1999.01. Final Report Biological Effects in The MM Wave Range

Final Report
Biological Effects in the cm/mm Wave Range Part II/III Determination of Material Parameters and Analysis of Field Strengths in Human Tissue
by Institute of Mobile and Satellite Communication Techniques GmbH, Germany Dr.-Ing. Frank Gustrau Dr.-Ing. Achim Bahr
14. January 1999
Biological Effects in the cm/mm Wave Range, Part II/III
Table of Contents
1 2 SUBJECT OF INVESTIGATION ........................................................................................................ 4 RADIOFREQUENCY FIELD EXPOSURE STANDARDS................................................................. 4 2.1 2.2 2.3 2.4 2.5 3 LEGAL CLASSIFICATION OF STANDARDS ................................................................................................... 4 STANDARDS ASSOCIATIONS ..................................................................................................................... 5 DISTINCTION BETWEEN AREAS, TIME OF EXPOSURE AND FREQUENCIES ...................................................... 5 BASIC RESTRICTIONS AND DERIVED REFERENCE LEVELS............................................................................ 6 MAXIMUM PERMISSIBLE EXPOSURE FOR HIGH FREQUENCY DEVICES FROM 3 GHZ TO 100 GHZ ...................... 7
DETERMINATION OF THE RELEVANT MATERIAL PARAMETERS........................................ 8 3.1 3.2 LITERATURE SURVEY AND MEASUREMENT TECHNIQUES ............................................................................. 8 MEASUREMENT RESULTS IN THE FREQUENCY RANGE 75 100 GHZ ........................................................ 12 MEASUREMENT OF DIELECTRIC PROPERTIES OF SKIN TISSUE .......................................................... 13 MEASUREMENT OF DIELECTRIC PROPERTIES OF EYE TISSUE ........................................................... 17
3.2.1 3.2.2 3.3
MEASUREMENT RESULTS IN THE FREQUENCY RANGE 200 MHZ 20 GHZ ............................................... 20 MEASUREMENT OF DIELECTRIC PROPERTIES OF SKIN TISSUE .......................................................... 21 MEASUREMENT OF DIELECTRIC PROPERTIES OF EYE TISSUE ........................................................... 24
3.3.1 3.3.2 4
ANALYSIS OF THE FIELD STRENGTHS IN HUMAN TISSUE ................................................... 26 4.1 MODELS OF THE OBJECTS UNDER INVESTIGATION ................................................................................... 26 LAYERED MODEL OF THE HUMAN SKIN ......................................................................................... 26 MODEL OF THE HUMAN EYE ........................................................................................................ 27
4.1.1 4.1.2 4.2
SIMULATION METHODS ........................................................................................................................ 29 ANALYTICAL METHOD ................................................................................................................. 29 THE FINITE DIFFERENCE TIME DOMAIN METHOD.......................................................................... 30
4.2.1 4.2.2 4.3
SIMULATION OF THE FIELD DISTRIBUTION INSIDE THE HUMAN EYE AND SKIN ............................................ 31 FIELD DISTRIBUTION IN THE HUMAN SKIN..................................................................................... 31 FIELD DISTRIBUTION IN THE HUMAN EYE...................................................................................... 35
4.3.1 4.3.2 5
INVESTIGATION OF THERMAL EFFECTS.................................................................................. 39 5.1 5.2 INFRARED THERMOGRAPHY .................................................................................................................. 39 HUMAN SKIN....................................................................................................................................... 39 EXPERIMENTAL SETUP................................................................................................................. 39 MEASUREMENT RESULTS ............................................................................................................. 40
5.2.1 5.2.2

5.2.3 5.3
DISCUSSION .............................................................................................................................. 42
PORCINE EYE ...................................................................................................................................... 43 EXPERIMENTAL SETUP................................................................................................................. 43 MEASUREMENT RESULTS ............................................................................................................. 43 DISCUSSION............................................................................................................................... 45
5.3.1 5.3.2 5.3.3 5.4
SIMULATION OF THERMAL EFFECTS ....................................................................................................... 45 MATHEMATICAL MODEL OF BIO-HEAT-TRANSFER ......................................................................... 45 LAYERED MODEL OF SKIN ........................................................................................................... 46 RESULTS.................................................................................................................................... 47 DISCUSSION............................................................................................................................... 48
5.4.1 5.4.2 5.4.3 5.4.4 6 7 8
SUMMARY.......................................................................................................................................... 48 REFERENCES .................................................................................................................................... 50 APPENDIX .......................................................................................................................................... 52 8.1 COMPARISON OF MEASURED DIELECTRIC PROPERTIES AND DATA FROM LITERATURE ................................. 52
1 Subject of Investigation
On behalf of the FGF the IMST has carried out an investigation referring to biological effects in the cm/mm wave range. The technical part of the project dealt with the analysis of field strengths inside human tissue. The investigation was made with respect to the electric and magnetic fields inside the human eye and the skin on the back. This restriction was made because of the low vascularity and bad thermoregulation of the human eye. A similar statement can be made for the skin on the back because of the relative bad vascularity and the low density of sweat glands. Part II of the project contained the determination of the dielectric parameters of human tissue in the frequency range from 200 MHz up to 100 GHz. The tissues under investigation included muscle, fat, skin (dermis), cornea, retina, lens, sclera, vitreous body, and liquid from camera anterior. First of all a literature study was made concerning the dielectric parameters of human tissue. In a next step the unknown dielectric parameters were measured with a material measurement system in the frequency range of interest. In part III the field strengths in the human eye and skin were simulated. The exposure realized was a linearly polarized plane wave. For the human skin a layered model was used. Besides epidermis and dermis, fat and muscle tissue were distinguished. The human eye was modeled as a quasi-ellipsoid with the tissues mentioned above.
