Huiming Dong1,2, Richard D White1,3, and Arunark Kolipaka1,2,3
1Department of Radiology, The Ohio State University Wexner Medical Center, Columbus, OH, United States, 2Department of Biomedical Engineering, The Ohio State University, Columbus, OH, United States, 3Internal Medicine-Division of Cardiology, The Ohio State University Wexner Medical Center, Columbus, OH, United States
Synopsis
In vivo MR elastography (MRE)
allows non-invasive estimation of aortic stiffness. Recently, the feasibility
of multi-slice free-breathing (FB) spin-echo echo-planar imaging (SE-EPI)
aortic MRE was investigated. In this study, a single breath-hold (SBH) SE-EPI aortic
MRE protocol was proposed and compared to FB SE-EPI aortic MRE. Moreover, in vivo FB and SBH SE-EPI aortic MRE were
studied using both1.5T and 3T scanners. The main aim was to assess
inter-scanner reproducibilities of FB and SBH SE-EPI aortic MRE.
Introduction
Recent development in aortic MR elastography
(MRE) has provided the opportunity to non-invasively study in vivo aortic stiffness in the healthy aorta [1-3] as well as in
the aorta with abnormalities [4-6]. Aortic stiffness can be a potential biomarker for
multiple cardiovascular diseases, such as systemic arterial hypertension (SAH)
[7] and abdominal aortic aneurysm (AAA) [8].
Recently, the feasibility of multi-slice free-breathing
(FB) SE-EPI aortic MRE and its comparison to the established GRE aortic MRE was
investigated at 3T [9]. In the present study, a single breath-hold (SBH) SE-EPI
aortic MRE protocol was proposed to acquire multi-slice and multi-directional in vivo aortic MRE data, reducing the
potential measurement errors caused by prolonged scan time and inconsistent
breath-holds.
Currently, the performance of FB and SBH SE-EPI
aortic MRE on a 1.5T scanner remains unknown. Compared to 3T scanners, 1.5T
scanners are more commonly employed for cardiovascular imaging. However, signal-to-noise
ratio (SNR) is reduced due to lower field strength when similar imaging
parameters are applied, which can potentially bias the stiffness estimates.
Therefore, the main goal of this work is to
study the inter-scanner reproducibilities of in vivo SE-EPI aortic MRE when using (1) a FB protocol and (2) a SBH
protocol.Methods
A cardiac-gated SE-EPI MRE sequence was implemented. Prospective gating was used to choose the appropriate trigger delay for
avoiding flow- or cardiac motion-induced artifacts by acquiring data during diastole.
The external mechanical vibration was synchronized to the cardiac gating signal
and triggered immediately after detecting the gating signal for the first MRE
phase offset. The number of vibration cycles was calculated based on the total
duration of the selected trigger delay and TR. The other MRE phase offsets were
triggered by shifting the starting time of the external mechanical vibration
accordingly.
To further reduce the effect of aortic flow on
MRE phase images, the motion-encoding gradient (MEG) was designed to be both zero-
and first-moment-nulled. Moreover, the first moment of the readout gradient was
nulled for the center k-space line. EPI ghosting and distortion were
effectively corrected via a non-phase-encoded reference scan and protocol
optimization. To further reduce the undesired signal that can potentially cause
ghosts, a saturation band was used to suppress the unwanted signal from the
anterior surface of the subject (i.e., eliminating signals from the anterior
abdomen and chest). Chemical shift and off-resonance artifacts were reduced
through gradient reversal technique [10].
In this ongoing study, 4 healthy volunteers
(age: 33±9.26 years) were recruited. MR imaging was performed on a 1.5T
(MAGNETOM Aera, Siemens Healthcare, Erlangen, Germany) and 3T (MAGNETOM Prisma,
Siemens Healthcare, Erlangen, Germany) MR scanners using the developed cardiac-gated
SE-EPI MRE sequence. Imaging parameters for FB and SBH protocols are summarized
in Table I.
TR was selected to accommodate three slices.
In the first scan, both FB and SBH scans were
performed during the same sitting. To determine the inter-scanner reproducibility
of SE-EPI aortic MRE, each volunteer was also scanned on a different scanner at
a different site after the first scan.
Aortic MRE data were processed using MRElab
(Mayo Clinic, Rochester, MN). Eight 4th-order Butterworth band-pass
directional filters with cutoff of 1-40waves/FOV were used to eliminate the
undesirable noise, longitudinal and reflected waves. Subsequently, 3D Local-Frequency
Estimation (LFE) inversion was performed to obtain the weighted effective stiffness
map from each motion-encoding direction [11]. The weighing was based on the
first-harmonic amplitude from each motion-encoding direction.
