Comparison of Delayed Contrast-Enhanced Magnetic Resonance Imaging of Myocardial Viability at 1.5 and 3 Tesla

Abstract

Imaging of myocardial viability using the delayed enhancement technique currently provides high image contrast between infarcted and normal tissue with the aid of a magnetization prepared fast gradient echo pulse sequence following the administration of an extracellular contrast agent. However, there exists a degree of image contrast variability and subjectivity due to contrast agent kinetics and user-specified imaging parameters. Also, the technique has not been explored at higher field strengths (3T), which offer greater inherent signal-to-noise ratio. The overall goal of this study is to compare magnetic resonance delayed contrast enhancement of myocardial infarction at 1.5T and 3T. The analysis was conducted by first developing a comprehensive mathematical simulation of the imaging sequence, which allowed modification of various imaging parameters. Simulations were performed to optimize the sequence for flip angle and inversion time, as well as to evaluate the influence of other image parameters that affected contrast. These theoretical results were validated experimentally with phantoms. In vivo post-contrast T1 measurements at 1.5T and 3T from normal volunteers (n=10) and patients (n=5) provided more precise input into mathematical optimization simulations. In both populations, longer T1 values were found at 3T compared to 1.5T for normal (pre-contrast: 1.24 .06s vs. 1.07 .05s; post-contrast: 0.34-0.59 vs. 0.33-0.54s, n=15) and infarcted myocardium (pre-contrast: 1.27 .06s vs. 1.04 .06s; post-contrast: 0.25-0.37s vs. 0.23-0.32s, n=5). Corresponding simulations using these T1 values revealed an infarct-to-normal tissue contrast gain at 3T of approximately 25%. In vivo image contrast between infarcted and normal tissue following contrast administration was also higher at 3T by approximately 37%. In conclusion, there was good correlation between mathematical simulations of delayed enhancement and experimental results, enabling parameters to be compared and optimized offline given input T1 values. Although contrast-enhanced viability imaging at 3T suffered from artifacts due to field, RF, and inversion pulse inhomogeneity, these results suggest that 3T offers higher contrast-to-noise ratio than 1.5T for this application.