An investigation on power loss of an out-to-in body wireless radio frequency link

Abstract

BACKGROUND:

Implantable medical sensors for monitoring and transmitting physiological signals like blood glucose, blood oxygen, electrocardiogram, and endoscopic video present a new way for health care and disease prevention. Nevertheless, the signals transmitted by implantable sensors undergo significant attenuation as they propagate through various biological tissue layers.

OBJECTIVE:

This paper mainly aims to investigate the power loss of an out-to-in body wireless radio frequency link at 2.45 GHz.

METHODS:

Two simulation models including the single-layer human tissue model and three-layer human tissue model were established, applying the finite element method (FEM). Two experiments using physiological saline and excised porcine tissue were conducted to measure the power loss of a wireless radio frequency link at 2.45 GHz. Various communication distances and implantation depths were investigated in our study.

RESULTS:

The results from our measurements show that each 2 cm increase in implantation depth will result in an additional power loss of about 10 dB. The largest difference in values obtained from the measurements and the simulations is within 4 dB, which indicates that the experiments are in good agreement with the simulations.

CONCLUSIONS:

These results are significant for the estimate of how electromagnetic energy changes after propagating through human tissues, which can be used as a reference for the link budget of transceivers or other implantable medical devices.

Keywords

Implantable medical sensors power loss radio frequency link finite element method

1. Introduction

It was reported by the World Health Organization (WHO) that the burden of chronic diseases is increasing at a sharp rate in the world. Statistics from WHO show that chronic diseases contributed about 60% of the 56.5 million total reported deaths worldwide in 2001. It has been predicted that nearly 75% of global deaths will be attributed to chronic diseases by 2020, such as hypertension, diabetes and cardiovascular disease [1]. It is essential to monitor the status of the patient’s health in real time, so that doctors can provide effective health care and disease prevention before the condition worsen. IEEE 802.15.6 is an efficient wireless body area network (WBAN) standard for connecting small devices deployed on the surface, inside or in the peripheral proximity of the human body. Combined with various biomedical sensors, monitoring and transmitting of various physiological signals in real time will become possible with this technology [2]. Physiological signals like blood glucose, blood oxygen, endoscopic video, electrocardiogram (ECG), and electroencephalograph (EEG) can provide useful information for doctors to diagnose and treat patients [3]. Currently, different implantable medical devices (IMD) such as implantable insulin pumps, cardiac defibrillators, pacemakers, and artificial cochlear have been widely applied in the human body [4].

To ensure optimal performance of such implantable devices, physiological signals should be transmitted outside the human body in an efficient and reliable manner. Microwave antenna and techniques have been widely used for monitoring and transmitting various physiological signals [5, 6]. Nevertheless, due to the dielectric properties of human tissues, the signals transmitted by implantable devices will sustain large power losses as they propagate through various biological tissue layers. To realize efficient performance of a wireless communication system, it is greatly important to study the attenuation characteristics of the radio channel. Software simulation provides a useful technique to estimate the power loss of a wireless communication link [7]. The power loss of a wireless communication link between an endoscopy capsule and a receiver outside the human body was investigated in [8]. All the work was performed in medical implant communication service (MICS) band and finished by a 3D electromagnetic solver, applying the finite difference time domain (FDTD) method. Similar investigation was presented in [9, 10, 11], but a path loss model of ultra-wideband (UWB) was investigated in the latter. A lot of work on channel attenuation characteristics of implantable human body communication have been done by many scholars, but most of them were focused on model simulation and mathematical calculation. To the best of our knowledge, very little attention was given to experimental measurement to verify the correctness and validity of the theory.

This paper mainly aims at investigating the power loss of an out-to-in body wireless radio frequency link by means of software simulation and experimental measurements. For this purpose, two simulation models including single-layer tissue model and three-layer tissue model were established, by applying the finite element method (FEM). Additionally, two experiments using saline solution and excised porcine tissue were conducted to validate the results of our simulations. All the investigations are performed in industrial, scientific, medical (ISM) bands at 2.45 GHz, which have the advantages of high data rate and low power consumption. Various communication distances from 10 cm to 70 cm and implantation depths from 1 cm to 6 cm are discussed in this paper.

