> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 1

METAMATERIAL-INSPIRED WEARABLE PAD FOR

ENHANCING EM COUPLING WITH BIOLOGICAL
TISSUES
Maria Koutsoupidou*, Dimitrios C. Tzarouchis*, Member, IEEE, Dionysios Rompolas, Ioannis
Sotiriou, George Palikaras, and Panagiotis Kosmas, Senior Member, IEEE

Overview of the functionality of the proposed metamaterial-inspired wearable pad for enhancing the EM
coupling with biological tissues at 2.4 GHz.

Take-Home Messages

 This paper proposes a totally passive, wearable, thin pad, comprising metallic structures embedded in a
dielectric material for effectively couplingVisual Summary
EM radiation to a biological tissue.
 The improvement in transmission to a small, implantedx antenna
Max 4 inch 6 inch was measured to be more than 4 dB in the 2.4-
2.5 GHz range.
 The proposed pad can be used as an auxiliary tool in various biomedical applications, from hyperthermia to
implant technology for enhancing communication and/or wireless charging.
 A thin, flexible and mechanically robust pad was developed for clinical and every-day biomedical applications.
 METAMATERIAL-INSPIRED WEARABLE PAD FOR
ENHANCING EM COUPLING WITH BIOLOGICAL
TISSUES
Maria Koutsoupidou*, Dimitrios C. Tzarouchis*, Member, IEEE, Dionysios Rompolas,
Ioannis Sotiriou, George Palikaras, and Panagiotis Kosmas, Senior Member, IEEE

Abstract Wearable, implantable, and ingestible antennas are continuously evolving in biomedical applications, as they are crucial
components in devices used for monitoring and controlling physiological parameters. This work presents an experimentally
validated wearable pad which can improve transmission of electromagnetic waves into the human body. This metamaterial-inspired
matching pad, which is based on small metallic loops encased in a thin dielectric layer, is mechanically stable, flexible, and passive.
As such, the pad can serve as a coupling medium for microwave medical systems and implantable device communication.
Operating in the 2.4-2.5 GHz range, the pad demonstrates significant improvement in signal penetration levels (and, hence, depth)
into a biological tissue. The study presents design methodology, simulation studies, in-lab development, and experimental
characterization of this pad, which can offer a practical solution for enhanced communication and functionality in various medical
diagnostic systems.

Keywords —antenna measurements, dielectric materials, electromagnetic coupling, implantable biomedical devices, metamaterials,
phantoms

tissues. Two main challenges arise in EM signal propagation

I. INTRODUCTION1 between the human body and an external source or detector:

Β IOMEDICAL applications covering imaging, sensing, a) significant reflection at the air-skin interface, and b) rapid
and communication across RF to mm-waves necessitate signal attenuation within the tissue due to its high losses,
the transmission and reception of electromagnetic (EM) resulting in limited penetration depths relative to the
waves to and from human tissues. Typically, one or more application frequency. This limitation is particularly
antennas are employed in close proximity to or within the pronounced in the 2.40-2.50 GHz range (2.4 GHz ISM
human body. Recent progress in the field of biomedical band), which is the frequency band allocated for various
engineering includes the development of systems of communication protocols like WLAN and Bluetooth [8], [9].
wearable, implantable, and ingestible antennas for Signal attenuation within tissues presents an inherent
monitoring and/or controlling various physiological challenge, and potential solutions include shifting to lower
parameters [1], [2], as well as managing or treating frequencies, targeting shallow tissue layers, or increasing the
symptoms [3], [4], [5]. The seamless functioning of these overall radiated power. However, these options may not
devices, especially the long-term operation of implantable
ones within the human body, relies heavily on their capacity
to establish communication with an external base. This
communication is essential for tasks such as data sharing,
control, and wireless charging. Similarly, the efficiency of
other microwave diagnostic and therapeutic applications,
including imaging [6] and hyperthermia [7], depends on the
enhanced coupling of the EM radiation with the tissue and
the localization toward a specific site or target. The impact
of these applications on patients’ well-being, quality of life,
and healthcare systems is very significant.
The widespread clinical adoption of innovative EM
Fig. 1. Concept image of the proposed wearable pad for enhancing implant
diagnostic and therapeutic tools, particularly implantable technology and theranostic systems.
technology, requires effective coupling with biological

