Jana Huiyue Zhang1, Aurélien Bustin2,3, Augustin C. Ogier1, Ruud B. van Heeswijk1, Christopher W. Roy1, Isabel Montón Quesada1, Matthias Stuber1,4, and Jérôme Yerly1,4
1Department of Diagnostic and Interventional Radiology, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland, 2IHU LIRYC, Electrophysiology and Heart Modeling Institute, INSERM U1045, Centre de recherche Cardio-Thoracique de Bordeaux, Université de Bordeaux, Bordeaux, France, 3Department of Cardiovascular Imaging, Hôpital Cardiologique du Haut-Lévêque, CHU de Bordeaux, Bordeaux, France, 4Center for Biomedical Imaging (CIBM), Lausanne, Switzerland
Synopsis
Keywords: Myocardium, Cardiovascular
Motivation: Most cardiac T1ρ mapping techniques collect data in 2D, while the few existing 3D methods use a Cartesian sampling trajectory.
Goal(s): In this study, we report the first use of a 3D radial trajectory for whole-heart T1ρ mapping at 1.5 T.
Approach: The proposed technique was validated in a phantom, tested in one volunteer, and compared to a 2D reference method.
Results: Our results show that the T1ρ values obtained from our 3D technique are comparable to the 2D reference values for both phantom and volunteer. However, we observed an underestimation of the T1ρ values using the 3D method in the phantom.
Impact: 3D radial free-breathing electrocardiogram-triggered
cardiac T1ρ mapping at 1.5 T has the potential to serve
as a contrast agent-free adjunct to late gadolinium enhancement imaging for the
whole-heart motion-robust quantification of tissue properties such as
myocardial fibrosis.
Introduction
T1-rho (T1ρ) mapping of the heart is a promising contrast
agent-free method for the quantitative characterization of diseased myocardial
tissue properties such as myocardial fibrosis.1 While most T1ρ mapping techniques are 2D and, therefore, not well-suited for whole-heart
assessment, a few 3D methods2,3 have been proposed. However, the existing 3D techniques use
Cartesian sampling, which is known to be more susceptible to motion artifacts
than radial sampling.4 To overcome these
limitations, we propose a free-breathing 3D whole-heart myocardial T1ρ mapping sequence at 1.5 T that leverages a motion-robust radial
trajectory. Methods
The proposed method (Fig. 1) uses a free-breathing ECG-triggered
balanced steady-state free precession (bSSFP) pulse sequence with a 3D spiral
phyllotaxis radial trajectory5 (TE/TR=1.29/2.58ms, RF excitation
angle=70°, 32 k-space lines per segment, 10 ramp up pulses before each segment,
1076 segments, field of view=(250mm)3, resolution=(1.5mm)3,
2 heartbeats/trigger).
For magnetization preparation, T1ρ preparation
modules with 5 different spin-lock times (SLT=0/10/20/35/50ms) were applied in
an interleaved fashion. The T1ρ preparation module consists of the RF
excitation sequence 90x – SLy – 180y – SL-y
– SLy – 180-y – SL-y – 90-x with a
spin-lock frequency of 500 Hz.
We reconstructed respiratory motion-corrected images with the focused
navigation (fNAV)6 technique, which derives 3D
rigid motion estimations using the respiratory signal extracted from the
patient table. Next, the images were denoised using the HD-PROST
patch-based algorithm.7 Finally, the 3D T1ρ map was calculated by fitting the
signal intensities of the different T1ρ weighted
images to the function S(SLT)=Soe-(SLT/T1ρ), where S is the signal intensity,
SLT the spin-lock time, and S0 the equilibrium magnetization.
