The effect of respiratory motion compensation in intracardiac 4D flow magnetic resonance imaging on left ventricular flow dynamics, multicomponent particle tracing, and valve tracking

Figure 1. CMR scans with segmentation and valve tracking, as used to perform four-component intracardiac particle tracing. (A–E) LV cine segmented fast gradient echo scans of short-axis, two-chamber left, four-chamber, sagittal aortic annulus, and coronal aortic annulus, respectively. Straight lines depict location and orientation of scans that are perpendicular to the image plane. Blue lines and arrows depict mitral annulus (B and C) or aortic annulus (D and E) location and flow direction. (F) Phase image of 4D flow MRI with velocity encoding in feet–head direction with time point chosen during peak-systole.

Figure 2. Intracardiac particle tracing flow components in 4D flow MRI with respiratory compensation (A) and non-compensated (B) in 15 healthy volunteers. Flow fractions and EF were compared with paired sample t-test (C), but no significant P-values were found. DF, direct flow; DEF, delayed ejection flow; RI, retained inflow; RV, residual volume; SV, stroke volume; Mismatch, |mitral flow–SV|.

Figure 3. Example of visual results of four-component intracardiac particle tracing. Blue segmentation model of end-diastolic endocardial contours is drawn. Particles and pathlines are colour coded by flow component. (A) Pathlines from release to early systole, juxtaposed against 4D flow MRI slice (phase contract with feet–head flow encoding). (B) Particles at end of systole after forward particle tracing. Particles are black if excluded in either forward or backward particle tracing. Green, direct flow; blue, delayed ejection flow; yellow, retained inflow; red, residual volume; black, excluded particles.

Figure 4. Intracardiac particle tracing flow components in 4D flow MRI with respiratory compensation (A) and non-compensated (B) in 10 healthy volunteers after exclusion of inaccurate data. Inaccurate particle tracing analyses were excluded from the data set. Flow fractions and EF were compared with paired t-test (C), but no significant P-values were found. DF, direct flow; DEF, delayed ejection flow; RI, retained inflow; RV, residual volume; SV, stroke volume; Mismatch, |mitral flow–SV|.

Figure 5. LV total KE (A) and total vorticity (B) at peak-systole (Sys), E-peak (E), and A-peak (A), indexed to end-diastolic volume. RC, respiratory motion-compensated MRI; NC, non-compensated MRI.