Wen-Tung Wang1, Dzung Pham1, and John A Butman1,2
1Center for Neuroscience and Regenerative Medicine, NIH/USU, Bethesda, MD, United States, 2Radiology and Imaging Sciences, NIH, Bethesda, MD, United States
Synopsis
Ultrashort echo-time (UTE) imaging can detect short- and ultrashort-T2
tissue components. e.g. tendons, ligaments, and cortical bone. Multi-echo UTE is used to generate the tissue
attenuation maps required for quantitative MRI PET, as short T2 skull is
visible on the first echo but not on the second. For accurate classification, it is assumed
that these are registered. Here we show that
geometric scaling issues of fat and water may be different between the 1st
and 2nd echoes – and hence lead to erroneous tissue classification.
INTRODUCTION
Multi echo UTE has been used for identification of short T2 tissues
based on the large drop in signal between the echoes1, 3,
particularly the first and second echoes, to identify ultrashort T2 tissues,
which are otherwise indistinguishable from air. For signal change between the two echoes to be
valid, there is an implicit assumption that the data from the two echoes at
each voxel correspond to the same tissue, i.e. the first and second echoes are
anatomically coregistered. Here we show, for particular implementations of a
UTE pulse sequence with a kooshball readout, that (1) the first echo is overall
scaled smaller than the second and (2) that within the first echo, chemical
shift results in an overlap between fat and water that is not present in the
second echo. METHODS
MRI acquisition:
MR
images were acquired using the vendor provided
two-echo 3D UTE pulse sequence with a kooshball readout (ute3d2_ns) on a
Siemens Biograph mMR 3T MRI system (Siemens, Erlangen, Germany) at software
VE11P. Contrast parameters were TR = 11.94 ms, TE1 = 0.07 ms, TE2 = 2.46 ms, and
flip angle = 10°. Geometric parameters were: image matrix 192×192×192, and
voxel size of 1.56×1.56×1.56 mm.
Phantom Experiments
Two
concentric oil/water phantoms were used, with an approximately 10 cm diameter
inner container fitted concentrically within a 15 cm diameter container. For phantom 1, the “W-o” phantom, vegetable
oil filled the inner container and water in the outer ring. For phantom 2, the “O-w” phantom, the
contents were reversed.
Healthy volunteer experiments:
HV volunteers were scanned under an
IRB approved protocol (NCT00001711) using the above MRI acquisition. UTE images of the head were obtained.RESULTS
In Figure 1, first- and second-echo images of kooshball UTE
acquisitions from the two phantoms were compared. Dependent on whether the oil
was at the center or not, there were a gap (W-o phantom) or an overlap (O-w
phantom) between the water and oil. This was clearly seen in the difference and
ratio images. Furthermore, the diameter of the phantoms on TE1 was smaller than
on TE2. Similar off-resonance effects were seen in brain images, as shown in
Figure 2. The head size on TE1 was smaller than on TE2. On TE1, skull bone
marrow fat overlapped with meninges, which is between the skull and brain.DISCUSSION
Because both echoes of the UTE are obtained synchronously,
true anatomical coregistration between the two sets of data is expected.
However, it may be the case that the gradients played out during acquisition
are not necessarily identical, perhaps due to sequence design or hardware
limitations. Furthermore, chemical shift between fat and water can result in
differential overlap of fat and water between the two echoes. Both
processes result in artifactual signal changes between the 1st and 2nd
echo which can be misattributed to the presence of e.g. bone.
We showed the degradation of UTE image quality due to
off-resonance fat signals. It manifests as artifactual rings, erroneous overlap
with water signal, and incorrect overall scaling of the first-echo images. It
is essential to have proper corrections, such as fat-water separation2, 5, 8,
careful gradient calibrations4, 7, and/or improved image
reconstruction, so that multi-echo UTE imaging can give accurate quantitation
of altra-short T2 components. Without corrections, incorporating the altrashort
TE image with late TEs and other images may workaround the issue of fat-water
chemical shift. For example, improved µmaps can be calculated based
non-supervised training using two-echo UTE images and CT images6.CONCLUSION
It is essential to correct off-resonance effects for
multi-echo UTE imaging to become a commonly useful imaging modality. Acknowledgements
This study was supported by the Department of Defense in the Center for Neuroscience and Regenerative Medicine and the Intramural Research Program of the National Institutes of Health. References
1 T. Boucneau, P. Cao, S. Tang, M.
Han, D. Xu, R. G. Henry, and P. E. Z. Larson, 'In Vivo Characterization of
Brain Ultrashort-T2 Components', Magn
Reson Med, 80 (2018), 726-35.
2 E. K. Brodsky, J. H. Holmes, H.
Yu, and S. B. Reeder, 'Generalized K-Space Decomposition with Chemical Shift
Correction for Non-Cartesian Water-Fat Imaging', Magn Reson Med, 59 (2008), 1151-64.
3 J. Du, G. Ma, S. Li, M. Carl, N.
M. Szeverenyi, S. VandenBerg, J. Corey-Bloom, and G. M. Bydder, 'Ultrashort
Echo Time (Ute) Magnetic Resonance Imaging of the Short T2 Components in White
Matter of the Brain Using a Clinical 3t Scanner', Neuroimage, 87 (2014), 32-41.
4 K. H. Herrmann, M. Kramer, and J.
R. Reichenbach, 'Time Efficient 3d Radial Ute Sampling with Fully Automatic
Delay Compensation on a Clinical 3t Mr Scanner', PLoS One, 11 (2016).
5 S. B. Reeder, Z. Wen, H. Yu, A. R.
Pineda, G. E. Gold, M. Markl, and N. J. Pelc, 'Multicoil Dixon Chemical Species
Separation with an Iterative Least-Squares Estimation Method', Magn Reson Med, 51 (2004), 35-45.
6 S. Roy, W. T. Wang, A. Carass, J.
L. Prince, J. A. Butman, and D. L. Pham, 'Pet Attenuation Correction Using
Synthetic Ct from Ultrashort Echo-Time Mr Imaging', J Nucl Med, 55 (2014), 2071-7.
7 H. Tan, and C. H. Meyer,
'Estimation of K-Space Trajectories in Spiral Mri', Magn Reson Med, 61 (2009), 1396-404.
8 D.
Wang, N. R. Zwart, and J. G. Pipe, 'Joint Water-Fat Separation and Deblurring
for Spiral Imaging', Magn Reson Med,
79 (2018), 3218-28.