Data were acquired on a 1.5 T clinical scanner
(Magnetom Sola, Siemens Healthcare, Germany). We validated our proposed method in
both an agar-NiCl2-gel phantom, as well as in one healthy
volunteer (m/29y). Bland-Altman plots and regression analyses were used to compare
the T1ρ
values derived from our proposed 3D method with those obtained from a 1-slice T1ρ map
in short-axis view that was acquired with the previously reported 2D breath-hold
single-shot ECG-triggered (3 heartbeats/trigger) bSSFP Cartesian T1ρ
mapping sequence.8 For
the phantom, we only evaluated the T1ρ values in the vials that contain values
relevant for cardiac imaging. For the volunteer, we measured and compared T1ρ values
in regions-of-interest drawn within the myocardium and the papillary muscle. Acquisition
times were recorded.Results
We successfully acquired T1ρ-weighted
images with our proposed method and calculated the corresponding 3D T1ρ maps in the phantom (Fig. 2). The T1ρ values for
the vials in the myocardial range showed good agreement between our 3D
technique and the 2D reference method, with a slope of 0.824 (P<0.0001) for
the linear regression (Fig. 2B). However, the Bland-Altman plot suggested a significant
bias of 11.74 ms, indicating a systematic underestimation of T1ρ values with our method (Fig. 2C). For the volunteer, the acquisition of
the 3D whole-heart T1ρ maps took ~35 min. We successfully
reconstructed high-resolution, isotropic, whole-heart T1ρ maps, which enable reformatting in any orientation (T1ρ map in coronal, sagittal, and axial orientation in Fig. 3). The
quantitative evaluation of the T1ρ maps in
short-axis view (Fig. 4A) showed an overestimation of T1ρ values in the myocardium and the papillary muscle by our proposed 3D method in
comparison to the 2D reference (Fig. 4B). Discussion
This preliminary study reports the first application
of a 3D radial trajectory for T1ρ mapping of the heart. Our proposed 3D radial technique was assessed in
both a phantom and a volunteer. Phantom results showed good correlation to the
2D reference method. However, a slight underestimation of T1ρ values was observed, which might be due to insufficient
magnetization recovery, as we used only 2 heartbeats per trigger in comparison
to the 2D reference that allows for 3 heartbeats between every T1ρ preparation module. In future studies, this can
be accounted for by using dictionary matching instead of exponential fitting
for the T1ρ mapping.
Our method was successfully applied in vivo in
a healthy volunteer. Unlike the phantom results, the T1ρ map obtained in vivo showed a slight
overestimation in T1ρ values for
both the myocardium and papillary muscle. From our results, the origin of this
overestimation is unclear, but it could be attributed to a noise bias in the
fit. In future studies, we plan to explore the potential of compressed sensing
and low-rank reconstruction as solution to these problems, while also seeking
to reduce overall scan time. Moreover, we aim to collect additional data from volunteers
to rigorously quantify any potential bias between our 3D method and the 2D
reference.Acknowledgements
This work was supported by funding from the French
National Research Agency under grant agreement ANR-22-CPJ2-0009-01 and from the
European Research Council (ERC) under the European Union’s Horizon 2020
research and innovation program (grant agreement N°101076351).References
1. Bustin A, Witschey WRT, van Heeswijk
RB, Cochet H, Stuber M. Magnetic resonance myocardial T1ρ mapping. J
Cardiovasc Magn Reson. 2023;25(1):34. doi:10.1186/s12968-023-00940-1
2. Qi H, Bustin A, Kuestner T, et al.
Respiratory motion-compensated high-resolution 3D whole-heart T1ρ mapping. J Cardiovasc Magn Reson. 2020;22(1):12.
doi:10.1186/s12968-020-0597-5
3. Qi H, Lv Z, Hu J, et al. Accelerated
3D free-breathing high-resolution myocardial T1ρ mapping at 3
Tesla. Magn Reson Med. 2022;88(6):2520-2531. doi:10.1002/mrm.29417
4. Xie J, Lai P, Huang F, Li Y, Li D.
Cardiac magnetic resonance imaging using radial k-space sampling and
self-calibrated partial parallel reconstruction. Magn Reson Imaging.
2010;28(4):495-506. doi:10.1016/j.mri.2009.12.018
5. Piccini D, Littmann A,
Nielles-Vallespin S, Zenge MO. Spiral phyllotaxis: the natural way to construct
a 3D radial trajectory in MRI. Magn Reson Med. 2011;66(4):1049-1056.
doi:10.1002/mrm.22898
6. Roy CW, Heerfordt J, Piccini D, et
al. Motion compensated whole-heart coronary cardiovascular magnetic resonance
angiography using focused navigation (fNAV). J Cardiovasc Magn Reson.
2021;23(1):33. doi:10.1186/s12968-021-00717-4
7. Bustin A, Lima da Cruz G, Jaubert O,
Lopez K, Botnar RM, Prieto C. High-dimensionality undersampled patch-based
reconstruction (HD-PROST) for accelerated multi-contrast MRI. Magn Reson Med.
2019;81(6):3705-3719. doi:10.1002/mrm.27694
8. Bustin A, Toupin S, Sridi S, et al.
Endogenous assessment of myocardial injury with single-shot model-based
non-rigid motion-corrected T1 rho mapping. J Cardiovasc Magn Reson. 2021;23(1):119.
doi:10.1186/s12968-021-00781-w