Quick Fire Session
SCMR 22nd Annual Scientific Sessions
Elizabeth Jenista, PhD
Research Scholar
Duke University
David Wendell, PhD
Senior Research Associate
Duke University Medical Center
Han Kim, MD
Associate Professor of Medicine
Duke University
Clerio Azevedo, MD, PhD
Medical Instructor
CDPI-DASA
Wolfgang Rehwald, MSc
Staff Scientist
Siemens Healthineers
Enn-Ling Chen, PhD
Research Professor
Duke Cardiovascular Magnetic Resonance Center
Stephen Darty, RT, FSCMR
Cardiac MRI Technologist
Duke University
Logan Smith
Undergraduate Research Associate
Duke University
Michele Parker, MSc
Statistician / Business Manager
Duke Cardiovascular MR Center
Raymond Kim, MD
Professor of Medicine and Radiology
Duke University Medical Center
Background:
Delayed enhancement MRI (DE-MRI) has become the reference standard for in-vivo imaging of myocardial infarction (MI). A limitation of DE-MRI is that infarcted myocardium and the blood pool often hyperenhance to a similar degree following contrast administration. Fortunately, dark-blood DE-MRI techniques (e.g. flow-independent dark-blood delayed -enhancement (FIDDLE)) allow visualization of tissue contrast-enhancement while simultaneously suppressing blood pool signal (1-4). While both magnetization transfer (MT) and T2-preparations have previously been described for dark-blood DE-MRI at 1.5T, only MT-preparation has been used at 3T. In this study, we evaluated the performance of different preparation modules for FIDDLE at 3T.
Methods:
Variants of FIDDLE were developed using MT and T2-preparation (HS (5) and BIREF1(6)) and tested in 24 patients at 3T (figure 1). Images were acquired with each variant of FIDDLE in an interleaved fashion 10 minutes after gadolinium administration with otherwise identical acquisition parameters. Images were visually and quantitatively assessed for artifacts and for differences in right ventricle to left ventricle (RV-to-LV) blood pool suppression. T1-maps (MOLLI) and T2-maps were acquired to investigate differences in RV-to-LV blood pool suppression.
Results:
Bright artifacts, reflecting incomplete blood pool suppression, were frequently observed in the left atrium with T2-prep FIDDLE (89% for HS and 78% for BIREF1)(table 1, figure 2) and often the artifact extended into the LV cavity (28% for HS and 6% for BIREF1). MT-prep FIDDLE resulted in significantly fewer patients with in-flow artifacts (22%), which never extended into the LV cavity. Additionally, visibly different RV-to-LV blood pool suppression was observed with T2-prep in up to 90% of patients. In these patients, the RV blood pool signal was elevated, reducing the conspicuity of the border between the RV blood pool and myocardium. In contrast, there were no visible differences in RV-to-LV blood pool suppression with MT-prep. Quantitative assessment of the artifacts and differences in RV-to-LV blood pool suppression was consistent with visual analysis (table 1). We investigated the mechanism behind this effect by comparing the observed T1 and T2 in the RV and LV blood pools to the FIDDLE image intensity. We found that differences in the measured T2 and T1 of the RV and LV blood-pool accounted for 91% and 9%, respectively, of the differences in RV-to-LV blood-pool suppression.
Conclusion:
Fewer bright blood pool artifacts and more consistent blood pool suppression throughout the heart was observed with MT-prep FIDDLE compared to T2-prep FIDDLE, suggesting that MT-prep is the superior choice for dark-blood DE-MRI at 3T.