Eva Oberacker1, Andre Kuehne2, Thomas Wilhelm Eigentler1, Cecilia Diesch1, Jacek Nadobny3, Pirus Ghadjar3, Peter Wust3, and Thoralf Niendorf1,2,4
1Berlin Ultrahigh Field Facility (B.U.F.F.), Max Delbrück Center for Molecular Medicine in the Helmholtz Association, Berlin, Germany, 2MRI.TOOLS GmbH, Berlin, Germany, 3Clinic for Radiation Oncology, Charité Universitätsmedizin, Berlin, Germany, 4Experimental and Clinical Research Center (ECRC), joint cooperation between the Charité Medical Faculty and the Max-Delbrück-Center for Molecular Medicine in the Helmholtz Association, Berlin, Germany
Synopsis
Ultrahigh field (UHF) MR employs higher radio frequencies
(RF) than conventional MR and has unique potential to provide focal temperature
manipulation and high resolution imaging (ThermalMR), as our simulations and
experimental work demonstrated for f=300MHz. Here we use EMF simulations to
demonstrate the added value for multi-frequency thermal intervention
(f=300MHz-500MHz) to boost the efficacy of the heating paradigm for clinical
target volumes (TVs) of two brain tumor patients. An optimization algorithm was
developed to study RF power deposition in the TV for a single, double or triple
frequency set permitting highest RF power deposition
in the TV.
Introduction
Hyperthermia has proven beneficial as an adjunct treatment
of Glioblastoma Multiforme (GBM)1. Advances in RF based hyperthermia
have evoked developments in the designs of annular2,3 or helmet
shaped4 RF applicators for the treatment of brain tumors.
Optimization algorithms used to confine the RF power deposition to the target
volume (TV) are under constant revision5,6. Introducing RF antenna
designs that support a wide frequency range up to Δf=300MHz enable the choice
of frequency to be based on the shaped and size of the TV rather than initial
engineering constraints7. Previous works claim that frequencies up
to 1GHz increase the focusing capabilities of RF applicators3,8.
Control over the resulting temperature distribution can be further increased by
driving wide band antennas at a RF frequency range that includes the Larmor
frequency of an MR system to enable anatomic MRI for treatment planning and MR
Thermometry (MRTh) for treatment supervision and dosage control, designated as
Thermal MR. Recognizing this opportunity, this work examines the efficacy of
multi-frequency thermal intervention targeting the clinical target volume of
two GBM patient-derived head voxel models using a 20 channel RF applicator
supporting a frequency range of 270-550MHz.Methods
For this simulation study, two head models with TV of a)
500ml (bounding box: 10.3x10.3x9.2cm³) and b) 170ml (bounding box:
5.7x6.4x6.8cm), were placed in an annular phased array Thermal MR RF applicator.
20 self-grounded wideband bowtie antenna building blocks9 were
azimuthally arranged in two rings around the head, with the center of the array
being aligned with the center of the TV in z-direction (Fig.1).
Broadband electromagnetic field (EMF) simulations
were performed using Sim4Life10. From the channelwise E- and H-field
data, SAR matrices were calculated, averaging over 10g of tissue11.
For phase and amplitude optimization, a quadratically constrained quadratic
program was simplified via its semidefinite relaxation12, which was
solved using the MatLab13-based optimization framework YALMIP14
with different constraints. The optimization goal is set to maximize total
power absorption in the TV:
maximize $$$ \sum_f tr\left(X_fQ_f\right) $$$
subject to $$$ \sum_f tr\left(X_fS_{f,i}\right) \leq SAR_{Lim} $$$ & $$$ \sum_f tr\left(X_fP_f\right) \leq P_{Lim} $$$
where tr=trace, f=frequency, Xf=frequency-specific
solution matrix, Q=tumor-SAR matrix, Si=SAR-matrices in the healthy
tissue and SARLim=40W/kg the exposure limit in the healthy tissue.
Head power deposition was constrained via the global SAR matrix P.
EMF simulations were performed for individual frequencies (300;
400; 500MHz). All possible combinations of frequencies were studied and
benchmarked against the single frequency results. For performance assessment of
the SAR10g based hyperthermia treatment planning the total RF power
delivered to the TV, PTV,
the performance indicator PI=SAR10g,max(TV)·SAF·TCSAR>Lim15
and the hotspot to target quotient HTQ were
determined.Results
Our
results demonstrate that the multi-frequency optimization algorithm was
successfully implemented. For the large TV, the best single frequency result is
obtained at the lowest investigated frequency (Table1, PTV(300MHz)=18.22W,
PTV(400MHz)=13.75W, PTV(500MHz)=14.42W). The total target
power achieved by the double frequency optimization is at least as good
(Table1, PTV(300MHz) = PTV(300+400Mz)=18.22W) as the
single frequencies used or better (PTV(300+500MHz)=19.38W, PTV(400+500MHz)=16.3W).
The triple frequency optimization identified the most powerful combination to
be a combination of 300MHz and 500MHz and sets the contribution of 400MHz to
zero in the additive process (Table1). The maximum intensity projections of the
additive SAR10g distributions are shown in Fig.2 for the large TV together with the double and triple frequency optimization results and the contributing SAR10g
distributions of each frequency.