The remaining sections of this paper are organized as follows: Section 2 mainly describes the finite element simulation models of human body tissues and the distribution of the power density in human body tissues. Two experimental measurements are conducted in Section 3. In Section 4, we analyze and discuss the results of the simulations and the experiments. Finally, we draw our conclusions in Section 5.

2. Simulation

2.1 The dielectric properties of human tissues

The dielectric properties of human body tissues, the conductivity and the relative permittivity, are related to the frequency of electric field, which can be represented by the complex relative permittivity. The four order Cole-Cole model provided by Gabriel et al. [12] describes both the conductivity and the relative permittivity of human body tissues in the frequency range from 10 Hz to 100 GHz. The model is given as Eq. (1), where $\varepsilon^{*}(\omega)$ is the complex relative permittivity and $\omega$ is the angular frequency; $\varepsilon^{\prime}(\omega)$ and $\varepsilon^{\prime\prime}(\omega)$ are the real and imaginary parts of the complex relative permittivity, respectively; $\varepsilon_{0}$ is the relative permittivity of vacuum and $\varepsilon_{\infty}$ is the relative permittivity at infinite frequency; $\Delta\varepsilon_{n}$ , $\tau_{n}$ , $\alpha_{n}$ , and $\sigma_{i}$ are the other parameters that can be found in [12]. The dielectric properties of different human tissues at 2.45 GHz provided by the model are listed in Table 1.

$\displaystyle\varepsilon^{*}(\omega)=\varepsilon^{\prime}(\omega)-j\varepsilon% ^{\prime\prime}(\omega)=\varepsilon_{\infty}+\sum_{n=1}^{4}\frac{\Delta% \varepsilon_{n}}{1+(j\omega\tau_{n})^{(1-\alpha_{n})}}+\frac{\sigma_{i}}{j% \omega\varepsilon_{0}}$ (1)

Table 1

The dielectric properties of different human tissues at 2.45 GHz

Tissue name	Conductivity (S/m)	Relative permittivity
Air	0	1
Dry skin	1.4640	38.0070
Wet skin	1.5919	42.8530
Fat	0.1045	5.2801
Muscle	1.7388	52.7290

Figure 1.

The simulation model of single-layer human tissue.

2.2 Simulation models

Our electromagnetic simulations are conducted using the finite element method (FEM) with the COMSOL Multiphysics 5.2 simulation tool [13]. Two models are established in our simulations. The single-layer model is used to simulate the human muscle layer, as illustrated in Fig. 1. The three-layer model includes skin, fat, and muscle layer, as illustrated in Fig. 2. The thickness of each tissue layer corresponds to the porcine tissue of the following ex vivo experiment. The models are exposed to an incoming plane electromagnetic wave, which can approximate the far-field of a communication transmitter. The electric field intensity of the plane electromagnetic wave is set to 1 V/m, and the direction is normal incidence. The leftmost boundary of the model is defined as scattering boundary condition, and the remaining boundary is defined as perfect electric conductor. In addition, the perfect matched layer is used for the simulation model, thus the human body environment reflections can be disregarded. The materials of each region are represented with the conductivity and the relative permittivity according to the four order Cole-Cole model. After meshing, the distribution of different electromagnetic parameters such as the electric field intensity, the magnetic field intensity, the current density, and the power density in human tissues can be obtained by the solver.

Figure 2.

The simulation model of three-layer human tissue.

2.3 The distribution of the power density in human tissues

Due to the dielectric properties of human tissues, the human body is not an ideal medium for electromagnetic wave propagating. Depending on the frequency of electric field, the human body will lead to large power losses caused by energy absorption, radiation pattern destruction, and central frequency shift [14, 15]. When the frequency of electric field is high, the signal strength received at the receiver is quite important since the transmitted power is usually small. The power loss is an important indicator to measure the communication quality from transmitter to receiver, which can be obtained from the distribution of the power density. The average power density in human tissues is usually expressed by the average Poynting vector, which describes both the direction and the power density of the electromagnetic wave propagation. The average Poynting vector is given by Eq. (2), in which $E$ (V/m) is the electric field intensity and $H^{*}$ (A/m) is the conjugation of the magnetic field intensity, Re represents the real part of the plural. By solving the simulation models, the distribution of the power density in human tissues at 2.45 GHz can be obtained, as illustrated in Figs 3 and 4.

