Vessel wall characterization using quantitative MRI: what’s in a number?

Figures

• 

  Fig. 1. 2D/3D blood suppression techniques. a 2D double/quadruple inversion recovery. Non-selective inversion of inflowing blood generates black-blood at a correctly timed inversion time. b 3D variable flip angle TSE. Frequency encoding gradients within the TSE readout have flow suppression properties in that direction. The point-spread function of the acquisition can be improved by variable flip angle schemes. c 3D motion-sensitized driven equilibrium (MSDE). Within a T2 preparation module, strong gradients dephase moving spins within a voxel, effectively crushing the blood signal. Magnetization of static tissue, although slightly T2-weighted, is restored by a final tip-up pulse. Subsequently, a 3D TFE/TSE readout can be applied

• 

  Fig. 2. Vessel wall imaging using 2D and 3D imaging sequences. Sequence parameters: a FOV = 144 × 144 × 35 mm, resolution = 0.7 × 0.7 × 0.7 mm. 3D MSDE, TR/TE = 10.0/3.5 ms, TEprep = 11.5 ms, TFE factor = 60, Acq time = 3 min. 3D VFA TSE, TR/TE = 1000/28 ms, ∆TE = 4.3 ms, TSE factor = 40, start-up echoes = 4, Acq time = 6 min. b 3D TSE, same as (a), 2D DIR, ST = 2 mm, resolution = 0.5 × 0.5 mm, TR/TE = 1000/3.5 ms, TI = 400 ms c 2D DIR, same as (b), but with resolution = 0.25 × 0.25 mm

• 

  Fig. 3. a Quantitative vessel wall T2 mapping. ECG-gated multi-echo spin-echo data are acquired with echo times between 25–100 ms. Blood suppression is obtained through double-inversion recovery preparation. b T2w image along with quantitative T2 map of an advanced plaque (SM smooth muscle, * lumen). Semi-automated segmentation of plaque components from the T2 map is shown on the right. Adapted from Biasiolli et al. (19) and Chai et al. [66]

• 

  Fig. 4. a Quantitative 3D vessel wall T1 mapping. The slice on the left shows regions of greatly reduced T1, indicating IPH. Adapted from Coolen et al. [31] b Masked images corresponding to panel a showing regions with T1 <400 ms, alongside registered SNAP images showing regions of IPH (SNAP+) or arterial blood (SNAP−). Regions with low quantitative T1 values match IPH-positive regions from SNAP (unpublished data)

• 

  Fig. 5. 3D quantitative diffusion vessel wall imaging a ECG-gated 3D diffusion-prepared variable-flip-angle TSE sequence. Motion-compensated diffusion gradients are combined with low b-value flow-sensitizing gradients for additional blood suppression. b Diffusion-weighted vessel wall images of a healthy volunteer at b values of 30 and 300 mm2/s, along with corresponding quantitative ADC map. c Diffusion-weighted imaging of a lipid-rich atherosclerotic plaque with high signal on the b = 300 mm2/s image and corresponding low ADC values. Adapted from Xie et al. [72]

• 

  Fig. 6. a Representative concentration versus time curve kinetics in atherosclerotic plaques. Data fitting is shown with 4 different models. Ext. extended. b Example of pixel-by-pixel parametric K trans maps of the carotid arteries, calculated using a Patlak model. Both maps were generated from the same individual and from images acquired 1 week apart, in order to evaluate inter-scan reproducibility. Adapted from Gaens et al. [79]

• 

  Fig. 7. 3D vessel wall imaging. a Multiplanar reconstruction (MPR) of 3D images acquired after the injection of Gd-DTPA into the aorta of an atherosclerotic rabbit. TFE turbo field echo. b Representative 2D turbo spin-echo (TSE) double-inversion recovery axial image acquired after injection of Gd-DTPA into the aorta of an atherosclerotic rabbit. c Corresponding AUC permeability maps (MPR) from 3D TFE DCE-MRI and near-infrared fluorescence (NIRF) after Evans blue (EB) injection in one representative atherosclerotic and control rabbit. Prominent uptake of MR contrast and EB is shown in the diseased animals compared to the control. Adapted from [86]

• 

  Fig. 8. a Peak systolic WSS vectors along the carotid (left) and thoracic aorta (right). Adapted from Potters et al. [135]. b Example of a 3D wall thickness (WT [mm]) map of the carotid bifurcation and corresponding time-averaged 3D wall shear stress WSS (Pa) map of the same subject (top). WT was significantly different in each tertile and the highest WT was found in the lowest WSS tertile (bottom). Adapted from Cibis et al. [136]

• 

  Fig. 9. Different techniques for determing pulse wave velocity (PWV) in the aorta. a foot-foot method. The path length between two slices perpendicular to the aorta is divided by the time difference between the initial upslopes of the two flow curves. b QA-method. Lumen area is plotted as function of flow during the cardiac cycle. PWV is defined as the change in lumen area (A) divided by the change in flow (Q) during the linear portion of the QA-plot. Adapted from [148]

• 

  Fig. 10. Comparison of post-contrast intracranial vessel wall imaging using 3T and 7T MRI. 7T MRI largely resulted in enhanced vessel wall delineation (due to enhanced CSF suppression) and more frequent detection of atherosclerotic lesions that were missed at 3T (rows 3 and 4). Depicted are the right proximal artery (row 1), left proximal vertebral artery (rows 2 and 3) and M1 segment of the middle cerebral artery (row 4). Adapted from Harteveld et al. [18]