2 Radiofrequency Field Exposure Standards

In nearly any country the protection of human beings against harmful influences is a task of government. Many different organizations are trying to get a work-out of national and international rules and standards to get technical conditions to put this political aim into practice. 2.1 Legal Classification of Standards To classify standards, rules and regulations it is important to distinguish between the following terms: National and international standards Laws (for example national laws of protection against immission like the German Bundes-Immissionsschutzgesetz [BImSchV 1996]) Recommendations (for example presented by the German national radiation protection commission Strahlenschutzkommission [SSK 1993]) Voluntary consumer protection standards (for example standard for low radiation computer screens, the MPR II standard) From the legislators point of view a careful distinction between the terms must be made. In the Federal Republic of Germany the standard DIN 0848 part 2 (similar to ANSI C95.1 [ANSI 1991]) dealing with the protection of human beings from electromagnetic fields, has not been put into asserted German right so far. A German law, called Bundes-Immissionsschutz-
gesetz, is just valid for wireless non mobile installations (indeed there is currently an attempt to extend this law in respect to the DIN 0848). The recommendation of the national German radiation protection association has no legally binding function as well as the voluntary consumer protection standards. 2.2 Standards Associations The approved institution by the German government for preparing standards is the Deutsches Institut fr Normung (DIN), which promotes the harmonization of the standards for Europe. In the field of high frequency electromagnetic fields the reference levels of the current CENELEC (Comit Europen de Normalisation Electrotechnique) prestandard [ENV 50166] are very close to the most recent draft of the DIN/VDE standard 0848 [DIN 0848 91]. The corresponding guideline ANSI C95.1 published by the American National Standards Institute [ANSI 1991] is not only used in the USA, but also by many other countries (for example Australia). The US Federal Communications Commission (FCC) issued a report and order on the 1st of August 1996 [FCC 1996], which requires routine dosimetric assessment of mobile telecommunications devices, either by laboratory measurement techniques or by computational modeling, prior to equipment authorization or use. One of the most important organizations deals with the international development of standards is the INIRC of the IRPA (INIRC: International Non-Ionizing Radiation Comitee; IRPA: International Radiation Protection Association). Its publications [IRPA 1988], [IRPA 1991] take a special place as they represent a summary of the Environmental Health Criteria published in the WHO (World Health Organization) [WHO 1993]. If they have not established national standards themselves, some countries, for example Norway, directly use the IRPA values (CENELEC survey [CENELEC 1995]). In all exposure limits safety factors have already been introduced, which partly explains the differences in the reference levels of the existing worldwide radiofrequency field exposure standards. 2.3 Distinction Between Areas, Time of Exposure and Frequencies In nearly all standards a distinction between exposure areas and exposure times is made. There is a general distinction between two different areas (with different names, but very close meaning), which are called exposure area number 1 and 2 in the DIN standard and controlled and uncontrolled environment in the ANSI document. Controlled environments are locations where there is exposure that may be incurred by persons who are aware of the potential for exposure as a concomitant of employment, by other cognizant persons. Concerning these areas, the maximum permissible exposure is defined in respect to human safety. These areas contain: Controlled areas, for example manufacturing plants
General accessible areas where it is secured that the exposure is only at short times due to the operation of equipment or due to the time of stay. Short time means due to DIN regulation up to 6 hours a day. The reference levels in uncontrolled environment have been fixed under consideration of additional safety precautions. These areas include long-term exposure and locations with exposure of individuals who have no knowledge or control of their exposure: Areas with residential and social buildings Facilities for sports, leisure and relaxation Working places where an electromagnetic field is unexpected. The limits for uncontrolled environment are lower or equal to those for controlled environment. In addition to the introduction of different exposure areas a distinction between exposure times is made. An international limit is made at 6 minutes exposure time ([ANSI 1991]: frequency dependent from 30 minutes at 3 GHz down to 0.62 minutes at 100 GHz). For short-term exposure, higher field strengths are admissible, because it takes a certain time until the human body warms up. Due to the influence of frequency on important parameters, as the penetration depth of the electromagnetic fields into the human body and the absorption capability of different tissues, the limits in general are frequency dependent. 