$\displaystyle P_{\textit{av}}=\frac{1}{2}\text{Re}\left[E\times H^{*}\right]% \text{W/m}^{2}$ (2)

3. Experiment

Experiments on real human bodies are the final tests to verify the correctness and validity of the theory. Due to a series of issues such as humanitarian and surgical techniques, it is not realistic to perform measurement campaigns directly inside living human bodies. Testing using physiological saline or excised animal tissue (ex vivo) before engaging in living human bodies (in vivo) can effectively reduce the risk from surgery [16]. Data measured from these tests can provide a useful reference for in vivo experiments and an estimate for the link budget of transceivers or implantable medical devices. In this paper, two test models, physiological saline and porcine tissue, are used to measure the power loss from the transmitting antenna to the receiving antenna.

Table 2
The materials and mass percentage of the solution

Materials	Distilled water	NaCl	DGBE
Mass percentage	73.2%	0.1%	26.7%

Figure 3.

The distribution of the power density in single-layer tissue model.

Figure 4.

The distribution of the power density in three-layer tissue model.

3.1 Physiological saline test

For simulating the dielectric properties in the human body environment, we first prepared a 500 mL solution containing distilled water, sodium chloride (NaCl) and diethylene glycol butyl ether (DGBE). The conductivity and the relative permittivity of the solution are about 1.78 S/m and 52.5, respectively, which are very close to the dielectric parameters of the human muscle tissue. The materials and mass percentage of the solution are listed in Table 2 [17].

The measurement setup for our experiments consists of two dipole antennas connected to an Agilent Technology E5061B vector network analyzer (VNA), which is used to measure the S-parameters. The antenna has the characteristics of small size (3 $\times$ 4 cm), high gain (4-dBi) and high efficiency in transmitting and receiving signals. Before experiments, the antennas were covered with a protective film, which can prevent the components or the printed circuit board from short-circuiting. By measuring the power received in free space with and without a protective film or a plastic solution container, we drew a conclusion that the protective film and the plastic container did not affect the results of the measurement. After all the preparation work, we first immersed the receiving antenna in the solution and placed 2 cm behind the surface of the container. Then we positioned the transmitting antenna at multiple distances (10, 20, 30, …, 70 cm) in front of the container and the corresponding values of the power loss were measured. The measurement campaigns were repeated as the implantation depths were set to 4 cm and 6 cm, respectively. Additionally, to prevent the human body from interfering with the experiment, we developed a measurement tool to fix the antenna, as illustrated in Fig. 5.

Figure 5.

The measurement setup for transmission in the saline solution.

Figure 6.

The measurement setup for transmission in the porcine tissue.

Figure 7.

The power received in free space versus implants (saline solution and porcine tissue).

Figure 8.

The power loss in simulations versus experiments at various implantation depths from 1 cm to 6 cm.

3.2 Porcine tissue test

According to [18, 19], porcine tissues have great similarities to human tissues in terms of anatomical properties. To obtain more accurate experimental data, we performed an ex vivo experiment by using a piece of porcine tissue. The experiment setup of the porcine tissue test is the same as the physiological saline test, as illustrated in Fig. 6. Firstly, we placed a piece of excised porcine tissue (20-cm-long, 13-cm-wide, 7-cm-thick) behind the receiving antenna and positioned the transmitting antenna at multiple distances (10, 20, 30, …, 70 cm) in front of the receiver, and the corresponding values of the power received in free space were measured. After that, the receiving antenna was embedded in the porcine tissue at about 2 cm depth from the skin. The transmitting antenna was positioned at multiple distances (10, 20, 30, …, 70 cm) in front of the tissue, and the corresponding values of the power received in porcine tissue were measured. These measurement campaigns were repeated with implantation depths of 4 cm and 6 cm. The difference in values of the received power between porcine tissue and free space represents the power loss through the tissue.

Besides, the transmission of in-body to out-body was also considered in our experiments. In our measurement setup, we only exchanged the position of the transmitting antenna and the receiving antenna, the communication distances and implantation depths were kept the same.

All the measurement campaigns were performed in an electromagnetic shielded room, which effectively prevents interference from external electromagnetic or electronic devices. In addition, a large sheet of RF absorber used in our experiments substantially reduced radiated noise and internal electromagnetic interference.