1
The paper was submitted for review on April 09, 2024. D. C. Tzarouchis is now with the Broadband systems & wireless
All authors were with Meta Materials Europe S.A., Ap, Pavlou 10a, Technologies Department, Intracom Telecom S.A., 19002, Peania, Greece.
Marousi, T.K. 15123, Athens, Greece. * These authors contributed equally to this work.
M. Koutsoupidou is now with Hellenic Military Academy, Evelpidon
Avenue,Vari, Greece. (e-mail: [email protected])
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 3

always be feasible for the intended application. Conversely,

technical interventions aim to mitigate reflections at the air-
skin interface through the use of high permittivity dielectrics
[10], [11], [12], [13] and metasurfaces [14], [15] as matching
mediums. High permittivity dielectrics are designed to align
the air impedance with skin impedance, but they typically
involve thick and bulky layers, rendering them possibly
suitable for clinical medical devices but not for everyday
applications like telemetric control or charging of body
Fig. 2. Side (left) and top (right) view of the design of the matching pad made
implants, which require matching layers that are comfortable of two parallel copper rings embedded in a thin silicon rubber disk.
and user-friendly. An alternative solution, explored by He et
al. [16], [17], utilizes metamaterial effective medium rings generate an effective magnetic dipole moment,
techniques. These methods aim to mitigate reflections facilitating the re-radiation of energy into the intricate and
through the design of an effective medium comprising lossy biological tissue. The rings were presumed to possess
subwavelength unit cells featuring metallic rings on a specific thicknesses as they are crafted from commonly used
dielectric substrate. However, these approaches entail copper sheets or copper tape. Our sole imposed constraint on
complex optimization and implementation processes these rings was to ensure they weren't positioned too close to
involving arrays of such minuscule unit cells. the surface of the pad and risk to touch the human skin. Their
Given these considerations, our paper proposes a placements were meticulously optimized to yield optimal
matching pad designed to conform to the human body above performance and effectiveness in the intended application.
the targeted area, significantly enhancing the external field's Both the metallic loops and the dielectric pad were inverse
impact on the body. The pad's metamaterial-inspired design designed: the resulted permittivity of the pad and the radii of
incorporates small parallel metallic loops within a thin the circular metallic rings were subject to optimization. The
dielectric of custom permittivity. Upon excitation by an design goal was to achieve maximization of the power
external field, these metal loops induce resonances at transmission from an external source to the tissue. The basic
frequencies which depend on their dimensions. Moreover, idea emerged from the contemporary trends of inverse design
the dielectric cover of these loops acts as a matching medium and metamaterial synthesis [18].
at the air-skin interface and contributes to the mechanical The setup was designed to operate at 2.45 GHz and to
stability of the setup. This design enables a pad which is exhibit a maximum thickness of 2 mm, to ensure that it is
mechanically stable, flexible, biocompatible, and entirely comfortable for users and thus useful in practical
passive. applications. The metal rings were designed to be made of
This pad is intended to serve as a coupling medium copper sheet of 0.12 mm thickness inserted into the dielectric
between incident EM radiation and human tissue for medical pad with dielectric permittivity, εr = 4, and loss tangent, tanδ
diagnostics or therapy systems requiring microwaves, as well = 0.02. The rings were positioned flexibly within the pad,
as for communication with implantable devices (Fig. 1). The ensuring they stayed at least 0.3 mm away from the outer
choice of the 2.4 to 2.5 GHz range for the operating edge.
bandwidth is specifically tailored to suit the latter With these design specifications and constraints, the pad
application. The upcoming sections will present the design was optimized to maximize the EM field inside the tissue
process and simulation outcomes across different setups beneath the pad in the 2.4-2.5 GHz range. The tissue was
(Section II), the in-lab development methodology for the pad modelled as a three-layer stack of: a) 5 mm thick skin with
(Section II), and its experimental characterization (Section dielectric permittivity εr_skin = 39.2 and conductivity σskin =
III). 1.8 S/m, b) 25 mm thick fat with εr_fat = 5.0 and σfat = 0.25
S/m and c) 30 mm muscle with εr muscle = 52.7 and σmuscle =
II. METHODS AND PROCEDURES 1.95 S/m. The EM field incident to the pad was linearly
A. Wearable pad design and specifications polarized, generated by a dipole with a resonance at 2.45
Our approach in designing the wearable matching pad GHz. Since the pad’s topology is symmetric, the orientation
prioritized simplicity, mechanical feasibility, and robustness. of the field does not affect its performance. The setup was
Instead of employing an array of metallic structures [14], modelled with CST Microwave Studio Suite® and optimized
[16], [17], we opted for a minimalist configuration consisting using the CST GA optimization tool. The pad’s dimensions
of only one set of rings embedded vertically within a medium were selected as, H1 = 1.16 mm, h2 = 0.3 mm, d1 = 15.5 mm,
with low permittivity and minimal losses. By doing so, we d2 = 25 mm, and d3 = 55 mm.
reduce the complexity of the problem, minimizing unknowns
and eliminating the need for homogenization techniques.
The proposed design features two parallel metal rings
encased within a thin dielectric circular pad (Fig. 2). The
underlying principle is that induced currents in the metallic
 B. Simulation results performance when conforming to curved body parts, such as
hands or legs. To access this, the pad was tested on a