2.4 Basic Restrictions and Derived Reference Levels There is a distinction between basic restrictions and derived reference levels concerning all normative regulations. Basic restrictions are defined for the specific absorption (SA, dimension: energy/mass), the specific absorption rate (SAR, dimension: power/mass) the electrical current density in the body and the current through the body because they can be referred directly to thermal based biological effects. It has been pointed out that in the high frequency range especially the specific absorption rate (SAR) is a useful and a biologically relevant quantity to describe the effect of the electromagnetic field. It is a measure of the power absorbed per unit mass. The unit of specific absorption rate is watt per kilogram (W/kg). The SAR may be spatially averaged over the total mass of an exposed body or its parts, and may be time-averaged over a given time of exposure or even a single pulse or modulation period of the radiation. A limitation of the specific absorption rate prevents an excessive heating of the human body by electromagnetic radiation. As it is difficult to determine these basic quantities directly by measurement, the standards specify a set of more-readily-measurable reference levels in terms of external electric and magnetic field strength and power density, derived from the basic restrictions. These limits have been fixed so that even under worst case conditions, the basic limits are not exceeded. It must be noted that already precaution factors have been introduced into the basic restrictions, which
are different from each other due to deviations in the rating of the potential of danger of electromagnetic fields. Thus there exist different values in the limits although all standards consider the latest scientific knowledge. The most general claim in every standard is: Compliance is established when the basic limits are not exceeded. At frequencies between 100 kHz and 6 GHz the limits for the electromagnetic field strengths may be exceeded if the exposure condition can be shown by appropriate techniques to produce SARs below the corresponding limits [ANSI 1991]. 2.5 Maximum permissible exposure for high frequency devices from 3 GHz to 100 GHz Having in mind a worst case consideration, all limits listed in the following count for the uncontrolled environment and for long time exposure. Table 1 contains the relevant basic limits for the specific absorption rate in the frequency range from 3 GHz to 100 GHz. Because the IRPA standard [IRPA 1991] only considers absorption rates related to the whole body (0.08 W/kg) this standard is omitted in Table 1.
Standard
Status
f [GHz] 3 - 100
Averaging
SAR limit [W/kg] 2.0
Reference
DIN VDE 0848 Teil 2, 1991 CENELEC ENV 50166-2, 1995 ANSI C95.11991
draft
10 g mass
[DIN 0848 91]
draft
3 - 100
10 g mass
2.0
[ENV 50166]
in force
6 >6
1 g mass -
1.6 -
[ANSI 1991]
Table 1: Relevant basic limits for the specific absorption rate (SAR), valid for high frequency devices in the frequency range of interest from 3 GHz to 100 GHz. In Table 2 the derived reference levels are listed. In contrast to the European and German prestandard the ANSI standard defines the exposure to radiofrequency electromagnetic fields above 6 GHz as quasi-optical. Therefore no SAR limit is valid. On the other hand the equivalent power density is limited as shown in Table 2. The rationale of the ANSI standard is to define the frequency region from 6 GHz up to 300 GHz as a transition area between the complex field behaviour at the lower frequencies up to 6 GHz and the simple surface heating process induced for electromagnetic waves in the optical frequency range. Therefore the ANSI standard defines a maximum permissible power density for partial body exposure of all parts of the body except the eyes and testes and device distances no closer than 20 cm, which starts from 40 W/m2 (3 GHz - 6 GHz) up to 200 W/m2 (30 GHz -100 GHz) and a linear increasing
function from 6 GHz up to 30 GHz. For comparison: the power density of the sun light hitting the earth is in the order of 1 kW/m2 [Bohrmann 1993]. The maximum permissible exposure listed in Table 2 according to the ANSI standard is valid for the eyes and the testes in order to take into account the worst case.