In both physiological saline test and porcine tissue test, and for each measurement point, the measurements were repeated 10–14 times over 3–4 minutes to reveal and account for any significant change in the propagation medium over time.

Figure 9.

Out-body to in-body propagation versus in-body to out-body propagation.

4. Results and analysis

Various communication distances from 10 cm to 70 cm and implantation depths from 1 cm to 6 cm are investigated in this paper. In our measurement setup, the transmitting power is set to 0 dBm, and the average received power of different cases (free space, saline solution, and porcine tissue) is assessed, as shown in Fig. 7. Note that the additional 4 dB gains from the antennas are subtracted from the recorded values, which are the average of multiple measurements. The results from our measurements show that each 2 cm increase in implantation depth results in an additional power loss of about 10 dB. Additionally, the average power received both in free space and implants decreases by about 14 dBm as the communication distance increases from 10 cm to 70 cm. The differences of the received power between physiological saline test and porcine tissue test can be attributed to the fact that the solution model lacks a full set of biological tissue layers.

In addition, to obtain the power loss of electromagnetic waves propagating through various biological tissue layers, we compare the measurements in free space with those done in the implants. Given a fixed communication distance, we subtract the power received in free space from that in the implants, and the difference in values between two measurements represents the power loss through the implants. The result is given by Eq. (3), where $PL_{e}$ represents the power loss, $P_{i}$ and $P_{f}$ are the power received in the implants and free space, respectively. For example, with the transmitting antenna positioned at 50 cm away from the porcine tissue, the power loss at various implantation depths from 1 cm to 6 cm in the tissue is plotted in Fig. 8. The measurement results of saline solution are also plotted in Fig. 8. The results from our studies show that electromagnetic waves propagating through a piece of 6-cm-thick tissue will attenuate by about 28 dB.

$\displaystyle PL_{e}=P_{i}-P_{f}$ (3)

The simulation results of the power loss are also shown in the Fig. 8. The power density in human tissues at various implantation depths are obtained by the finite element simulations. We choose the skin surface as the reference plane and the power loss at various implantation depths away from the skin can be calculated by Eq. (4), where $P_{r}$ (9.36 $\times$ 10 ${}^{-4}$ W/m ${}^{2}$ in our simulations) is the power density at the reference plane and $P_{d}$ is the power density at a certain depth in the model. Compared with the values we have measured in the implants, the largest difference is within 4 dB, which presents a good agreement with the experimental results. The slight difference in power loss obtained from the experiments and the simulations can be attributed to differences in dielectric properties and loss factors.

$\displaystyle PL_{s}=10\log\frac{P_{d}}{P_{r}}$ (4)

In the end of our studies, we compare the measurement result of out-body to in-body propagation with that of in-body to out-body propagation. The communication distance is set to 50 cm, and the received power of the two measurements at various implantation depths from 1 cm to 6 cm are plotted in Fig. 9. The largest difference of the two measurements is within 2 dBm, which indicates that the direction of propagation of a wave (out-body to in-body and in-body to out-body) does not affect significantly the efficiency of receiving signals.

5. Conclusion

This paper mainly presents the investigations on the power loss of an out-to-in body wireless radio frequency link at 2.45 GHz by means of software simulation and experimental measurements. The finite element simulations incorporated with single-layer tissue model and three-layer tissue model were established to provide an estimate of the channel characteristics. Besides, two experiments using physiological saline and excised porcine tissue were conducted to validate the results of our simulations. The ex vivo experiments provide a rational basis before engaging living human bodies (in vivo), which can effectively reduce the risk from surgery.

The power loss of the transmission is quantified for various implantation depths inside the human body. The results from our measurements show that each 2 cm increase in implantation depth will result in an additional power loss of about 10 dB. The largest spread in the values obtained from the measurements and the simulations is within 4 dB, which indicates that the experiments are in good agreement with the simulations. Finally, we also make a comparison between out-body to in-body propagation and in-body to out-body propagation, and the result indicates that the direction of propagation of a wave (out-body to in-body and in-body to out-body) does not affect significantly the efficiency of RF communication. These investigations are significant for the estimate of how electromagnetic energy propagates through human tissues, which is essential for the link budget of transceivers or other implantable medical devices.

