Dominique Franson1, Rajiv Ramasawmy2, Nicole Seiberlich1,3, and Adrienne Campbell-Washburn2
1Case Western Reserve University, Cleveland, OH, United States, 2National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD, United States, 3University of Michigan, Ann Arbor, MI, United States
Synopsis
A free-breathing late gadolinium enhancement scan may be beneficial for
patients who have difficulty with breath-holds.
However, the fast data acquisition needed may be difficult to achieve at
low field strengths. Here, an
undersampled spiral (R=3) acquisition with a through-time spiral GRAPPA
reconstruction is used to achieve high spatial (1.25 x 1.25 mm2) and
temporal (139 ms acquisition window) resolution for LGE imaging at 0.55T. The resulting images have qualitatively good
contrast-to-noise between lesions and healthy myocardium, with acceptable SNR
and without motion blurring.
Introduction
Low field MRI systems equipped with contemporary hardware and imaging
methods can offer opportunities for cardiac imaging due to low cost, reduced SAR,
reduced susceptibility and short T1/long T2* of tissue1.
Spiral trajectories are especially amenable for low field imaging because the
prolonged T2* enables long SNR-efficient readouts, and improved field
homogeneity can reduce spiral image artifacts.
Late gadolinium enhancement (LGE) is an
important tool to assess myocardial infarctions and other tissue lesions2. Free breathing LGE offers benefits for
patients unable to hold their breath2, but depends on high acceleration rates that
may be challenging at lower fields due to coil geometry and low SNR. This study
aimed to show the feasibility of an under-sampled spiral acquisition to achieve
a high spatial and temporal resolution for LGE imaging at 0.55T.Methods
Imaging
was performed on an MRI system modified to operate at 0.55T (prototype MAGNETOM
Aera, Siemens, Erlangen, Germany)3. The proposed sequence was tested in healthy
volunteers, patients with known myocardial infarctions, and swine models of
myocardial infarction. Human imaging was approved by local Institutional Review
Board, and swine imaging was performed with local ethics approval.
Data were acquired using a balanced steady-state free precession (bSSFP)
sequence at end diastole following a non-selective adiabatic inversion pulse. Eight images were acquired with one recovery
heartbeat between each, in a total breath-hold of 16 heartbeats. The final image was reconstructed from an
average of the last 6 images, excluding the first 2 images to reach a steady T1
recovery signal. Imaging was performed
10 minutes following gadolinium administration (0.15 mL/kg gadobutrol).
We
used a 48 arm, variable density spiral-out trajectory deign4 with
zeroth and first moment gradient nulling.
Single-shot images were acquired with an under-sampling rate of 3 (16/48
arms) and reconstructed using through-time spiral GRAPPA5. Sixty fully-sampled, ungated calibration
frames were acquired before or after the accelerated data (acquisition time ~ 60
seconds). An inversion pulse was played
out every five frames to induce contrast variation in the calibration data. The scan parameters were: TR/TE = 8.66/1.04
ms, FOV = 320 x 320 mm2, matrix size = 256 x 256, in-plane
resolution = 1.25 x 1.25 mm2, slice thickness = 8 mm, flip angle =
70°, dwell time = 2.5 μs, typical inversion time = 300 ms, temporal acquisition
window = 139 ms.
Comparison images were acquired using a single-shot Cartesian phase sensitive
inversion recovery bSSFP sequence2 with the following parameters: TR/TE = 3.4/1.35 ms, FOV = 380 x 323 mm2,
in-plane resolution = 2.97 x 1.48 mm2, slice thickness = 8 mm, flip
angle = 120°,
temporal acquisition window = 320 ms. As with the
spiral images, these data were acquired during a breath-hold.
Results
Figures 1 and 2 show example images from a swine
model and from a patient with a known infarction, respectively. The images show qualitatively good contrast
between regions of late enhancement and the healthy myocardium. The myocardium nulling is reduced with the
spiral LGE sequence compared to conventional Cartesian LGE because the center
of k-space is acquired with each spiral acquisition, meaning that the inversion
time is spread over the 139 ms acquisition window. Using 6 averages, the SNR is acceptable at an
in-plane resolution of 1.25 x 1.25 mm2. The short acquisition window permits
visualization of the heart without blurring from myocardial motion. Discussion
This preliminary work demonstrates the use of an accelerated spiral
acquisition and a through-time GRAPPA reconstruction to achieve high spatiotemporal
resolution LGE images at 0.55T. Each average is acquired in a single heartbeat
rather than in a segmented manner as a precursor to a free-breathing acquisition
at low field. Toward this end, future
work will include the addition of a motion correction algorithm to allow averaging
of multiple, free-breathing averages. Quantitative
SNR and contrast-to-noise evaluation of the images will also be performed. Conclusion
bSSFP LGE images can be acquired within a reasonable acquisition window
(<200 ms) and at a diagnostic spatial resolution using an under-sampled spiral trajectory and a
through-time spiral GRAPPA reconstruction at 0.55T.Acknowledgements
Funding was provided by the National Heart, Lung, and Blood Institute’s
Division of Intramural Research. We would
like to acknowledge the assistance of Siemens Healthcare in the modification of
the MRI system for operation at 0.55T under an existing cooperative research
agreement (CRADA) between NHBLI and Siemens Healthcare. References
1. Simonetti OP, Ahmad
R. Low-Field Cardiac Magnetic Resonance Imaging: A Compelling Case for Cardiac
Magnetic Resonance’s Future. Circ. Cardiovasc. Imaging 2017;10:1–7 doi:
10.1161/CIRCIMAGING.117.005446.
2. Kellman P, Larson AC,
Hsu L-Y, et al. Motion‐corrected free‐breathing delayed enhancement imaging of
myocardial infarction. Magn. Reson. Med. 2005;53:194–200 doi:
10.1002/mrm.20333.
3. Campbell-Washburn AE,
Ramasawmy R, Restivo MC, et al. Opportunities in Interventional and Diagnostic
Imaging by Using High-Performance Low-Field-Strength MRI. Radiology
2019;293:384–393.
4. Hargreaves B.
Variable-Density Spiral Design Functions.
http://mrsrl.stanford.edu/~brian/vdspiral/.
5. Seiberlich N, Lee G,
Ehses P, Duerk JL, Gilkeson R, Griswold M. Improved temporal resolution in
cardiac imaging using through-time spiral GRAPPA. Magn. Reson. Med.
2011;66:1682–1688 doi: 10.1002/mrm.22952.