Ultra-Wideband Self-Grounded Bow-Tie Antenna Building Block for Thermal Intervention, Diagnostic MRI and MR Thermometry at 7.0 Tesla
Thomas Wilhelm Eigentler1, Lukas Winter2, Haopeng Han1, Eva Oberacker1, Andre Kuehne3, Laura Boehmert1, Hana Dobsicek Trefna4, and Thoralf Niendorf1,3,5

1Berlin Ultrahigh Field Facility, Max-Delbrück Center for Molecular Medicine in the Helmholtz Association, Berlin, Germany, 2Physikalisch-Technische Bundesanstalt (PTB), Berlin, Germany, 3MRI.TOOLS GmbH, Berlin, Germany, 4Department of Electrical Engineering, Chalmers University of Technology, Gothenburg, Sweden, 5Experimental and Clinical Research Center (ECRC), a joint cooperation between the Charité Medical Faculty and the Max-Delbrück Center for Molecular Medicine, Berlin, Germany


A compact resonator antenna for broadband thermal intervention, diagnostic proton (1H) and fluorine (19F) imaging and MR-Thermometry was developed for operating at 7.0T MRI. The antenna is based on the concept of a self-grounded bow-tie (SGBT) antenna to enable thermal intervention frequencies ranging from 250MHz to 516MHz. The self-grounded bow-tie resonator antenna is smaller and lighter compared to previous dipole-antenna designs, enabling a high-density multi-channel array. The proposed antenna is demonstrated to be suitable for diagnostic imaging, thermal intervention, and MR thermometry.


Temperature is a physical parameter with diverse biological implications and crucial clinical relevance 1. Targeted temperature increase can be beneficial for thermal manipulation of inflamed tissues, treatment of cancer by potentiating cells for chemo- and radiotherapy, targeted drug delivery afforded by thermo-responsive nano-carriers and increased blood-brain barrier permeability 2,3. Focal and targeted temperature manipulation facilitated by radiofrequency applicators is highly dependent on the operating frequency, which ranges between 50 and 1000MHz for published implementations 2,4,5. Recent reports on self-grounded bow-tie SGBT antennae use a broadband approach, enabling frequency and focal temperature adjustment 6,7. Here we report on the development of a compact ultra-wideband self-grounded bow-tie resonator antenna that supports targeted thermal intervention, diagnostic proton (1H) MRI, 1H MR temperature mapping and fluorine (19F) MRI in a single device on a 7.0T MRI system.


CST Microwave Studio 2018 (CST–Computer Simulation Technology GmbH, Darmstadt, Germany) was used for electromagnetic field (EMF) and temperature simulations. The antenna and resonator geometry optimization was performed on a rectangular phantom (240x240x150mm³) with muscle tissue mimicking material 8. For obtaining the resonator and antenna dimensions, a genetic algorithm was used for optimization, pursuing the reflection coefficient S11<-13dB for the targeted operating bandwidth. Based on the findings, an antenna was manufactured and placed in an additive manufactured housing filled with deuterium oxide (99.9% D2O, Sigma Aldrich GmbH, Munich, Germany). An exponential tapered strip-line balanced-unbalanced-transformer with an overall size of 39.3x26.5mm² was employed to enable wideband response 9,10. A phantom with muscle tissue mimicking material was constructed for MR imaging and radiofrequency (RF) heating experiments. The dielectric properties of the phantom were measured for correct reproduction in the EMF and temperature simulations. Phantom density (1230.89 g/l), heat capacity (2.9635 J/g/K) and thermal conductivity (0.4355 W/m/K) estimations were based on 11,12. Transmission (B1+) fields at f=280 MHz (19F) and f=298 MHz (1H) were examined numerically and validated with pre-saturation based B1+ mapping 13. Temperature and SAR simulations based on IEEE/IEC standard 62704-1 were performed at f=300MHz, f=400MHz, and f=500MHz. Magnetic resonance thermometry was conducted using a 7.0T whole-body MRI system (MAGNETOM, Siemens Healthineers, Erlangen, Germany). For this purpose proton resonance frequency shift (PRFS) thermometry based on double echo method (TR=102ms, TE1=2.26ms and TE2=6.34ms at a special resolution of 1.5x1.5x4.0mm³) was combined with fiberoptic probe measurements (Neoptix, Québec, Canada) 14.


Figure 1 shows the antenna building block design and the reflection coefficient S11 for a bandwidth of 467MHz on muscle tissue. The dielectric parameters of the phantom are shown in Figure 2 (a). Figure 2 (b-c) shows the reflection coefficient S11 of the resonator antenna when placed on a phantom and on the thigh of the human voxel model Duke 15. Results of EMF simulations for B1+ at f=280MHz and f=298MHz together with B1+ mapping at f=298MHz of the phantom are demonstrated in Figure 3. B1+ simulations and measurements show a difference of approximately 8% offering transmission fields of >15μT/√kW at 40mm distance to the antenna which is very well suitable for 1H MRI and MR themometry. The losses of the imaging signal chain of -2.12dB at f=298MHz were considered in the simulations. SAR and temperature simulation results obtained for f=300MHz, f=400MHz, and f=500MHz are shown in Figure 4 (a-c). The heating transmit chain exhibited frequency-dependent losses of -1.95dB, -2.26dB and -2.59dB for the 3 thermal intervention frequencies. The experimental results obtained from MR thermometry matched the temperature simulations qualitatively and quantitatively. Figure 4 (d-e) highlights the results deduced from MR thermometry mapping and fiberoptic probe temperature reference measurements for f=300MHz, f=400MHz, and f=500MHz. Considering the match of MR thermometry and the fiberoptic probes, a difference of <1K for the first two measurement positions was achieved.