Footnotes

Acknowledgments

This work was supported by the National Natural Science Foundation of China (grant nos. U1505251 and 61201397), the Project of Chinese Ministry of Science and Technology (grant no. 2016YFE0122700), and the S&T Project of Fujian Province (grant no. 2018I0011).

Conflict of interest

None to report.

References

World Health Organization (homepage on the Internet). Background: The global burden of chronic (2015, Jan 19). Available: http://www.who.int/nutrition/topics/2_background/en/.

Ieee B. Element. IEEE standard for local and metropolitan area networks – Part 156: Wireless body area networks. IEEE Std 2012; 1-271.

Chavez-Santiago

Sayrafian-Pour

Khaleghi

Takizawa

. Propagation models for IEEE 80215.6 standardization of implant communication in body area networks. Communications Magazine IEEE 2013; 51(8): 80-87.

Islam

Yuce

. Review of Medical Implant Communication System (MICS) band and network. Ict Express 2016; 2(4).

Rosen

Stuchly

Vorst

. Applications of RF/microwaves in medicine. IEEE Transactions on Microwave Theory and Techniques 2002; 50(3): 963-974.

Soontornpipit

Furse

You

. Design of implantable microstrip antenna for communication with medical implants. IEEE Transactions on Microwave Theory and Techniques 2004; 52(8): 1944-1951.

Bose

Khaleghi

Albatat

Bergsland

Balasingham

. RF channel modeling for implant to implant communication and implant to sub-cutaneous implant communication for future leadless cardiac pacemakers. IEEE Transactions on Biomedical Engineering 2018; PP(99): 1-1.

Roman

Vermeeren

Thielens

Joseph

Martens

. Characterization of path loss and absorption for a wireless radio frequency link between an in-body endoscopy capsule and a receiver outside the body. Eurasip Journal on Wireless Communications and Networking 2014; 2014(1): 21.

Khaleghi

Balasingham

Chavez-Santiago

. Computational study of ultra-wideband wave propagation into the human chest. Iet Microwaves Antennas and Propagation 2011; 5(5): 559-567.

10.

Khaleghi

Vez-Santiago

Liang

Balasingham

Leung

VCM

. On ultra-wideband channel modeling for in-body communications. 2010 5th IEEE International Symposium on Wireless Pervasive Computing, IEEE, 2010, 140-145.

11.

Khaleghi

Chavez-Santiago

Balasingham

. Ultra-wideband statistical propagation channel model for implant sensors in the human chest. Iet Microwaves Antennas and Propagation 2011; 5(15): 1805-1812.

12.

Gabriel

Lau

Gabriel

. The dielectric properties of biological tissues: III. Parametric models for the dielectric spectrum of tissues. Physics in Medicine and Biology 1996; 41(11): 2271-2293.

13.

Chen

Gao

. A study of galvanic-coupled intra-body communication based on circuit-coupled finite element model. 2015 IEEE International Conference on Consumer Electronics, Shenzhen, Apr 2015.

14.

Yazdandoost

Kohno

. UWB antenna for wireless body area network. Asia-Pacific Microwave Conference IEEE, 2006; 1647-1652.

15.

Yazdandoost

Kohno

. Wireless communications for body implanted medical device. Asia-Pacific Microwave Conference IEEE, 2007; 1-4.

16.

Chow

Morris

Irazoqui

. Implantable RF medical devices: The benefits of high-speed communication and much greater communication distances in biomedical applications. IEEE Microwave Magazine 2013; 14(4): 64-73.

17.

OET Bulletin 65, Edition 97-01, FCC Std Supplement C Jun 2001.

18.

Vodicka

Smetana

Dvoránková

Emerick

Ourednik

Motlík

. The miniature pig as an animal model in biomedical research. New York Acad Sci 2010; 1049(1): 161-171.

19.

Dale Bass

Rafaels

Salzar

Carboni

Kent

Lloyd

Lucas

Meyerhoff

Planchak

Damon

Bass

. Thoracic and lumbar spinal impact tolerance. Accident Anal and Prevention 2008; 40(2): 487-495.

An investigation on power loss of an out-to-in body wireless radio frequency link

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

1. Introduction

2. Simulation

2.1 The dielectric properties of human tissues

Table 2 The materials and mass percentage of the solution

Footnotes

Acknowledgments

Conflict of interest

References

Table 2
The materials and mass percentage of the solution