temporal gyrus.(C)bilateral occipital lobe.(D)basal ganglia (bilateral thalamus, globus pallidus, putamen, caudate nucleus)

Figure 4. Regions of positive correlation between MDI、PDI and CBF in preterm infants aged 60 days-4 years of age. MDI (pink): left frontal、occipital lobe and caudate nuclear. PDI (blue): left frontal and occipital lobe.

BTI images of a 24-year-old woman with multiple thrombosed segments of varied stage. Coronary BTI image showed isointense signal indicative of acute thrombus in right sigmoid sinus (red arrows in A). Coronary BTI image showed hyperintense signal indicative of sub-acute thrombus in right jugular vein (red arrows in B). Sagittal BTI image showed isointense signal with flow voids indicative of chronic thrombus (red arrows in C) in right sigmoid sinus.

BTI images of a 13-year-old female with suspected CSVT. Sagittal, coronal, axial images of BTI (A, B, and C) and corresponding color graphs made by commercial software (Object Research System, Montreal, Quebec, Canada) depicted the hyperintense signal of subacute thrombi in Inferior sagittal sinus, vein of Galen, straight sinus, right sigmoid sinus and right transverse sinus.

Figure 3: Representative reconstructions of aortas demonstrating differences in plaque accumulation by diet. A representative aorta from each group is presented, Control Diet (A), High Fat Diet (B), and Ketogenic Diet (C). White coloring represents the lumen of the area, dark pink denotes the atherosclerotic plaque.

Figure 2: Plaque burden varies by dietary macronutrient content in the brachiocephalic artery (A), the aortic arch (B) and total aorta (C). Data shown are the mean ± SEM, n = 4–8/group; bars not sharing not sharing letters differ by p < 0.01.

Figure1. Segmentation as well as the centerline of the vessel were acquired. The measurements included tortuosity index (TI), bending length (BL), and inflection count metric (ICM). (A) Left ICA vessel segmentation of a young subject; (B) TI is actual length over direct length; (C) BL is the distance between the foremost points on the vessel path and the start-end points link; (D) ICM is the product of turning points number and TI. (E-H) are corresponding vessel segmentation and measurements of an old subject.

Figure 2. Illustrations of vessel segmentation for each decade over the lifespan. (A) Coronal TOF images of subjects from each decade demonstrate arteries become more tortuous with aging. (B) Upper: left ICA segmentation results of subjects from 20 to 80; Lower: right ICA segmentation results of subjects from 20 to 80.

Figure 4: Dynamic MRA (dMRA) images obtained by (a) iSNAP-HS and (b) iSNAP-FOCI of healthy volunteer 2. Readout excitation thickness is 137.5mm, small/large inversion slab thickness is 154/500mm. Fat related artifacts are totally removed in the first frames.

Figure 5: Static MRA (sMRA) images obtained by iSNAP-HS and iSNAP-FOCI of healthy volunteer 1. Readout excitation thickness is 137.5mm, small/large inversion slab thickness is 137/500mm.

Fig.2 MIP images of a 32 years old healthy man (a, b). TOF-MRA reveals signal loss in the cavernous segment of the internal carotid artery, but the signal is uniform in the same position for PETRA-MRA(box). MIP images of a 41 years old healthy woman (c, d). In addition to the loss of signal, TOF-MRA also shows a coloboma in the vascular profile, but in PETRA-MRA, it is smooth and complete (arrow). Source images of a 55 years old healthy woman (e, f). Low signal intensity occurs in the center of the lumen and part of the boundary in TOF-MRA, but it is not visualized for PETRA-MRA (double-headed arrow).

Fig.1 All points, median, 25% quartile, 75% quartile, the minimum, and the maximum value of 37 subjects on SNR, CNR, uniformity, as measured by Reader 1 and Reader 2, are shown in the box-and-whisker plot. The plot reveals higher SNR and CNR, lower CR and uniformity in TOF-MRA compared with PETRA-MRA. The differences between the two MRA images in these four indicators are statistically significant (all P < 0.001, which is indicated by * in the plot).

Figure 2. Pipeline of the proposed segmentation method (1) Perform maximum intensity projection along the temporal dimension of dMRA to generate tMIP image volume. (2) Enhance columnar structure with Frangi filter. (3) Use Frangi filter output to enhance dMRA. (4) Regenerate tMIP from the enhanced dMRA. (5) Use threshold method to yield initial segmentation. (6) Iteratively expand the segmentation based on spatial-temporal correlation between voxels. Light color in the bottom images indicates new vessel pixels found by the iteration. (7) Cleanup small connected components.

Figure 3. Vascular segmentation results by the proposed method on volunteer #1 (female, 24 years). The images in the upper row are MIP of tMIP along feet-head direction. To better distinguish the vessels, the depth images calculated from MIP of the final segmentations, instead of the binary MIP, are shown in (e) and (f). (d) shows the extracted and smoothed centerline obtained from the final segmentation. Example distal vessels extracted by correlation-based method but not by the threshold method are indicated with arrows.

Figure 5: Coronal MIPs of the sub-volume containing the lenticulostriate arteries in the same reconstructions as in Figure 4. Green arrows indicate examples of improved visibility and sharpness of the arteries when using optimized sampling parameters, for both acceleration factors.

Figure 4: Axial MIPs of reconstructions from fully-sampled and prospectively undersampled datasets. Undersampled datasets are shown for R=7.2 and R=15 with optimized (calib = 12, pp = 2.0) and literature-based (calib = 32, pp = 2.4) undersampling parameters. (a) Axial MIPs of the entire imaging volume. (b) Close-ups of the MIPs (blue box in fig. (a)) for improved visibility of smaller vessels. Especially for R=15, the optimized acquisition shows improved visibility (green arrows) and sharpness (white arrows) of the smaller vessels.

