Quick Fire Session
SCMR 22nd Annual Scientific Sessions
Luuk Hopman, MSc
PhD Student
Research Institute of the McGill University Health Center
Elizabeth Hillier, BSc
PhD Candidate
Research Institute of the McGill University Health Center
Yuchi Liu, PhD
Research Associate
Case Western Reserve University
Jesse Hamilton, PhD
Research Associate
Case Western Reserve University
Nicole Seiberlich, PhD
Associate Professor
Case Western Reserve University
.png "Matthias Friedrich, MD photo")
Matthias Friedrich, MD
Professor
McGill University Health Centre
Background:
Cardiac Magnetic Resonance Fingerprinting (cMRF) is a novel technique that enables the rapid and simultaneous mapping of myocardial T1 and T2. Recent advances in the cMRF technique allow for acquisition times as short as 5 seconds. This opens the possibility to acquire T1 and T2 maps in dynamic settings. Standardized breathing maneuvers have been successfully applied to induce a predictable response of the coronary circulation. We tested the feasibility of sequential rapid cMRF acquisitions (dynamic cMRF) during standardized breathing maneuvers.
Methods:
T1 and T2 values generated with cMRF were first assessed in a (ISMRM/NIST) phantom. In five healthy volunteers, we compared T1 (MOLLI) and T2 (T2-prepared balanced SSFP) results to standard (acquisition time 15 heartbeats) and ultrafast (acquisition time 5 heartbeats) cMRF at 3T. Ultrafast cMRF was also used to assess T1 and T2 signal evolution over time during breathing maneuvers (60s hyperventilation followed by a long breath hold). All cMRF data were reconstructed using a low rank method with a dictionary including slice profile and preparation pulses efficiency corrections.
Results:
In the range of myocardial T1 and T2 values in humans, ISMRM/NIST phantom data showed a strong linear correlation between both T1 and T2 maps collected with conventional methods, 15-heartbeat cMRF T1 maps, and 5-heartbeat cMRF T1 maps (R2=0.99 and R2=0.99, respectively), see Fig. 1. In healthy volunteers, the mean MOLLI T1 value was 1223±79 ms, the mean T1/cMRF (15 HB) was 1362±92 ms, and the mean T1/cMRF (5 HB) was 1396±133 ms. For T2, the mean value measured with the conventional mapping technique was 41.0±5.7 ms, the mean T2/cMRF (15 HB) was 29.±6.1 ms, and the mean T2/cMRF (5 HB) was 30.1±7.5 ms (Fig. 2). In the dynamic cMRF scans, the myocardial T1 changed from 1400±83 ms at baseline to 1377±403 ms at 60 seconds (p=0.42) and mean T2 changed from 27.9±5.6 ms at timepoint 0 seconds to 32.3±8.5 ms at 60 seconds (p=0.02)(Fig. 3).
Conclusion:
Our results indicate that in the ISMRM/NIST phantom, standard (15 HB) and ultrafast (5 HB) cMRF show excellent agreement with conventional T1 and T2 values. In healthy volunteers, myocardial T1 values derived from cMRF were higher than those obtained in conventional T1 mapping methods, while T2 was lower. These differences may be due to confounders such as magnetization transfer, flow artifacts, and partial volume effects, and are expected based on known limitations of conventional cardiac T1 and T2 mapping techniques. During a post-hyperventilation breath-hold, known to induce coronary vasodilatation, there was no significant change of myocardial T1. However, there was an observed trend of an increase in myocardial T2 using ultrafast dynamic cMRF. These results suggest that dynamic cMRF using fast sequences has the potential for assessing changes in myocardial relaxation times during dynamic interventions.