Introduction

Cardiac Magnetic Resonance Fingerprinting¹ (MRF) has shown promising results for myocardial fibrosis and inflammation characterization, enabling simultaneous T₁ and T₂ parameter mapping as well as extracellular volume (ECV) mapping, when performed pre- and post-contrast injection^2,3. In addition, T_1ρ mapping has shown promising results for detection of focal and diffuse myocardial fibrosis without the need of exogenous contrast agents⁴. However, multiparametric T₁, T₂ and T_1ρ mapping requires sequential acquisitions under several breath-holds, that can lead to non-registered maps, bias due to inter-parameter dependencies, as well as potential patient discomfort. Simultaneous T₁, T₂ and T_1ρ MRF has been recently proposed for liver imaging⁵; however, cardiac MRF requires significantly higher undersampling factors due to ECG gating and breath-holding limiting the scan time. In this work, we propose a novel simultaneous T₁, T₂ and T_1ρ cardiac MRF approach to enable comprehensive contrast-free myocardial tissue characterization in a single breath-hold scan.

Methods

An ECG-triggered 2D cardiac MRF gradient echo sequence for simultaneous quantification of T₁, T₂ and T_1ρ is proposed (Fig.1). Inversion recovery (IR), T₂ preparation (T₂p, echo times T₂pTEs = {0, 40, 80 ms}) and T_1ρ spin-lock preparation pulses (spin lock times SLTs = {0, 40, 80 ms} at 350 Hz) are applied at different heartbeats to encode T₁, T₂ and T_1ρ. T_1ρ preparation pulses robust to both B₁ and B₀ inhomogeneities were considered⁶. The acquisition consists of 16 heartbeats, performed under breath-hold. Highly undersampled MRF images are reconstructed using a multi-contrast patch-based higher-order low-rank reconstruction⁷ and matched to a subject-specific EPG-based⁸ dictionary (considering ~140000 different T₁/T₂/T_1⍴ combinations) to obtain the T₁, T₂ and T_1ρ maps.
Acquisitions were performed on phantoms and 5 healthy subjects on a 1.5T scanner (Ingenia, Philips Healthcare). Acquisition parameters included: 30 spiral readouts per heartbeat, sinusoidally varying FA of 5°-10°, TE/TR = 0.91/7.11 ms, scan time ~16s. Phantom scans were performed on a set of sample tubes filled with water and different solutions of varying concentrations of agar (for T₂ and T_1⍴ variation) and a Gd-based contrast agent (for T₁ variation). Conventional, sequential T₁-MOLLI, T₂-GRaSE and T_1⍴-reference scans were performed with clinically used parameters for comparison purposes. The T_1⍴-reference scan consisted of 5 consecutive single shot TFE acquisitions (SLTs = {0, 10, 25, 50, 60 ms}, at 350 Hz) with a 4.5s gap between SLT acquisitions to allow for magnetization recovery.

Results

Representative phantom T₁, T₂ and T_1⍴ maps are shown in Figure 2 in comparison to the corresponding reference maps. Values and correlation from T₁, T₂ and T_1⍴ cardiac MRF quantification against their reference value are shown in Figure 3. Good linear fit, close to the identity line (slopes: m = 1.10, m = 0.85 and m = 0.99 for T₁, T₂ and T_1⍴ respectively, with their y-intercepts not significantly different from 0) and excellent correlation (r² = 0.99, r² = 0.99 and r² = 0.95 for T₁, T₂ and T_1⍴ respectively) was achieved for all parameters. Representative T₁, T₂ and T_1⍴ maps for two healthy subjects are shown in Figure 4 in comparison to clinical conventional T₁-MOLLI, T₂-GRaSE and T_1⍴-reference maps. ROIs were manually drawn inside the myocardial septum. Quantification and comparison against reference values are shown in Figure 5 for all healthy subjects. A slight overestimation in T₁ and underestimation in T₂ and T_1⍴ can be found, which is consistent with previous findings^1,5. Bland-Altman plots of T₁, T₂ and T_1⍴ MRF values against their references (Figure 5b) show that deviation from the reference does not increase at higher values.

Conclusion

We propose a cardiac MRF acquisition scheme for simultaneous quantification of myocardial T₁, T₂ and T_1ρ in a contrast-free single breath-hold MR scan. The proposed approach has been investigated in phantoms and healthy subjects. Further studies are warranted to validate the technique in patients with cardiovascular disease.

Acknowledgements

This work was supported by EPSRC (EP/P001009, EP/P032311/1, EP/P007619/1) and Wellcome EPSRC Centre for Medical Engineering (NS/ A000049/1).