Standard
Status
f [GHz]
Eeff [V/m] Heff [A/m]
S [W/m]
Reference
DIN VDE 0848 Teil 2, 1991 CENELEC ENV 50166-2, 1995 ANSI C95.1
draft
3 - 100
61.4
0.16
10
[DIN 0848 91] [ENV 50166]
draft
3 - 100
61.4
0.5
0.163
0.5
10
in force
3 - 15 50.11 f 15 - 100
0.133 f
f / 0.15 [ANSI 1991] 100 10 [IRPA 1991]
194.1 61
0.515 0.16
IRPA
in force
3 - 100
Table 2: Relevant derived reference levels for the electromagnetic field valid for high frequency devices in the frequency range of interest from 3 GHz to 100 GHz.
3 Determination of the Relevant Material Parameters

3.1 Literature Survey and Measurement Techniques Investigations of the dielectric properties of human tissues are presented in the literature since more than 40 years (e.g. [Schwan 57][Schwan 80][Gabriel 96a]). The results are obtained by measurements of animal and human tissue in the frequency range up to 90 GHz [Edrich 1976]. Depending on the frequency range of interest three classes of measurement systems have to be distinguished. Impedance measurement systems in liquid cells with a typical frequency range from 10 Hz up to 30 MHz. An LCR meter or impedance analyzer is required to measure the relative permittivity r and the conductivity . Measurement of the reflection coefficient of an open-ended line, which is immersed in the liquid under test or attached to the solid under test (Fig. 1). The magnitude and phase of the signal reflected at the open-end depends on the dielectric properties of the material under test. A vector network analyzer measures the reflection coefficient of the sample. A measurement software converts the measured data into r and . The typical frequency range is 200 MHz to 20 GHz.
Fig. 1: Material measurement system II of the IMST. Measurement of the reflection coefficient of an open-ended coaxial line in the frequency range 200 MHz - 20 GHz. Measurement of the reflection and transmission coefficients of a line, which is partly filled with a sample of material. This measurement requires a vector network analyzer too. The lowest measurement frequency is restricted to about 100 MHz. The upper frequency limit is not restricted. The measurement system used in this investigation is a W-band system operating in the frequency range 75 - 100 GHz with rectangular waveguides.
The IMST has material measurement systems at its disposal, which work according to the both last-mentioned principles. With this systems dielectric properties can be measured in the frequency range 200 MHz - 20 GHz and from 75 GHz - 100 GHz. Gabriel [Gabriel 96b] introduced a parametric model, which describes the frequency dependency of the dielectric properties of 17 different human tissues from 10 Hz to 100 GHz. The model is based on measurements [Gabriel 96a] and on experimental data reported in the literature [Gabriel 96c]. The frequency range from 10 Hz to 100 GHz is divided into four main dispersions with different properties. The frequency dependency of each dispersion is described by a Cole-Cole relation according to equation 1. This enables a closed representation which can be directly implemented into existing simulation models of the human body. The dielectric properties of the following tissues are available: blood, bone (cancellous), bone (cortical), brain (gray matter), brain (white matter), fat (average infiltrated), fat (not infiltrated), heart, kidney (cortex), lens (cortex), liver, lung (inflated), muscle, skin (dry), skin (wet), spleen, tendon). In the following Figs. the frequency dependencies evaluated by
( ) = +
are shown for the relevant tissues.
1 + ( j )(
n =1 n
1n )
j0
(1)
10
10 10 10 10 10 10 10 10 10 10
5 4 3 2 1 0
Tissue: skin (wet) rel. permittivity conductivity [S/m]
-1 -2 -3 -4
10
10
10
10
10 10 10 10 Frequency [Hz]
10
10
10
10
11
Fig. 2: Dielectric properties of skin (wet) as a function of frequency.
10 10 10 10 10 10 10 10 10 10
5 4 3 2 1 0
Tissue: skin (dry) rel. permittivity conductivity [S/m]
-1 -2 -3 -4
10
10
10
10
10
10
10
10
11
Fig. 3: Dielectric properties of skin (dry) as a function of frequency.
11
10 10 10 10 10 10 10 10 10 10
7 6 5 4 3 2 1 0
Tissue: fat (average infiltrated) rel. permittivity conductivity [S/m]
-1 -2
10
10
10
10
10
10
10
10
11
Fig. 4: Dielectric properties of fat (average infiltrated) as a function of frequency.
