Effects of Hypertrophic and Dilated Cardiac Geometric Remodeling on Ejection Fraction

FIGURE 1. Visualization of the myocardial numerical model at three layers, the endocardium (red), the mid-wall layer (green), and the epicardium (blue), for the (A,B) normal heart, and during (C,D) myocardial wall thickening and (E,F) chamber dilation.

FIGURE 2. (A–D) Anatomic and functional characteristics of the porcine animal models, (A) Endocardial ejection fraction (B) LVM normalized by body weight, (C) relative wall thickness, and (D) spherical index. (E–H) Myocardial strains exhibited by the animal subjects. (E,G) Global longitudinal strain and (F,H) global circumferential strain at (E,F) the mid-wall and (G,H) the endocardium, measured from the healthy (sham), LVH and CAD porcine CMRI. *p < 0.05 with baseline. ∼ p < 0.05 with healthy for LVH only.

FIGURE 3. (A,B) The dependency of endocardial EF and mid-wall EF on (A) myocardial thickness ratio and (B) chamber dilation ratio, when strains at the mid-wall layer were fixed at either −0.12 or −0.16. (C–F) The complete dependency of (C,D) endocardial EF and (E,F) mid-wall EF on thickness ratio (four ratios from purple to yellow are 0.84, 1.00, 1.16 and 1.40), dilation ratio (four ratios from purple to yellow are 0.8, 1.0, 1.2 and 1.4), and longitudinal and circumferential strains at the mid-wall layer.

FIGURE 4. The dependency of SV on (A) myocardial thickness ratio and (B) chamber dilation ratio, when strains at the mid-wall layer were fixed at either −0.12 or −0.16. Results for other myocardial strains are given in Supplementary Figure S2.

FIGURE 5. Longitudinal strain (LS) and circumferential strain (CS) at the endocardium and epicardium at various (A,C) wall thickness ratios and (B,D) dilation ratios, where the mid-wall ED-to-ES strains were kept constant at −0.12. In (A) and (B), myocardium was appropriately modelled as incompressible, and LS and CS resulted in non-zero radial strains. In (C) and (D), radial strains were assumed to be zero, to show that without radial displacements, there would be minimal epi-to-endo spatial strain variability. (E,F) the radial displacements of the epicardial and endocardial surfaces from the mid-wall when the heart has (E) wall thickening or (F) dilation.

FIGURE 6. Schematic to explain the effects of radial displacement on endocardial strains. During contractions, the endocardial boundary not only shortens circumferentially (and longitudinally) but it is also pushed inwards by radial displacement, which shortens the endocardial boundary further. This additional shortening effect is enhanced when the heart is hypertrophic and the wall is thicker, since thicker wall will have higher radial displacement if the radial strain is not changed.

FIGURE 7. After removing radial strain from our numerical model, endocardial EF became largely insensitive to geometric changes, in contrast with results in Figure 3, demonstrating that radial displacement was the mechanism for endocardial EF shifting with geometry changes. (A,B) The dependency of endocardial EF on (A) myocardial thickness ratio and (B) chamber dilation ratio, when longitudinal and circumferential strains at the mid-wall layer were fixed at either −0.12 or −0.16, and when radial strain was 0. (C–D) The complete dependency of endocardial EF on thickness ratio, dilation ratio, and longitudinal and circumferential strains at the mid-wall layer.

FIGURE 8. (A) Endocardial EF and (B) mid-wall EF of LVH model and CAD model at termination, compared to their appropriate controls. The control for LVH was sham at termination, while that for CAD group was its own data at baseline, Significance: *p < 0.05 comparing to the control group.

FIGURE 9. Correlation analysis between (A,B) endocardial EF or (C,D) mid-wall EF and (A,C) normalized LV mass and (B,D) mid-wall circumferential fractional shortening.