Fig. 4. (a) Field distribution of the electric field excited by a dipole antenna
located at 130 mm above a three-layer tissue comprising (from top to bottom)
skin, fat, and muscle, with the pad lying on the tissue surface. (b) Power
density calculated at depths of 10 mm, 20 mm, and 30 mm below the tissue
surface with (solid line) and without (dashed line) the pad in the 2-3 GHz
range. (c) Improvement in dB of the power density with the pad for different
depths in the tissue at 2.45 GHz.

Fig. 5. (a) Transmission coefficient and (b) reflection coefficient of the

external antenna and the implantable antenna at depths 8 mm and 15 mm
from the tissue surface with (solid line) and without the pad (dashed line) on
the tissue surface.

cylindrical setup of 120 mm diameter comprising the

previous setup layers (Fig. 3a). The simulation results
depicting the electric field distribution are presented in Fig.
3b. Bending the pad induces a slight frequency shift in its
resonance from 2.46 GHz to 2.42 GHz (Fig. 3c).
Nevertheless, for both configurations, there is a noticeable
enhancement across the entire 2.4 GHz ISM range. When
bent, the improvement ranges from 8 dB at 5 mm within the
tissue to 4 dB at 40 mm (Fig. 3d). In both cases, strong
attenuation is noticeable at a depth of 30 mm (fat-muscle
interface) due to the high conductivity of the muscle,
resulting in strong signal attenuation. However, the
Fig. 3. (a) Simulation setup where the matching pad conforms to tissue of degradation occurs more gradually as the signal continues to
cylindrical shape of 120 mm diameter that comprises skin, fat, muscle. (b) propagate deeper into the muscle layer.
Field distribution of the electric field excited by a dipole antenna 130 mm These results clearly demonstrate the enhanced power
above the cylindrical tissue and the conforming matching pad. (c) Power
density calculated at depths of 10 mm, 20 mm, and 30 mm below the tissue penetration into the tissue when utilizing the proposed pad.
surface with (solid line) and without (dashed line) the pad in the 2-3 GHz In addition to this, and to consider various microwave
range. (d) Improvement in dB of the power density with the pad for different applications which involve implantable antennas within the
depths in the tissue at 2.45 GHz.
tissue, (e.g., telemetry and wireless charging), we conducted
tests on the transmission between an external antenna and a
The electric and magnetic fields were monitored with
dipole implanted in the fat layer. To achieve this
probes at various tissue depths to calculate power density
measurement, the previous flat-tissue model was updated to
with and without the pad (Fig. 4a). Fig. 4b illustrates the
incorporate a small dipole antenna operating at 2.45 GHz,
electric field distribution inside the tissue at 2.45 GHz
positioned at depths of 8 mm and 15 mm within the fat layer.
(central frequency) when the pad is employed. The power
The magnitude of the transmission coefficients, |S21|, for
density near the pad (5 mm deep in the tissue) shows a 9 dB
both implant positions, with and without the pad, are
improvement, extending to 4 dB at a depth of 40 mm within
depicted in Fig. 5a, while the corresponding reflection
the tissue (Fig. 4c, d). Importantly, the polarization of the
coefficients (|S11|, |S22|) of the external and implanted
field inside the tissue remains unaffected by the presence of
dipole are presented in Fig. 5b. The transmission
the pad.
Additionally, the proposed pad should maintain its
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 5