Table 2

Table 3

Table 2. Logistic regression analysis of risk factors for postoperative cerebral hyperperfusion.

Receiver operating characteristic (ROC) curves of the preoperative Fazekas score for white matter hyperintensities (WMHs) (A) and the number of lacunes (B) for prediction of cerebral hyperperfusion (CH) after surgery. (C) ROC curves of preoperative radiological markers (Fazekas score of WMHs and the number of lacunes combined with the presence of carotid near-occlusion) for prediction of CH after surgery. (D) Relationships between the Fazekas score of WMHs, the number of lacunes, carotid near-occlusion, and postoperative CH.

Figure 1: The schematic diagram and image quality of two ZTE-MRA sequences. Schematic diagram of the two MRAs (a). VR and MIP images for cASL-ZTE-MRA (c,d) and hASL-ZTE-MRA (e,f). Line chart of vessel contrast on hASL-ZTE-MRA (f)

Figure 3: A stent-assisted coil embolization using an Enterprise stent. Coiled aneurysm and residual contrast on DSA (a, b). The flow signal in the stent and coiled aneurysm on cASL-ZTE-MRA (c, d) and hASL-ZTE-MRA (e, f).

Figure 1. A 33-year-old male patient with bilateral MMV after bilateral bypass surgery. The vasculopathy was observed approximately 35 months after the surgery was performed on the right side. Intracranial collaterals originating from the right external carotid artery were clearly shown on sagittal 4D-sPACK (b), consistent with DSA (a); however, they were not clearly observed on axial, coronal, and sagittal images (c–e) obtained on TOF MRA because of overlap with other vessels. The anastomosis was shown on axial TOF MRA image (e, red arrow) and axial 4D-sPACK (f–i).

Table 1. Results for anastomosis patency in 4D-sPACK and 3D TOF MRA images using DSA as the gold standard.

Figure .5. Comparison between the conventional UTE 4D MRA and VTI UTE 4D MRA. Note that different time scales were used to display the full range of information in each dataset and the times shown are relative to start of labeling. (a) VTI UTE 4D MRA with 6 phases. (b) Conventional UTE 4D MRA with 12 phases. The sequence chart is as shown in Figure 3. VTI UTE 4D MRA better image quality comparing to conventional method without reducing structural information. The late phase of image with VTI method has better visibility of arteries compared to conventional method.

Figure .3. The graphical sketches showing UTE 4D-MRA pulse sequences used in the phantom and volunteer study. (a) Variable TI UTE 4D-MRA, where the TI intervals were increased with a manner of progression of difference, TI = 200, 480, 840, 1280, 1800, and 2400 ms, (b) conventional UTE 4D-MRA, 12 readout blocks from 200 to 2400 ms each blocks having 200 ms of delay time from the previous blocks. Each readout block consists of 53 series of UTE acquisitions. The data from each readout blocks were used to construct a single image.

Fig 2. The maximum intensity projection of the reconstruction images are compared. Flow void can be found around bifurcation (red arrow). Inhomogeneous signal distribution can also be found in the upper stream of carotid arteries (yellow arrow). Temporal Harmonic Encoding helps reduce these artifacts and the quality becomes better as $$$ R_p $$$ increases.

Fig 3. The images are from the root-mean-square of the harmonic components in the two Temporal Harmonic Encoding experiments. The bright signals indicate region with high motion contents. Although these images also contains artifacts due to parallel imaging decomposition, the dynamic information in the image may still serve as reference.

Fig. 4. Combine-model nomogram for the assessment of vulnerability plaque. A. The combine-model was visualized through linearized image, which combined radscore and LHR in training group. The AUC showed the diagnostic performance among radscore, LHR and nomogram in training group (B) and testing group (C).

Fig. 3. Diagnostic performance of univariate logistic regression model. A and B show the ROC curves of the radscore that the AUC is higher than LHR in the training and testing groups.

Figure 1. A patient with R-MCA undergoing stenting before (top row) and after (bottom row) surgery. A 65-year-old male with a severe R-MCA stenosis. a. DSA：after stenting, the stenosis of the R-MCA was significantly alleviated, the blood flow was recanalization.b. Preoperative ASL ：low perfusion area was visible in M6, postoperative low perfusion basically returned to normal level. c.d. Preoperative tASL ： the perfusion in M3, M5 and M6 areas was lower than the contralateral side, and the postoperative hypoperfusion area basically returned to normal level.

Fig 2 The ROC curve shows the sensitivity and specificity values for predicting good clinical outcomes with the ASPECTS score of ASL and tASL at various levels of cut-off scores. Predictive value of good prognosis based on the ASPECTS score in postoperative tASL technology(AUC=0.829；95%CI:0.651-1.000 P<0.05) higher than ASL (AUC=0.760；95%CI:0.574-0.946 P<0.05).

Figure 1. A workflow used here to design a thin-film based stent for intracranial aneurysm treatment: after understanding the problem (1) and current solutions (2), a stent was designed (3) and tested virtually (4). Using the resulting data, the design was improved and fabricated (5), then evaluated experimentally with 4D flow MRI in vitro (6). Again, the results were used to improve the design that leads the hitherto final stent (7).

Figure 3. Visualization of flow patterns in aneurysm sac with velocity streamlines obtained with 4D flow MRI in an experiment without a device, with FD 0, FD 1, and FD 2 (a); net flow rate calculated in the branches outlet 2-4 (b), forward intra-aneurysmal flow (b). All FDs resulted in a reduction of aneurysmal flow rate.

MIP of the coronary right ICA (A) of TOF-MRA showing an obvious flow dephasing artifact within the vessel lumen.

Corresponding images of PETRA-MRA are shown as (B)with no sign of any dephasing artifact.