10 10 10 10 10 10 10 10 10 10
7 6 5 4 3 2 1 0
Tissue: fat (not infiltrated) rel. permittivity conductivity [S/m]
-1 -2
10
10
10
10
10
10
10
10
11
Fig. 5: Dielectric properties of fat (not infiltrated) as a function of frequency.
12
10 7 Tissue: muscle 10 6 rel. permittivity 10 conductivity [S/m] 5 10 4 10 3 10 2 10 1 10 0 10 -1 10 -2 10 1 2 3 4 5 6 7 8 9 10 11 10 10 10 10 10 10 10 10 10 10 10 Frequency [Hz]

Fig. 6: Dielectric properties of muscle as a function of frequency. 3.2 Measurement Results in the Frequency Range 75 100 GHz The experimental investigation of dielectric properties of different human tissues in the frequency range 75 100 GHz was carried out using the above mentioned W-band measurement system from Damaskos, Inc.. A rectangular waveguide is partially filled with a small sample of the biological tissue. The relative permittivity and the conductivity are calculated from the measured transmission coefficients. Fig. 7 shows the arrangement of the sample in the W-band measurement system. The sample has dimensions of 2.54 mm 1.27 mm 1.5 mm (width height length) for non-liquid biological tissues. For liquid material one end of the sample is attached to a foil, on the other end the exact shape of the sample is unknown because of the surface tension of the liquid material (as illustrated in Fig. 7). A first study has shown that this phenomenon results in a reduction of the effective length of the sample with respect to its transmission characteristics and therefore has to be considered for the calculation of the dielectric properties. However, in practice this effective length cannot be determined. In order to minimize the effect of surface tension, we soaked a piece of cotton wool with the liquid material. The cotton wool has no significant effect on the calculated dielectric properties. Fig. 8 visualizes a comparison of the measurement results for water using this procedure and theoretical data from literature [Gabriel 96c][Duck 1990]. In this case the original length of the sample was used for the calculation of the dielectric properties. The temperature of the sample was T = 27C. This comparison shows a reasonable agreement between measurement and theory (Fig. 8). The maximum deviation from the theoretical model amounts to 30 50%.
13
rectangular waveguide
tissue sample (liquid)
foil
m 4m 2.5
1.50 mm
Fig. 7: Rectangular waveguide partially filled with a liquid sample of biological tissue.
100 80 60 40 20 0 75 Tissue: water, T=27C theoretical rel. permittivity theoretical conductivity [S/m] measured conductivity [S/m] measured rel. permittivity
80
85 90 Frequency [GHz]
95
100
Fig. 8: Theoretical and measured dielectric properties of water. The temperature of the sample was T = 27C. 3.2.1 Measurement of Dielectric Properties of Skin Tissue Fig. 9 14 present the measurement results of in vitro porcine skin, fat and muscle tissue at temperatures of T = 27C and T = 37C. The dielectric characteristic of porcine tissue is ex-
1.27 mm
14
pected to differ not significantly from human tissue, because there is insufficient evidence to identify consistent variation between species [Duck 1990]. Species-specific variations are probably masked by other sources of variability (local tissue inhomogeneities, changes following death, age). The only tissue showing clear species-dependent variation is skin [Duck 1990]. Our measurements show similar results for porcine skin tissuea and data from human skin tissue reported in literature. In the frequency range from 75 GHz up to 100 GHz all tissues show similar dielectric characteristics as water. The permittivity falls monotonically with frequency and the conductivity shows in a first approximation a constant curve. Skin and fat are showing a lower conductivity and relative permittivity than muscle. The measured values of the relative permittivity and electric conductivity of skin and muscle correspond well with results reported in literature [Gabriel 1996c] and with the parametric curves from [Gabriel 1996b]. Only fat tissue shows a significant deviation from the parametric model. This result may be explained by the large variations for adipose tissue and bone marrow reported in literature. These variations are caused by the wide range of water content in these tissues [Duck 1990]. For all three tissues the high temperature (T = 37C) measurements showed no significant temperature coefficient for the conductivity and relative permittivity. The changes measured were within the measurement uncertainties. From literature [Duck 1990] a temperature coefficient of 1-2% can be expected, which results in a variation of 10-20% for both, conductivity and relative permittivity. This changes are below the resolution of the 75-100 GHz measurement setup, as discussed in the next section.
