Jürgen Braun1, Heiko Tzschätzsch2, Clara Körting3, Marika Jenderka3, Angela Ariza de Schellenberger2, Toni Drießle4, Michael Ledwig4, and Ingolf Sack2
1Department of Medical Informatics, Charité - Universitätsmedizin Berlin, Berlin, Germany, 2Department of Radiology, Charité - Universitätsmedizin Berlin, Berlin, Germany, 3Beuth Hochschule für Technik Berlin, Berlin, Germany, 4Pure Devices GmbH, Würzburg, Germany
Synopsis
A compact
tabletop MR elastography (MRE) device was employed for rheological tests of
soft tissue samples to measure the change of viscoelastic powerlaw constants in
liver and brain tissue during progressive fixation. Shear-modulus dispersion
functions were acquired from 300 to 5700Hz in animal tissues at different
states of formaldehyde fixation and fitted by the rheological springpot-powerlaw
model. Formalin fixation reduced viscosity and increased elasticity of liver
tissue faster and to a higher degree than in brain tissue similar to the
alteration of mechanical properties observed by in vivo elastography of hepatic
fibrogenesis.
Purpose
Magnetic resonance elastography (MRE) shows a higher consistency of rheological constants of macroscopic soft biological tissue samples1 than surface based mechanical tests.2 We therefore developed a compact, automated tabletop MRE device and applied the new methods to test the change of viscoelastic powerlaw constants of tissue samples during progressive fixation. Fixation effects on tissue structures are highly relevant in pathology and transplantation medicine3 but have never been addressed by the springpot model – the most basic two-parameter powerlaw model in rheology.Methods
A tabletop MRI scanner (Pure Devices GmbH, Würzburg, Germany)
with 0.5T permanent magnet was customized to MRE (Fig. 1) as described in a
preliminary version in 4. Trapezoidal bipolar motion encoding
gradients (MEG, amplitude=0.7T/m) centered around the refocusing pulse of a
spin echo sequence were synchronized to the vibration frequency f. The number N of MEG-cycles varied from N=4
and N=78 for f=300 and 5700Hz, respectively. Further acquisition parameters: repetition
time=300ms, echo time=32ms, slice thickness=2mm, matrix size=64×64, field of view=12.8mm, driving frequencies
300-4500 (500-5700)Hz for porcine brain (calf liver), eight wave
dynamics, one average, total acquisition time for a single frequency including
toggling of MEG polarity=5.2min.
Fresh calf liver tissue and porcine brain tissue
(brain stem) were cut in cylindrical cores (7.5mm diameter, 5 to 8mm length) were
transferred into glass tubes for experiments. A total of 13 (14) samples of liver
(brain) were immersed in formalin (4% formaldehyde in buffer solution) and
measured at the following times (in hours): 0 (0, native sample), 1(1), 2(2),
3(3), 4(4), 5(5), 6(6), 7(7), 8(8), 10(10), 12(12), 14(14), 16(16), 18(18),
24(24), 26(26), 28(28), 36(36), 48(48), 55(50), 77(72).
Due to the constrained axial motion, cylindrical
shear waves emanated concentrically from the tissue border towards the center. Therefore,
motion field acquisition was limited to the component perpendicular to the transverse
image orientation. Postprocessing consisted of unwrapping phase difference data
and Fourier transform in time. Resulting 2D-wave fields at driving frequencies f (Fig. 2) were fitted by the analytical
solution of shear waves in a cylinder5,6 (Fig. 3). The calculated
complex-valued wave number was translated into two real-valued quantities
related to stiffness (shear wave speed c)
and viscous attenuation (shear wave penetration rate a).7 To derive shear modulus-related parameters, c and a were directly fitted by the springpot model which predicts a
monotonic increase of storage and loss modulus over increasing vibration
frequencies by the parameters μ (shear
modulus) and α (dimensionless
powerlaw exponent).Results
Reference
measurements showed very good agreement with high field MRE performed on a
preclinical scanner (Bruker PharmaScan 70/16, Ettlingen, Germany) using
ultrasound gel of the same batch (Fig.4). Unfixated tissue of porcine brain
stem was stiffer and higher dispersive than calf’s liver as reflected by μ=2.66kPa and α=0.66 versus μ=1.92kPa
and α=0.42. Upon fixation, liver μ increased faster and reached higher
values than brain μ, with decay times
of approximately six versus ten hours and asymptotic shear moduli of 564±40 versus 134±25kPa. α in
brain tissue decayed
towards 0.39±0.02 in <1 hour and towards 0.23±0.01 in liver tissue in approximately 3 hours. Figure 5 shows springpot fit
parameters μ and α versus the fixation time.Discussion
The harmonic stimulation of cylindrical waves
allowed us to reduce the acquired information to single-slice, single-component
encoding and analysis based on 2D-wave fitting. Unlike Laplacian-based
reconstruction the model-fitting approach, (i) needs no filtering, (ii) shows
no noise amplification, and (iii) is not affected by discretization artifacts up
to the Nyquist limit. Thus, resolution limits and low SNR within short scan
times can be alleviated making multifrequency MRE and viscoelastic dispersion
measurements feasible for low-field MRI.
Formalin as a crosslinker of proteins is known to
increase tissue stiffness and reduce damping properties. This behavior was well
reflected in the results obtained for tissue samples of porcine brain and calf
liver. Brain tissue stiffening occurs less correlated with the reduction of
loss properties compared to liver stiffening. For the latter formalin fixation
seems to simulate the progressive change of viscoelastic tissue properties
during hepatic fibrogenesis and might be a model of tissue mechanics and liver
fibrosis. Conclusion
The
progressive change of viscoelastic powerlaw constants of liver and brain tissue
samples during fixation could be measured by a compact tabletop MRE device at
0.5T in automated fashion. The motion characteristics of the integrated
piezoelectrical actuator allowed measurements in a range of mechanical
frequencies from 300 to 5700Hz and supported filter-free fit-based cylindrical
wave inversion. Formalin fixation changed liver tissue from highly viscous to
elastic properties more than brain tissue.Acknowledgements
No acknowledgement found.References
1. Reiter R, Freise C, Jöhrens K, et al.
Wideband MRE and static mechanical indentation of human liver specimen:
Sensitivity of viscoelastic constants to the alteration of tissue structure in
hepatic fibrosis. J Biomech. 2014;47:1665-1674.
2. Dittmann F, Hirsch S, Tzschaetzsch H, et al. In vivo wideband multifrequency MR elastography of the human brain and
liver. Magn Reson Med. 2016;76:1116–1126.
3. Ayyildiz M, Aktas RG, Basdogan C. Effect of solution and post-mortem
time on mechanical and histological properties of liver during cold
preservation. Biorheology. 2014;51:47-70.
4. Samavati N, Körting C, Drießle T, et al.
Compact and fully automated 3D multifrequency tabletop MR elastography for the
measurement of viscoelastic parameters in small tissue samples. Proc ISMRM.
2016:4653.
5. Yasar TK, Royston TJ, Magin RL. Wideband MR
elastography for viscoelasticity model identification. Magn Reson Med. 2013;70:479-489.
6.
Okamoto RJ, Clayton EH, Bayly PV. Viscoelastic
properties of soft gels: comparison of magnetic resonance elastography and
dynamic shear testing in the shear wave regime. Phys Med Biol. 2011;56:6379-6400.
7.
Tzschätzsch H, Guo J, Dittmann F, et al. Tomoelastography by multifrequency
wave number recovery from time-harmonic propagating shear waves. Med Image Anal. 2016;30:1-10.