Magnetic resonance elastography as a method to estimate myocardial contractility

Abstract

Purpose: To determine whether increasing epinephrine infusion in an in vivo pig model is associated with an increase in end-systolic magnetic resonance elastography (MRE)-derived effective stiffness.

Materials and methods: Finite element modeling (FEM) was performed to determine the range of myocardial wall thicknesses that could be used for analysis. Then MRE was performed on five pigs to measure the end-systolic effective stiffness with epinephrine infusion. Epinephrine was continuously infused intravenously in each pig to increase the heart rate in increments of 20%. For each such increase end-systolic effective stiffness was measured using MRE. In each pig, Student's t-test was used to compare effective end-systolic stiffness at baseline and at initial infusion of epinephrine. Least-square linear regression was performed to determine the correlation between normalized end-systolic effective stiffness and increase in heart rate with epinephrine infusion.

Results: FEM showed that phase gradient inversion could be performed on wall thickness ≈≥1.5 cm. In pigs, effective end-systolic stiffness significantly increased from baseline to the first infusion in all pigs (P = 0.047). A linear correlation was found between normalized effective end-systolic stiffness and percent increase in heart rate by epinephrine infusion with R(2) ranging from 0.86-0.99 in four pigs. In one of the pigs the R(2) value was 0.1. A linear correlation with R(2) = 0.58 was found between normalized effective end-systolic stiffness and percent increase in heart rate when pooling data points from all pigs.

Conclusion: Noninvasive MRE-derived end-systolic effective myocardial stiffness may be a surrogate for myocardial contractility.

Figure 1

Schematic of the pig showing the MRE driver placed on the chest wall to induce external motion into the heart and an IV placed in the ear to infuse epinephrine.

Figure 2

Plot demonstres the deviation of calculated stiffness values obtained using phase gradient inversion on a spherical shell when compared to true values with varying thickness of the shell. X-axis represents the ratio of calculated wavelength to the thickness. Y-axis represents ratio of true stiffness value to that of calculated stiffness value. True stiffness value is provided as input to finnite element modeling for generating displacement field. Calculated stiffness value is obtained by processing the displacment field using phase gradient inversion algorithm. Calculated wavelength is obtained by using calculated stiffness value in the following equation. μ_cal=ρ (fλ_cal) ², where μ_cal is the stiffness obtained using phase gradien inversion, f is the excitation frequency (80 Hz), ρ density of the material 100kg/m³ and λ_cal is the calculated wavelength. From the plot it can be observed that phase gradient inversion is stable when the thickness almost equal to or greater than 1.5cm.

Figure 3

a, f) End-systolic short axis SSFP image of the LV in pig #1 at baseline and after 5^th infusion of epinephrine respectively. The green and orange contours show the delineation of LV myocardium. b, c, d, g, h, i) Wave images in x, y, and z motion sensitization directions during end-systole. e, j) Corresponding stiffness maps at end-systole with a mean stiffness of 5.95 ± 2 kPa and 14.03 ± 7.6 kPa respectively.

Figure 4

Plot of effective end-systolic stiffness versus baseline and 20% increased heart rate during initial infusion of epinephrine.

Figure 5

Plot of normalized effective end-systolic stiffness versus percent increase in heart rate in pig 1.

Figure 6

Plot of normalized effective end-systolic stiffness versus percent increase in heart rate in pig 2.

Figure 7

Plot of normalized effective end-systolic stiffness versus percent increase in heart rate in pig 3.

Figure 8

Plot of normalized effective end-systolic stiffness versus percent increase in heart rate in pig 4.

Figure 9

Plot of normalized effective end-systolic stiffness versus percent increase in heart rate in pig 5.

Figure 10

Plot of normalized effective end-systolic stiffness versus percent increase in heart rate pooled from all the pigs.