enhancement at 2.45 GHz is 5.4 dB and 4.8 dB for positions

at 8 mm and 15 mm, respectively. As anticipated and is
evident from the reflection coefficient results, the pad
influences the antenna’s performance when the implant is
closer to the tissue’s surface. Conversely, the external
antenna is positioned far enough to avoid impacting its input
resistance.
C. Wearable pad development
The pad is composed of copper rings embedded in a
dielectric material, for which we opted for a blend of silicon
rubber with calcium copper titanite (CCTO). Silicon rubber
was chosen for its mechanical strength and flexibility [19],
while CCTO powder was introduced to further control the
dielectric permittivity and reduce the loss tangent of the
mixture. The copper rings, crafted from copper sheet cut with
a laser cutter, were intentionally designed to be fully
immersed in the pad, ensuring no contact with the user’s
skin.
For the mixture, the silicon rubber base and hardener
(Zhermack GmbH, Germany) and CCTO powder (Lori
Industry Co., Ltd., China) were blended in a ratio of
100:3:200 by mass. In its semi-solid state, the mixture Fig. 6. (a) Transmission coefficient and (b) reflection coefficient of the
external antenna and the implantable antenna at depths 8 mm and 15 mm
underwent degassing in a vacuum chamber to eliminate air from the tissue surface with (solid line) and without the pad (dashed line) on
bubbles. Subsequently, its dielectric properties were the tissue surface.
measured within the 2-3 GHz range using the open-ended
coaxial probe DAK 12 by SPEAG® (Fig. 6a). The semi-solid
mixture was then poured into a 3D printed mold and left to
cure. The resulting pad exhibits excellent mechanical
stability and flexibility, enabling it to conform to various
body parts. Its overall dimensions are 2.13 mm in thickness
and 54.6 mm in diameter (Fig. 6b). Moreover, it is
lightweight, weighing only 8 g.
D. Experimental setup
To test the developed pad and validate the simulation
results, we constructed the final simulation setup,
incorporating a dipole antenna positioned within the fat layer
of a flat phantom. We assembled a box with acrylic walls and
3D printed components (Fig. 7a), including sockets of
varying heights designed to accommodate the implanted
dipole. Specifically, two sockets were made: one for placing
the antenna at a depth of 8 mm below the surface and another
at 15 mm (Fig. 7b). Fig. 7. (a) Transmission coefficient and (b) reflection coefficient of the
external antenna and the implantable antenna at depths 8 mm and 15 mm
The muscle, fat and skin gel phantoms were made with the from the tissue surface with (solid line) and without the pad (dashed line) on
following process. For the muscle and skin phantoms, DI the tissue surface.
water, gelatin and glycerol were used. For the skin, gelatin TABLE I
and water were mixed in a mass ratio 1/20 at 75 oC until the DIELECTRIC PROPERTIES OF THE TISSUE MODELS AND PHANTOMS
mixture was homogenous, and subsequently, glycerol was Simulation Experimental Experimental
gradually added in the mixture until the desired dielectric 2.45 model setup 1 (antenna setup 1 (antenna
GHz at 8 mm) at 15 mm)
properties were reached by continuously measuring its
dielectric properties. Finally, the mixture was molded and εr σ (S/m) εr σ (S/m) εr σ (S/m)
placed in the fridge to solidify for a few hours. Similarly, DI skin 39.2 1.8 39.7 3.9 38.8 3.1
water, gelatin, safflower oil and wheat flour were used to
fat 5 0.25 5.7 0.16 5.2 0.14
fabricate the fat phantom. Gelatin and water were again
initially mixed, and then flour was added in the mixture at a muscle 52.7 1.95 52.7 1.85 53.4 2.35
gelatin/flour mass ratio equal to 1/5. Finally, safflower oil
Fig. 8. (a) Magnitude of transmission coefficient, |S21|, and (b) Magnitude the reflection coefficients of the external and the immersed dipole, |S11|, |S22|, with
(solid line) and without (dashed line) the pad when the immersed dipole lied 8 mm below the phantom surface. (c) Magnitude of transmission coefficient, |S21|,
and (d) Magnitude the reflection coefficients of the external and the immersed dipole, |S11|, |S22|, with (solid line) and without (dashed line) the pad when the
immersed dipole lied 15 mm below the phantom surface.
TABLE II
was gradually added until the desired dielectric properties TRANSMISSION IMPROVEMENT WHEN THE PAD IS USED AT SELECTED
were reached, again, by continuously measuring mixture FREQUENCIES