Discussion and Conclusion

The presented compact SGBT resonator broadband antenna supports thermal intervention, diagnostic imaging, and temperature mapping at 7.0T, while offering a frequency bandwidth ranging from 250MHz to 516MHz. The proposed SGBT building block pushes the boundaries of antenna design by affording 54.9% or 72.2% reduced antenna volume than previously reported for SGBT (107x78x31mm³) or bow-tie (150x70x40mm³) antenna configurations 4,7. In addition to 19F and 1H imaging, this antenna supports RF-induced heating for a broad frequency range. Temperature simulations, MR thermometry and temperature measurement showed good agreement between simulations and measurements. To take this project to the next level, we envision a high-density array of SGBT building blocks customized for 1H and 19F MRI, broadband thermal interventions, and MR-thermometry all being integrated into a 7.0T MRI scanner.


This project was funded in part by an advanced ERC grant (EU project ThermalMR: 743077).


1. Lamb G M Gedroyc W M Interventional magnetic resonance imaging. Br J Radiol. 1997;70:81-88.

2. Wust P, et al. Hyperthermia in combined treatment of cancer. Lancet Oncol. 2002;3(8):487-497.

3. Issels R, et al. Effect of Neoadjuvant Chemotherapy Plus Regional Hyperthermia on Long-term Outcomes Among Patients With Localized High-Risk Soft Tissue Sarcoma. JAMA Oncol. 2018;4(4):483

4. 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):e61661.

5. Guérin B, et al. Computation of ultimate SAR amplification factors for radiofrequency hyperthermia in non-uniform body models: impact of frequency and tumour location. Int J Hyperth. 2018;34(1):87-100.

6. Takook P, et al. Compact self-grounded Bow-Tie antenna design for an UWB phased-array hyperthermia applicator. Int J Hyperth. 2017;33(4):387-400.

7.Winter L, et al. Ultrahighfield, One for all: Ultra-wideband (279-500MHz) self-grounded bow-tie antenna for MR. ISMRM-ESMRMB. 2018:#4281.

8. IT’IS Foundation. Tissue Properties Database V4.0. 2018.

9. Yang J, Kishk A. A Novel Low-Profile Compact Directional Ultra-Wideband Antenna: The Self-Grounded Bow-Tie Antenna. IEEE Trans Antennas Propag. 2012;60(3):1214-1220.

10. Kazemipour A, Begaud X. Calculable Dipole Antenna for EMC Measurements with Low-Loss Wide-Band Balun from 30 MHz to 2 GHz. Electromagnetics. 2005;25(3):187-202.

11. Asadi M. Beet-Sugar Handbook. John Wiley & Sons; 2007

12. Buchanan EJ. Economic Design and Operation of Process Heat Exchange Equipment. Proc S Afr Sug Technol Ass. 1966;44:89-101

13. Yarnykh VL. Actual flip-angle imaging in the pulsed steady state: A method for rapid three-dimensional mapping of the transmitted radiofrequency field. Magn Reson Med. 2007;57(1):192-200.

14. Rieke V, Pauly KB. MR thermometry. J Magn Reson Imaging. 2008;27(2):376-390.

15. Christ A, et al. The Virtual Family - Development of surface-based anatomical models of two adults and two children for dosimetric simulations. Phys Med Biol. 2010;55(2)


Figure 1: Optimization result of the resonator cavity, balun and Self-Grounded Bow-Tie antenna design (a). Additive manufactured resonator cavity with antenna design inside the housing and balun on the resonator top (b). Numerical reflection coefficient S11 of the resonator antenna on a phantom with 20 mm water bolus between antenna and phantom (c).

Figure 2: Measured phantom material parameter with sucrose (994.0 g/l), NaCl (38.8 g/l), Agar (2.0 g/l) and CuSO4 (0.75 g/l) (a). Simulated and measured reflection coefficient S11 on Phantom (b), on the thigh of the human voxel model Duke and a human suspect (c).

Figure 3: Transmission (B1+) field at f=280MHz and f=298MHz of the resonator antenna on the phantom (a). Comparison of the B1+ field along the center line of simulations of Duke’s thigh and the phantom at f=280MHz and f=298MHz and the mean and interquartile range of the center area of the phantom measurement (b). Pre-saturation based experimental B1+ mapping result of the phantom (c). Transmission (B1+) field at f=280MHz and f=298MHz of the resonator antenna on the thigh of the human voxel model Duke (d).

Figure 4: Specific Absorption Rate (SAR) averaged over 10g tissue/material at f=300MHz, f=400MHz and f=500MHz of the torso phantom (a) and human voxel model Duke's thigh (b). Temperature simulation results using the phantom (c) after 2 min thermal intervention with measured amplifier output of 90.2W, 79.8W and 80.7W for f=300MHz, f=400MHz, and 500MHz. d) Temperature difference maps obtained from proton resonance-frequency (PRF) shift MR thermometry. The positions of the fiber optic probes are highlighted by the black square (d). e) Comparison between the results obtained from temperature simulations and experimental measurements confirming the agreement between the simulations and the experiments.

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)