Focus Session
SCMR 22nd Annual Scientific Sessions
Yuchi Liu, PhD
Research Associate
Case Western Reserve University
Jesse Hamilton, PhD
Research Associate
Case Western Reserve University
Mark Griswold, PhD
Professor
Case Western Reserve University; University Hospitals
Nicole Seiberlich, PhD
Associate Professor
Case Western Reserve University
Background:
Cardiac Magnetic Resonance Fingerprinting (cMRF) has recently been introduced for simultaneous T1 and T2 quantification in the myocardium during a single scan1. One important feature of the MRF framework is the potential to measure multiple tissue properties beyond T1 and T2. Fat-water separation is valuable in various clinical applications such as evaluation of myocardial lipomatous infiltration2,3. Here we propose an approach for simultaneous fat mapping and T1 T2 quantification based on cMRF framework1 with a rosette trajectory4,5. The rosette trajectory allows for data reconstruction at either the water or fat resonance frequency due to its spectral selectivity4. During post-processing, the k-space data acquired with water on-resonance can be shifted to fat on-resonance simply by multiplying a phase term that compensates the off-resonance accumulation of fat during the readout period4. Thus, separate water images and fat images can be reconstructed from one dataset. Using this approach combined with cMRF, T1, T2, water proton density (PD) and fat PD maps were generated from one scan.
Methods: Two healthy volunteers were scanned after written informed consent in this IRB-approved study on a 3T scanner (Siemens Skyra) at University Hospitals Cleveland Medical Center. A Class I rosette trajectory with five zero-crossings (Fig1.a) and a Class II rosette trajectory with four zero-crossings (Fig1.b)6 were used to collect data with water on-resonance at end-diastole in a 15-heartbeat breath-hold with ECG-triggering. Acquisition parameters for these two types of trajectories are summarized in Table 1. Magnetization preparation pulses and flip angle pattern were employed as in this reference7 (Fig1.c). The chemical shift of fat was obtained from the shimming procedure before the acquisition started. A dictionary including slice profile and preparation pulses efficiency corrections8 and low rank reconstruction9 were used. For water image series, the dictionary covers a range of T1/T2 values from 10/2 to 3000/500ms for water and 300/40 to 500/100ms for fat (limiting the dictionary for fat T1 and T2 at 3T)10,11. T1, T2, and PD maps of water and PD map of fat were produced from the water image and fat image series, respectively.
Results:
Figure 2 shows the T1, T2, water PD, and fat PD maps at mid-ventricular level in two healthy subjects acquired by sequence A and B, respectively. Myocardial T1 and T2 values are 1353.4 ± 49.6 and 28.1 ± 2.7ms in subject A; 1319.6 ± 70.7 and 30.2 ± 3.8ms in subject B. Epicardial fat was shown clearly in the fat PD map in both cases.
Conclusion:
This study shows that fat mapping can be achieved along with T1 and T2 quantification in the heart in a single scan of less than 15s using rosette cMRF. This approach could provide multiple biomarkers (i.e. fat deposition, T1, T2) for myocardial diseases simultaneously and potentially be used to quantify proton density fat fraction.