Data are presented as mean ± standard
deviation. Bland-Altman analysis was performed to assess inter-scanner
reproducibility. Results and Discussion
No significant artifacts were observed in FB or
SBH SE-EPI MRE. Figure
1 demonstrates the magnitude images and raw MRE phase offsets.
Discernible waves were observed in both FB and SBH
SE-EPI scans. Figure
2 displays the anatomical image, snapshots of propagating waves
within the aorta and the corresponding stiffness maps measured using two
different scanners. Similar aortic stiffness maps were observed in all cases.
The reported aortic stiffness is the average stiffness across 3 slices.
Table II shows
the effective aortic stiffness for individual volunteers. FB and SBH SE-EPI aortic MRE yielded similar stiffness measures except in one volunteer (i.e., volunteer 3) who experienced variable heart rates and high flow during the second scan.
FB and SBH SE-EPI aortic MRE are reproducible on
both scanners except for volunteer 3. Figure 3(a-d) demonstrate Bland–Altman plots. Conclusion
The inter-scanner reproducibilities of FB and
SBH SE-EPI aortic MRE were assessed in 4 volunteers at 1.5T and 3T. FB SE-EPI
aortic MRE can potentially be a good alternative for patients who have
challenges in holding their breath. To comprehensively assess the
reproducibilities of FB and SBH SE-EPI aortic MRE on both 1.5T and 3T scanners,
further investigation is needed by recruiting additional volunteers.Acknowledgements
The authors acknowledge grant sponsor: NIH–NHLBI (grant number: NIH-R01HL124096). References
1. Damughatla AR, Raterman B, Sharkey-Toppen T,
et al. Quantification of aortic stiffness using MR Elastography and its
comparison to MRI-based pulse wave velocity. J Magn Reson Imaging.
2015;41(1):44-51. doi:10.1002/jmri.24506
2. Kenyhercz
WE, Raterman B, Illapani VSP, et al. Quantification of aortic stiffness using
magnetic resonance elastography: Measurement reproducibility, pulse wave
velocity comparison, changes over cardiac cycle, and relationship with age. Magn
Reson Med. 2016;75(5):1920-1926.
3. Woodrum DA, Romano AJ, Lerman A, et al.
Vascular wall elasticity measurement by magnetic resonance imaging. Magn Reson
Med. 2006;56(3):593-600. doi:10.1002/mrm.20991
4. Dong H, Mazumder R, Illapani VSP, Mo X,
White RD, Kolipaka A. In vivo quantification of aortic stiffness using MR
elastography in hypertensive porcine model. Magn Reson Med.
2017;78(6):2315-2321.
5. Kolipaka A, Illapani VSP, Kenyhercz W, et
al. Quantification of abdominal aortic aneurysm stiffness using magnetic
resonance elastography and its comparison to aneurysm diameter. J Vasc Surg.
2016;64(4):966-974.
6. Kolipaka A, Woodrum D, Araoz PA, Ehman RL.
MR elastography of the in vivo abdominal aorta: A feasibility study for
comparing aortic stiffness between hypertensives and normotensives. J Magn Reson
Imaging. 2012;35(3):582-586.
7. Payne RA, Wilkinson IB, Webb DJ. Arterial
stiffness and hypertension: Emerging concepts. Hypertension. 2010.
doi:10.1161/HYPERTENSIONAHA.107.090464
8. Vorp DA, Vande Geest JP. Biomechanical
determinants of abdominal aortic aneurysm rupture. Arterioscler Thromb Vasc
Biol. 2005;25(8):1558-1566.
9. Dong H, Jin N, Kalra P, White R, Kolipaka A.
In Vivo Quantification of Aortic Stiffness using a Multi-Slice Spin-Echo
Echo-Planar Imaging Sequence: A Comparison to a Gradient-Recalled Echo Sequence.
In: 27th Scientific Meeting of International Society of Magnetic Resonance in
Medicine (ISMRM). Montreal, QC; 2019.
10. Nagy Z, Weiskopf N. Efficient fat
suppression by slice-selection gradient reversal in twice-refocused diffusion
encoding. Magn Reson Med. 2008. doi:10.1002/mrm.21746
11. Knutsson H, Westin CF, Granlund G. Local
multiscale frequency and bandwidth estimation. In: 1st International Conference
on Image Processing. 1994: 36–40.