properties. The mixture was poured in the mold and placed Depth of implanted 2.4 GHz 2.45 GHz 2.5 GHz
in the fridge. antenna
The dielectric properties of the developed phantoms 8 mm 1.8 dB 4.1 dB 5.4 dB
measured at 2.45 GHz with the DAK12 probe (SPEAG®)
15 mm 2.0 dB 3.9 dB 4.4 dB
and the results vs the model values are shown in Table I.
Initially, the socket for 8 mm was used and each gel mixture
was poured following solidification of the previous one. The been significantly detuned (Fig. 8b). Additionally, the
final experimental setup is shown in Fig. 7c: the metallic transmission enhancement because of the pad is higher than
loop-based pad lies on the three-layer phantom and the 2 dB from 2.4 GHz to 2.6 GHz, peaking at 2.5 GHz. We get
external antenna is attached to the top cover of the box, to similar results for the 15 mm depth, where the pad improved
ensure that it is stable and located at specific, controlled the |S21| at 2.45 GHz by 3.9 dB (Fig. 8c). The immersed
distances during the experiment. The antennas’ S-parameters dipole is more stable at this depth because of its increased
with and without the pad were measured using an distance from the pad’s inductive loops. Generally, the
R&S®ZNBT VNA. To replicate the experiment with the transmission enhancement matches well the simulation
antenna positioned inside the 15 mm socket within the tissue, results of Section II.B. A summary of the measured
new phantoms were fabricated after destroying the previous improvement in S21 scattering at specific frequencies in the
ones, and the experiment was conducted again. 2.4-2.5 GHz frequency range is shown in Table II.

III. EXPERIMENTAL RESULTS IV. DISCUSSION AND CONCLUSIONS

The experimental results for the S-parameters of the Our findings indicate that employing the proposed metal-
dipole antennas in the 2-3 GHz frequency range for both loop-based pad enhances the penetration depth of incident
depths, 8 mm and 15 mm, are depicted in Fig. 8. When the EM radiation through a combination of dielectric matching
immersed dipole lies at 8 mm from the phantom surface, the mechanisms and the excitation of a resonance near the tissue.
magnitude of the transmission coefficient, |S21|, is enhanced We note that the observed improvement of 4-5 dB in
by 4.1 dB at 2.45 GHz (Fig. 8a), although this dipole had transmission to an implant holds exceptional significance, as
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 7

