Maksym Yushchenko1, Mathieu Sarracanie1,2, Jens Wuerfel2,3, and Najat Salameh1
1Laboratory for Adaptable MRI Technology (AMT lab), Dpt of Biomedical Engineering, University of Basel, Allschwil, Switzerland, 2Medical Image Analysis Center (MIAC AG), Basel, Switzerland, 3Quantitative Bioimaging Group (qbig), Dpt of Biomedical Engineering, University of Basel, Allschwil, Switzerland
Synopsis
We describe the fabrication of a low-cost,
silicone-based phantom for elastography, which provides precise geometries and controlled
stiffness. Simulations were carried out to produce synthetic MRE data and
verify the performance of MRE reconstruction in ideal conditions. To our
knowledge, this is the first time that simulations can be validated on exact
same physical objects, making this approach very handy to optimize acquisition
parameters and establish reconstruction validity criteria of elastography data.
Such phantoms can be prepared with different mechanical properties, and could thus
be broadly used as an accessible mean to assess the robustness of MRE tools.
Introduction
Elastography is an imaging technique that assesses the
stiffness of body tissues remotely1,2. Several elastography
approaches exist that involve complex combinations of many interdependent
elements (vibration source, motion capture, reconstruction algorithm). Hence, it
is paramount to assess the accuracy of elastography protocols in phantoms
before applying it to patients, or before drawing diagnostic or research
conclusions.
Commercial phantoms are usually optimized only for
ultrasound elastography and are very expensive despite their lack of stability
over time. Typical elastography materials used for custom phantoms have
different drawbacks: water-based polymers lacking long-term stability, irregular
geometries and/or inhomogeneous mechanical properties3.
We propose to use a low-cost, silicone-based material
for custom-built elastography phantoms, which provides precise geometries and desired
stiffness properties in a reproducible way. We show that our reconstruction algorithm
can be validated by comparison with MRE results of ideal synthetic data, simulated
on a ground-truth phantom.Methods
Simulations of a CAD-designed ground-truth phantom were
done using the Structural Mechanics module of COMSOL Multiphysics© (COMSOL AB,
Sweden). The phantom comprises 4 inclusions in a homogeneous background (Fig
1a: cylinders of diameters 10, 20, 30, and 40 mm), and viscoelastic parameters were
set in the software (background: Gd = 5.1 kPa, Gl = 1.5
kPa; inclusions: Gd = 4.1 kPa, Gl = 1.4 kPa). Wave propagation
was simulated using a FEM forward approach for a given frequency (57 Hz) and
the displacement fields were extracted at the resolution employed in MRE (3 x 3
x 3 mm3 voxels, 9 slices). This information was converted into
synthetic raw phase data that was processed by curl-based direct inversion algorithm4.
The physical phantom was made using a custom 3D-printed
cast and a bicomponent RTV silicone Eurosil4 with the addition of a Eurosil Softener
(Schouten Syntec, Netherlands). Firstly, the background was created by filling
the mold (182 x 108 x 105 mm3) with the first silicone mixture
(ratio by weight 1A : 1B : 1.75 Softener) with solid 3D-printed cylinders (Ø
10, 20, 30, 40 mm) positioned inside. After 1 day of curing time, the cylinders
were extracted and the remaining holes were filled with a second mixture (1A :
1B : 1.9 S) to create softer inclusions. At the beginning of the curing
process, the silicone was degassed in a vacuum chamber.
MRE was performed at
3 T (Prisma – Siemens, Germany), using a custom 57 Hz vibration source5
and a fractional gradient-echo encoding sequence6. The phantom was
positioned in its cast to constrain the lateral faces’ movement, and the
acoustic vibration created at its surface. A 32-channel head coil was used for
the experiments. Raw data were processed using the same method used for the
synthetic data.Results
The silicone used for the phantom yields accurate and
precise contours of the inclusions (Fig 1b). No discontinuity arises between
the background and the inclusions as seen from the MR images (data not shown),
which ensures continuous wave propagation. Furthermore, the material preserves
its geometric shape without any visible degradation after 5 months.
The elastograms obtained from the simulation (Fig 2a)
clearly show the inclusions with stiffness values corresponding to the desired
parameters. On the elastograms of the physical phantom (Fig 2b), the three
largest inclusions are correctly detectable. In both elastograms,
non-homogeneous regions are present both in the background of the phantom and
in the inclusions.Discussion
We showed that our silicone-based material enables the
fabrication of objects with precise contours, helping to assess, for a given
elastography protocol, the minimum detectable object size, which can be
extremely useful for the detection of focal lesions such as tumors. This
phantom can be used in parallel to simulations assessing the performance of the
reconstruction method for different elastography protocols, by varying for
example spatial resolution and/or vibration frequency.
The elastogram obtained in the physical phantom showed
a central spot with higher stiffness that is due to strong waves in the region
in contact with the vibration source. One can also observe that the left side
of the elastogram is very heterogenous and this is due to poor wave propagation
in this area. The next step would consist in determining confidence values for
our elastograms based on those considerations.
Finally, in order to obtain stiffness values
comparable to biological tissues, we had to use softener concentrations higher
than what the material is designed for. This caused a slight leakage of the oil
softener, which might alter the phantom’s mechanical properties over time.
Further investigations will include rheological studies to assess the mechanical
stability of the silicone and fully characterize the relation between the
desired stiffness and the material composition.Conclusions
We showed that it is possible to fabricate a
silicone-based phantom with materials that are cheap, easy to handle, and
stable over time. Those phantoms can be prepared with different mechanical
properties and consist in a convenient tool for MRE methods’ validation.
To the best of our knowledge, this is the first time
that simulations can be validated on exact same physical objects, making this
approach very useful for optimizing acquisition parameters and establishing
reconstruction validity criteria of elastography data.Acknowledgements
No acknowledgement found.References
1. Muthupillai R, Lomas D, Rossman P, et al.
"Magnetic resonance elastography by direct visualization of propagating acoustic
strain waves." Science (80- ). 1995;269(5232):1854-1857.
2. Krouskop TA, Dougherty DR,
and Vinson FS. 1987. “A Pulsed Doppler Ultrasonic System for Making
Noninvasive Measurements of the Mechanical Properties of Soft Tissue.” Journal
of Rehabilitation R&D 24 (2): 1–8.
3. Solamen LM, McGarry MD, Tan L, et al. 2018. “Phantom Evaluations of Nonlinear Inversion MR Elastography.”
Physics in Medicine & Biology 63 (14): 145021.
4. Sinkus R, Tanter M, Xydeas T, et al. 2005. “Viscoelastic Shear Properties of in Vivo
Breast Lesions Measured by MR Elastography.” Magnetic Resonance Imaging 23 (2
SPEC. ISS.): 159–65.
5. Salameh N, Souris L, Sarracanie M, et al. 2011.
“Feasibility of Brain MR-Elastography at 1.5 T with a Novel Wave Generator : An
Animal Study” Proc ISMRM2011: 3492.
6. Garteiser P, Sahebjavaher RS, Ter Beek LC, et al. 2013. “Rapid
Acquisition of Multifrequency, Multislice and Multidirectional MR Elastography
Data with a Fractionally Encoded Gradient Echo Sequence.” NMR in Biomedicine 26
(10): 1326–35.