For the small TV, the best single frequency result is obtained for the highest
investigated frequency (Table1, PTV(300MHz)=8.18W, PTV(400MHz)=7.92W,
PTV(500MHz)=9.49W), see Fig.4. Here, all double frequency results
outperform the single frequencies used in the combination (Table1, PTV(300+400Mz)=9.2W,
PTV(300+500MHz)=10.8W, PTV(400+500MHz)=10.3W). For the
small target volume, the triple frequency set proved to achieve the highest
total RF power deposition (PTV(300+400+500MHz)=10.9W).Discussion and Conclusion
Our results demonstrate that multi-frequency Thermal MR
holds the potential to increase the power delivered to the TV in the brain. Our
optimization algorithm successfully discarded a frequency that did not boost
the RF power delivery to the large TV and combined all frequencies used in this
study to obtain the largest RF power delivery to the small TV. Our results also
show that increasing the frequency does not necessarily result in a better
target delineation if the TV exceeds the wavelength at said frequency. The high
focusing at smaller wavelengths leads to insufficient coverage of the TV and
thus smaller total RF power deposition. For the small TV, the highest frequency
did show the best result for the single frequency optimization but was
outperformed by the triple frequency optimization.
To conclude, our preliminary results obtained for two
patient data sets motivate inclusion of more patient data sets into our
optimization algorithm to examine the correlation between the TV size and the
(combination of) frequencies. In this process, we will also enhance the
frequency sampling rate within the operational range of the RF applicator.Acknowledgements
This
project has received funding from the European Research Council (ERC) under the
European Union's Horizon 2020 research and innovation program under grant
agreement No 743077 (ThermalMR).References
1. Sneed,
P.K., et al., Survival benefit of
hyperthermia in a prospective randomized trial of brachytherapy
boost±hyperthermia for glioblastoma multiforme. International Journal of
Radiation Oncology* Biology* Physics, 1998. 40(2): p. 287-295.
2. Winter,
L., et al., Design and Evaluation of a
Hybrid Radiofrequency Applicator for Magnetic Resonance Imaging and RF Induced
Hyperthermia: Electromagnetic Field Simulations up to 14.0 Tesla and
Proof-of-Concept at 7.0 Tesla. PLoS One, 2013. 8(4): p. e61661.
3. Winter,
L., et al., Thermal magnetic resonance:
physics considerations and electromagnetic field simulations up to 23.5 Tesla
(1GHz). Radiation Oncology, 2015. 10(1):
p. 1.
4. Dobsicek
Trefna, H., et al. Antenna Applicator for
Microwave Hyperthermia Treatment of Pediatric Brain Cancer. in 8th European Conference on Antennas and
Propagation, EuCAP 2014, The Hague, The Netherlands 6-11 April 2014. 2014.
5. Rijnen,
Z., et al., Clinical integration of
software tool VEDO for adaptive and quantitative application of phased array
hyperthermia in the head and neck. International Journal of Hyperthermia,
2013. 29(3): p. 181-193.
6. Bellizzi,
G.G., et al., Selecting the Optimal
Subset of Antennas in Hyperthermia Treatment Planning. IEEE Journal of
Electromagnetics, RF and Microwaves in Medicine and Biology, 2019: p. 1-1.
7. Takook,
P., et al., Compact self-grounded Bow-Tie
antenna design for an UWB phased-array hyperthermia applicator.
International Journal of Hyperthermia, 2017. 33(4): p. 387-400.
8. Guérin,
B., et al., Computation of ultimate SAR
amplification factors for radiofrequency hyperthermia in non-uniform body
models: impact of frequency and tumour location. International Journal of
Hyperthermia, 2018. 34(1): p.
87-100.
9. Eigentler, T.W., et
al., Ultra-Wideband Self-Grounded Bow-Tie
Antenna Building Block for Thermal Intervention, Diagnostic MRI and MR
Thermometry at 7.0 Tesla. Proceedings of the 27th ISMRM Annual
Meeting and Exhibition, Montreal, Canada, 2019.
10. Sim4Life V3.4, ZMT,
Zurich, Switzerland
11. IEEE,
Recommended Practice for Determining the
Spatial‐Peak Specific Absorption Rate (SAR) in the human body Due to Wireless
Communications Devices: Measurement Techniques. 2003..
12. Luo,
Z.-Q., et al., Semidefinite relaxation of
quadratic optimization problems. IEEE Signal Processing Magazine, 2010. 27(3): p. 20.
13. MatLab, The MathWorks
Inc., Natick, Massachusetts, US
14. Löfberg J. YALMIP : A Toolbox
for Modeling and Optimization in MATLAB. In: Proceedings of the CACSD
Conference. Taipei, Taiwan; 2004.
15. Oberacker,
E., et al., Radiofrequency applicator concepts for RF hyperthermia treatment
and MR imaging of glioblastoma multiforme at 7.0 T (298 MHz). Proceedings of
the 26th Joint ISMRM-ESMRMB Annual Meeting and Exhibition, Paris,
France, 2018.