this difference can often make a signal connection feasible [9] M. J. Christoe, J. Yuan, A. Michael, and K. Kalantar-Zadeh,
“Bluetooth signal attenuation analysis in human body tissue
and reduce the power consumption of active implantable analogues,” IEEE Access, vol. 9, pp. 85144–85150, 2021.
devices. In implant technology, overcoming connection gaps [10] M. Koutsoupidou, N. Uzunoglu, and I. S. Karanasiou, “Antennas on
is a substantial challenge, and this enhancement addresses metamaterial substrates as emitting components for THz biomedical
imaging,” in 2012 IEEE 12th International Conference on
that issue effectively. Furthermore, the chosen materials and
Bioinformatics & Bioengineering (BIBE), IEEE, 2012, pp. 319–322.
design at 2.4 GHz ensures comfortable, safe, and user- [11] G. G. Bellizzi, K. Sumser, and M. T. Bevacqua, “On the optimal
friendly application to users in everyday life. The pad's small matching medium and the working frequency in deep pelvic
size, light weight, and ability to be attached to any body part hyperthermia,” IEEE J. Electromagn. RF Microw. Med. Biol., vol.
5, no. 3, pp. 223–230, 2020.
using an available adhesion mechanism contribute to its [12] C. Rappaport and others, “Determination of bolus dielectric constant
practicality. Additionally, a simulation-based sensitivity for optimum coupling of microwaves through skin for breast cancer
analysis was conducted to assess the system's robustness. imaging,” Int. J. Antennas Propag., vol. 2008, 2008.
[13] P. M. Meaney, C. J. Fox, S. D. Geimer, and K. D. Paulsen,
The ±3σ analysis revealed that, with a 10% standard “Electrical characterization of glycerin: Water mixtures:
deviation (σ = 10%) in the electromagnetic properties Implications for use as a coupling medium in microwave
(permittivity and loss tangent) of both the tissues and the pad, tomography,” IEEE Trans. Microw. Theory Tech., vol. 65, no. 5, pp.
1471–1478, 2017.
there is a ±1.5 dB variation in the calculated transmission. [14] E. Razzicchia, I. Sotiriou, H. Cano-Garcia, E. Kallos, G. Palikaras,
Thus, the system demonstrates robustness against and P. Kosmas, “Feasibility study of enhancing microwave brain
perturbations and uncertainties. imaging using metamaterials,” Sensors, vol. 19, no. 24, Art. no. 24,
2019.
Based on these findings, we believe that the proposed pad [15] D. C. Tzarouchis, M. Koutsoupidou, I. Sotiriou, K. Dovelos, D.
can tackle critical challenges in the realm of biomedical Rompolas, and P. Kosmas, “Electromagnetic Metamaterials for
applications involving the transmission and reception of EM Biomedical Applications: Short Review and Trends,” EPJ Appl.
Metamaterials, vol. (accepted), 2024.
waves within the RF to mm-wave spectrum. Aware of the
[16] Y. He and G. V. Eleftheriades, “A thin double-mesh metamaterial
limitations posed by signal attenuation and reflections at the radome for wide-angle and broadband applications at millimeter-
air-skin interface, we have developed an innovative solution wave frequencies,” IEEE Trans. Antennas Propag., vol. 68, no. 3,
in the form of a metamaterial-inspired matching pad. The pp. 2176–2185, 2019.
[17] Y. He and G. V. Eleftheriades, “Magnetoelectric uniaxial
next steps in this research involve scaling the proposed metamaterials as wide-angle polarization-insensitive matching
design to higher frequency ranges, such as 3.6 GHz (mid- layers,” Phys. Rev. B, vol. 98, no. 20, p. 205404, 2018.
band 5G), and incorporating multispectral characteristics. [18] V. Nikkhah, D. C. Tzarouchis, A. Hoorfar, and N. Engheta,
“Inverse-designed metastructures together with reconfigurable
This expansion aims to enable simultaneous operation across couplers to compute forward scattering,” ACS Photonics, vol. 10,
various frequency ranges and communication protocols. no. 4, pp. 977–985, 2022.
Overall, we believe that this work can contribute to the [19] M. Zare, E. R. Ghomi, P. D. Venkatraman, and S. Ramakrishna,
“Silicone-based biomaterials for biomedical applications:
advancement of microwave medical diagnostics and therapy antimicrobial strategies and 3D printing technologies,” J. Appl.
systems, offering promising potential for both clinical and Polym. Sci., vol. 138, no. 38, p. 50969, 2021.
everyday use by patients.

REFERENCES
[1] J. Duun-Henriksen et al., “A new era in electroencephalographic
monitoring? Subscalp devices for ultra–long-term recordings,”
Epilepsia, vol. 61, no. 9, pp. 1805–1817, 2020.
[2] S. Simovic et al., “The use of remote monitoring of cardiac
implantable devices during the COVID-19 pandemic: an EHRA
physician survey,” EP Eur., vol. 24, no. 3, pp. 473–480, 2022.
[3] A. M. Dydyk and P. Tadi, “Spinal cord stimulator implant,” 2020.
[4] K. P. Iyengar, B. T. V. Gowers, V. K. Jain, R. S. Ahluwalia, R.
Botchu, and R. Vaishya, “Smart sensor implant technology in total
knee arthroplasty,” J. Clin. Orthop. Trauma, vol. 22, p. 101605,
2021.
[5] K. Zhang, G. A. Vandenbosch, and S. Yan, “A novel design
approach for compact wearable antennas based on metasurfaces,”
IEEE Trans. Biomed. Circuits Syst., vol. 14, no. 4, Art. no. 4, 2020.
[6] N. Ghavami, I. Sotiriou, and P. Kosmas, “Preliminary experimental
validation of radar imaging for stroke detection with phantoms,” in
2019 Photonics & Electromagnetics Research Symposium-Fall
(PIERS-Fall), IEEE, 2019, pp. 1916–1923.
[7] M. Paulides, H. D. Trefna, S. Curto, and D. Rodrigues, “Recent
technological advancements in radiofrequency-andmicrowave-
mediated hyperthermia for enhancing drug delivery,” Adv. Drug
Deliv. Rev., vol. 163, pp. 3–18, 2020.
[8] M. Fernandez, H. G. Espinosa, D. Guerra, I. Pena, D. V. Thiel, and
A. Arrinda, “RF energy absorption in human bodies due to wearable
antennas in the 2.4 GHz frequency band,” Bioelectromagnetics, vol.
41, no. 1, pp. 73–79, 2020.