T1, T2, and Fat Fraction Cardiac MR Fingerprinting: Preliminary Clinical Evaluation

Abstract

Background Dixon cardiac magnetic resonance fingerprinting (MRF) has been recently introduced to simultaneously provide water T1, water T2, and fat fraction (FF) maps. Purpose To assess Dixon cardiac MRF repeatability in healthy subjects and its clinical feasibility in a cohort of patients with cardiovascular disease. Population T1MES phantom, water‐fat phantom, 11 healthy subjects and 19 patients with suspected cardiovascular disease. Study Type Prospective. Field Strength/Sequence 1.5T, inversion recovery spin echo (IRSE), multiecho spin echo (MESE), modified Look–Locker inversion recovery (MOLLI), T2 gradient spin echo (T2‐GRASE), 6‐echo gradient rewound echo (GRE), and Dixon cardiac MRF. Assessment Dixon cardiac MRF precision was assessed through repeated scans against conventional MOLLI, T2‐GRASE, and PDFF in phantom and 11 healthy subjects. Dixon cardiac MRF native T1, T2, FF, postcontrast T1 and synthetic extracellular volume (ECV) maps were assessed in 19 patients in comparison to conventional sequences. Measurements in patients were performed in the septum and in late gadolinium enhanced (LGE) areas and assessed using mean value distributions, correlation, and Bland–Altman plots. Image quality and diagnostic confidence were assessed by three experts using 5‐point scoring scales. Statistical Tests Paired Wilcoxon rank signed test and paired t‐tests were applied. Statistical significance was indicated by *(P < 0.05). Results Dixon cardiac MRF showed good overall precision in phantom and in vivo. Septal average repeatability was ~23 msec for T1, ~2.2 msec for T2, and ~1% for FF. Biases in healthy subjects/patients were measured at +37 msec*/+60 msec* and –8.8 msec*/–8 msec* when compared to MOLLI and T2‐GRASE, respectively. No statistically significant differences in postcontrast T1 (P = 0.17) and synthetic ECV (P = 0.19) measurements were observed in patients. Data Conclusion Dixon cardiac MRF attained good overall precision in phantom and healthy subjects, while providing coregistered T1, T2, and fat fraction maps in a single breath‐hold scan with similar or better image quality than conventional methods in patients. Level of Evidence 2. Technical Efficacy Stage 2.

Myocardial T1, T2, T2*, and fat fraction quantification via low‐rank motion‐corrected cardiac MR fingerprinting

“…26,57 Moreover, this framework is expected to also enable post-contrast ECV mapping (thus producing all SCMR recommended parameters for myocardial tissue characterization with a single sequence), as shown in our previous work. 69 F I G U R E 7 T1, T2, T2*, and FF AHA segmental analysis for the mean values measured with different regions of the left ventricle, obtained with NMC-MRF, the proposed LRMC cardiac MRF (LRMC-MRF) approach, and the corresponding conventional methods: MOLLI, T2-GraSE, eight-echo GRE, and six-echo GRE. The proposed LRMC-MRF estimates higher T1 values than MOLLI, lower T2 and T2* values than T2-GraSE, and eight-echo GRE (respectively), and slightly higher FF values than six-echo GRE.…”

Simultaneous T1, T2, and T1ρ cardiac magnetic resonance fingerprinting for contrast agent–free myocardial tissue characterization

“…However, a regularized reconstruction like multicontrast patch-based higher-order reconstruction is not expected to obscure tissue characterization when its regularization parameters are optimized, as shown by previous studies. 3,45,46 Nevertheless, a further optimized sequence design for this T 1 /T 2 /T 1⍴ MRF sequence could lead to more efficient parametric encoding/scan efficiency, thus allowing higher resolution, and/or a shorter cardiac windows could be investigated to decrease any potential remaining in-plane cardiac motion. Quantification of relaxation times in myocardial tissue shows slightly F I G U R E 4 (A) Violin plots of T 1 , T 2 , and T 1ρ cardiac MRF quantification compared with their reference values obtained from conventional T 1 -MOLLI, T 2 -GRASE, and T 1⍴ -TFE.…”

“…Magnetic resonance fingerprinting could address some of these problems, as it provides inherently co‐registered multiparametric maps from a single scan. Cardiac MRF has been shown to provide simultaneous T 1 and T 2 mapping in a single breath‐hold scan of about 16 seconds 1,19 and more recently has been extended to also provide fat fraction estimation from a single scan of similar duration 3,20 . Simultaneous T 1 , T 2 , and T 1ρ MRF has been recently proposed for liver, 21 knee, 22 and lower leg muscle 23 imaging; however, it has not been demonstrated for cardiac MRF, which requires significantly higher undersampling factors due to electrocardiogram (ECG) gating and breath‐holding, limiting the scan time.…”

Simultaneous comprehensive liver T1, T2, , T1ρ, and fat fraction characterization with MR fingerprinting

“…The MRF underestimation bias for T 2 has also been previously reported in several studies. 18,20,34,39,54,55 Diffusion effects 56 or imperfect T 2 relaxation during the adiabatic T 2 preparation pulses could be some of the sources of this underestimation. A clinical reference technique for T 1⍴ quantification does not currently exist in abdominal imaging.…”