Zhongzheng He1, Paul Soullié1, Khalid Ambarki2, Pauline M Lefebvre1, and Freddy Odille1,3
1IADI U1254, INSERM, Université de Lorraine, Nancy, France, 2Siemens Healthcare SAS, Saint Denis, France, 3CIC-IT 1433, CHRU Nancy, INSERM, Université de Lorraine, Nancy, France
Synopsis
Keywords: Electromagnetic Tissue Properties, Electromagnetic Tissue Properties, MR EPT, conductivity, liver, brain, age-related changes
Motivation: Age-related study of electrical conductivity may be useful for personalized SAR modelling.
Goal(s): To evaluate in-vivo conductivity changes in a preliminary study in adults (brain and liver), using MR Electrical Properties Tomography, at 3T.
Approach: We used a UTE sequence to obtain RF-weighted images suitable for conductivity reconstruction. Data from 10 subjects were analyzed: in the brain, white matter and grey matter values were quantified; in the liver, fat volume fraction was also obtained for comparison.
Results: In-vivo conductivities were close to literature values, except for white matter which was higher in vivo. A significant dependency with age was found in the liver.
Impact: Knowledge
of electrical properties of tissues is required for SAR modelling. Currently, values from ex-vivo/animal
studies are used. In-vivo measurements by MR-EPT could bring valuable insights,
in particular age-related changes, if existing, could be considered for MRI
safety studies.
INTRODUCTION
Knowledge
of age-related changes of electrical conductivity (𝜎) is important for MRI safety modelling, in order to get personalized
specific absorption rate estimation. MR electrical properties (EPs) tomography
(MR-EPT) is a non-invasive imaging technique that reconstructs the EPs of biological
tissues using high radiofrequency (RF) signature pulse sequences. In this study,
we validated MR-EPT against ground truth probe measurements in phantom
experiments, and we performed a preliminary in-vivo investigation in adults,
including brain and liver measurements. In particular, we sought to study
age-related changes of conductivity. It may be assumed that liver conductivity
varies with fat volume fraction, which is age-dependent 1.MATERIALS AND METHODS
MR data acquisition: MRI measurements
were performed on 10 volunteers (5 male, 5 female, 32.5 ± 11 years old
and BMI ranging from 19 to 32 kg/m²) using a 3T MR scanner (Siemens Prisma,
Erlangen, Germany). For all scans, a 2-port body coil was employed for transmission
while different combinations of surface coils were used for reception. 3D
images were acquired using a UTE Spiral VIBE sequence 2 (TE = 50 μs, FA = 3°, voxel size = 1x1x1 mm3
(phantom) / 1.2x1.2x1.2 mm3 (brain) / 2×2×2.5 mm3 (liver)).
Phantom
(agar and various concentrations of NaCl) scans and ground truth probe measurements
were made following He Z. et al. 3, in the 0~ 3S/m range.
The
liver water and fat maps were acquired in a 18 s breath-hold using a dual-echo Dixon
VIBE sequence and then reformatted to match the dimensions of UTE scan. MPRAGE
was acquired for brain anatomy.
Electrical properties
reconstruction: Image-based EPT 4 uses the complex image of a
UTE/ZTE sequence with a low flip angle (I≈ B0 =B1+B1-), to reconstruct conductivity with:
σ = Re(ΔB/(jωµ0B)) , where μ0 is the vacuum permeability, ω is the angular frequency.
To mitigate noise amplification resulting from the differentiation
operator, we computed the second derivatives of B using a large 3D second-order
Savitzky-Golay filter. We employed a filter size of [11 11 11] for phantoms/liver
and [7 7 7] for the brain. Furthermore, after differentiation, we applied a [3
3 3] median filter to reduce boundary effects.
Image analysis: For
liver data, the FVF maps 5 were generated by combining the fat image (F)
and water image (W) as:
FVF= F/(F+W)×100%.
The liver segmentation was performed
with k-means algorithm (k=5) and corrected manually.
For brain data, white matter
and grey matter were segmented using software SPM 12 (WTCN, UCL, London, UK). All
ROIs were eroded with a disk-shaped structure element with a radius of 2.
