High-resolution 3D whole-heart bright- and black-blood imaging with co-registered T2 mapping at 0.55 T

Figures

• 

  Figure 1. (A) Sequence diagram over three interleaved heartbeats (HBs): T2prep (50 ms) IR (TI = 90 ms)—none—T2prep (30 ms). (B) Respiratory binning of data acquired with VD-CASPR to derive 3D non-rigid motion fields. Images are reconstructed in-line using motion-corrected itSENSE followed by off-line HD-PROST for denoising. (C) Contrasts are shown for the mid-coronal slice. 3D datasets for each volume can be found in the Supplementary Material. The bright-blood image is acquired in the 1nd HB, while the black-blood image is obtained through direct magnitude subtraction of the 1st from the 2nd volume. T2 maps are generated voxel-wise by dictionary matching.

• 

  Figure 2. Optimization on sequence parameters using EPG simulations at 0.55 T for a HR of 60 bpm. Simulated signals of the arterial blood (T1=1100ms,T2=260ms), healthy myocardium (T1=700ms,T2=58ms) and epicardial fat (T1=180ms,T2=93ms) are shown for the 1st volume (A) and 2nd–1st subtracted volume (B). Contrast values for the bright- and black-blood are defined as a difference between the corresponding signals. In each graph one of the following parameters was varied, while the other two were kept constant: imaging flip angle (FA), inversion time (TI) and T2prep pulse duration. Signal dependency on (i) FA with fixed inversion TI of 90 ms and T2prep of 50 ms, on (ii) TI with fixed FA of 110∘ and T2prep of 50 ms, and on (iii) T2prep pulse duration with fixed FA of 110∘ and TI of 90 ms.

• 

  Figure 3. Validation of the proposed 3D T2 multi-contrast mapping sequence in the standardized T1MES phantom with simulated HR of 60 bpm at 0.55 T against a 2D SE reference scan with correlation plot (A) and Bland-Altman analysis (B). The relevant T2 myocardial region is zoomed in. (C) Phantom image of the proposed and reference sequences with the parametric values given in Table 1. The T2 value for H vial was out of range and is therefore not shown.

• 

  Figure 4. High-resolution bright-blood images are illustrated for four subjects in multiplanar reformation. The course of the left (LCA) and right (RCA) coronary arteries is indicated by white arrows.

• 

  Figure 5. 3D reconstructed data for two healthy subjects, showing bright- and black-blood images with co-registered T2 maps in coronal, vertical long-axis (for evaluating the anterior/inferior walls and apex of the LV) and horizontal long-axis (for assessment of LV septal/lateral walls and apex, as well as RV free wall and chamber sizes) orientations. On the black-blood images different intrapericardial structures are indicated as: LV/RV—left/right ventricle, LA/RA—left/right atrium, Ao—ascending aorta, PA—pulmonary artery. To highlight relevant cardiovascular structures, a binary contrast-generated mask was applied to the T2 map dataset.

• 

  Figure 6. Bright- and black-blood images with co-registered T2 maps of SAx cross-section slices from apex to base are displayed for a healthy subject with HR of 67 bpm. The mean T2 values, SD and CoV in each of 16 segments are shown in the bull’s-eye-plot generated by projecting LV myocardial tissue onto a plane.

• 

  Figure 7. (A) Anatomical images and mapping shown for four representative subjects in mid-cavity SAx view. (B) Violin plot comparing the mean T2 values from apical, mid-cavity and basal regions. (C) Comparison of T2 mean and SD values across all subjects.