RESULTS
Results in phantoms (Figure 1)
show excellent agreement between ground truth and MRI measurements. Brain conductivity
maps of two subjects are shown in Figure 2. In-vivo conductivity mean values in
the brain were: 0.56 ± 0.09 S/m (grey
matter, GM), 0.71 ± 0.04 S/m (white matter, WM). The WM value was higher than that
reported in the literature for ex-vivo human studies 6–8 : 0.34 S/m (WM). No significant changes with age (Figure
3) were observed (p-value > 0.05). Liver FVF and conductivity maps of two subjects
are shown in Figure 4. In-vivo conductivity values in the liver were 0.55 ± 0.16
S/m (vs. 0.51 S/m in the literature 6–8). Significant correlations (Figure 5) were found between age
and conductivity (r = -0.814, p = 0.004), and between age and FVF (r = 0.688, p
= 0.028).DISCUSSION
It
is expected to measure higher conductivities in-vivo compared to ex-vivo, as
shown in Peyman et al., 2007 9 in animals (0.09 S/m higher
values in-vivo at 150 MHz in pigs). However, this may not explain the difference
measured in this study which is larger 0.71 S/m in WM. Few ex-vivo data are available. Age-related
changes in the brain were not expected in the age range of this study (23 to 54
years old), but would be expected between children/adults/elderly people10. Age-related changes in the
liver are expected, and our results in comparison
with FVF are consistent with the literature 1.CONCLUSION
This
study needs to be confirmed by larger studies to clarify the actual range of
conductivity values in different organs in-vivo, and check whether it is
relevant to use values from ex-vivo or animal studies. Age-related changes also
need to be investigated more thoroughly, as shown in the liver here, where conductivity
ranged from 0.33 to 0.81 S/m. This could have a significant impact on SAR
studies.Acknowledgements
Funding:
ANR-21-CE19-0040 (ELECTRA project), CPER IT2MP, FEDER (European Regional Development
Fund). The authors thank Thomas Benkert (PhD, Application Development, Siemens
Healthcare GmbH, Erlangen, Germany) for providing the UTE sequence.References
- Ulbrich EJ, Fischer MA, Manoliu A, et al. Age- and Gender Dependent Liver Fat Content in a Healthy Normal BMI
Population as Quantified by Fat-Water Separating DIXON MR Imaging. Zhang Z, ed.
PLoS ONE. 2015;10(11):e0141691. doi:10.1371/journal.pone.0141691.
- Mugler JP, Fielden SW, Meyer CH, et al.
Breath‐hold UTE lung imaging using a stack-of-spirals acquisition. In: 23rd
Annual Meeting International of the Society for Magnetic Resonance in Medicine.
Toronto, Canada; 2015:1476.
- He Z, Doguet M, Soullié P, De Sousa PL, Lefebvre
P, Odille F. Comparison and validation of multiple MR-EPT methods with ground
truth vector network analyzer measurements. In: London, England, UK; 2022:2910.
doi:10.58530/2022/2910
- Lee SK, Bulumulla S, Wiesinger F, Sacolick L,
Sun W, Hancu I. Tissue Electrical Property Mapping from Zero Echo-Time Magnetic
Resonance Imaging. IEEE Trans Med Imaging. 2015;34(2):541-550.
doi:10.1109/TMI.2014.2361810
- Reeder SB, Sirlin CB. Quantification of Liver Fat
with Magnetic Resonance Imaging. Magnetic Resonance Imaging Clinics of North
America. 2010;18(3):337-357. doi:10.1016/j.mric.2010.08.013
- Gabriel C, Gabriel S, Corthout E. The dielectric
properties of biological tissues: I. Literature survey. Phys Med Biol.
1996;41(11):2231-2249. doi:10.1088/0031-9155/41/11/001
- Gabriel S, Lau RW, Gabriel C. The dielectric
properties of biological tissues: II. Measurements in the frequency range 10 Hz
to 20 GHz. Phys Med Biol. 1996;41(11):2251-2269.
doi:10.1088/0031-9155/41/11/002
- Gabriel S, Lau RW, Gabriel C. The dielectric
properties of biological tissues: III. Parametric models for the dielectric
spectrum of tissues. Phys Med Biol. 1996;41(11):2271-2293.
doi:10.1088/0031-9155/41/11/003
- Peyman A, Holden SJ, Watts S, Perrott R, Gabriel
C. Dielectric properties of porcine cerebrospinal tissues at microwave
frequencies: in vivo , in vitro and systematic variation with
age. Phys Med Biol. 2007;52(8):2229-2245. doi:10.1088/0031-9155/52/8/013
- Gabriel C, Peyman A. Dielectric Properties of
Biological Tissues; Variation With Age. In: Conn’s Handbook of Models for
Human Aging. Elsevier; 2018:939-952. doi:10.1016/B978-0-12-811353-0.00069-5