80
60
40
Tissue: skin, T=27C conductivity [S/m] rel. permittivity
20
0 75
80
95
100
Fig. 9: Measured dielectric properties of skin tissue. The temperature of the sample was T = 27C.
15
80
60
40
Tissue: skin, T=37C conductivity [S/m] rel. permittivity
20
0 75
80
95
100
80
60
40
Tissue: fat, T=27C conductivity [S/m] rel. permittivity
20
0 75
80
95
100
Fig. 11: Measured dielectric properties of fat tissue. The temperature of the sample was T = 27C.
16
80
60
40
Tissue: fat, T=37C conductivity [S/m] rel. permittivity
20
0 75
80
95
100
120 100 80 60 40 20 0 75 80 85 90 Frequency [GHz] 95 100 Tissue: muscle, T=27C conductivity [S/m] rel. permittivity
Fig. 13: Measured dielectric properties of muscle tissue. The temperature of the sample was T = 27C.
17
120 100 80 60 40 20 0 75 80 85 90 Frequency [GHz] 95 100 Tissue: muscle, T=37C rel. permittivity conductivity [S/m]
Fig. 14: Measured dielectric properties of muscle tissue. The temperature of the sample was T = 37C. 3.2.2 Measurement of Dielectric Properties of Eye Tissue Fig. 15 to 20 present the measurement results of in vitro porcine eye tissue.
80
60 Tissue: vitreous body, T=27C conductivity [S/m] rel. permittivity
40
20
0 75
80
95
100
Fig. 15: Measured dielectric properties of vitreous body tissue. The temperature of the sample was T = 27C.
18
80
60
40 Tissue: cornea, T=27C conductivity [S/m] rel. permittivity
20
0 75
80
95
100
Fig. 16: Measured dielectric properties of cornea tissue. The temperature of the sample was T = 27C.
80 Tissue: lens, T=27C conductivity [S/m] rel. permittivity
60
40
20
0 75
80
95
100
Fig. 17: Measured dielectric properties of lens tissue. The temperature of the sample was T = 27C.
19
80
60
40 Tissue: retina, T=27C conductivity [S/m] rel. permittivity
20
0 75
80
95
100
Fig. 18: Measured dielectric properties of retina tissue. The temperature of the sample was T = 27C.
100 80 60 40 20 0 75 Tissue: sclera, T=27C conductivity [S/m] rel. permittivity
80
95
100
Fig. 19: Measured dielectric properties of sclera tissue. The temperature of the sample was T = 27C.
20
100 80 60 40 20 0 75 Tissue: liquid from camera anterior, T=27C conductivity [S/m] rel. permittivity
80
95
100
Fig. 20: Measured dielectric properties of liquid from camera anterior. The temperature of the sample was T = 27C. As already seen for the skin tissues all eye tissues show similar dielectric characteristic as water. The permittivity falls monotonically with frequency and the conductivity shows in a first approximation a constant curve. The results contain some experimental uncertainties because of the small dimensions of the sample. Especially liquid tissues show significant loss in water content due to drying effects during probe preparation and measurement. Therefore measurements at higher temperatures (i.e. T = 37C) are extremely difficult for liquid material and dropped in this report. Due to the following effects: shape of the small probe, inhomogeneity of tissue material, dynamic of the measurement system, drying of tissue, the uncertainty of the measurement setup can be estimated by about 20-40%. However, the dielectric properties of biological tissue itself show a significant variability. There are several factors, which affect the dielectric properties of tissue: post-mortem changes, local tissue inhomogeneities, age, animal species, and temperature [Duck 1990] [Edrich 1976]. 3.3 Measurement Results in the Frequency Range 200 MHz 20 GHz The experimental investigation of dielectric properties of different human tissues in the frequency range 200 MHz 20 GHz was carried out using the above mentioned measurement
21
system II shown in Fig. 1. The dielectric properties are calculated from the measured reflection coefficient of an open-ended line attached to the tissue sample. 3.3.1 Measurement of Dielectric Properties of Skin Tissue Fig. 21 and Fig. 22 show the measurement results of porcine skin tissue for a tissue temperature of 27C and 37C, respectively. Fig. 23 and Fig. 24 show the measurement results of porcine fat tissue for a tissue temperature of 27C and 37C, respectively. Fig. 25 and Fig. 26 show the measurement results of porcine muscle for a tissue temperature of 27C and 37C, respectively. Figures 21-26 include the results from the 75-100 GHz measurement. Dielectric properties of porcine skin, muscle and fat tissue show characteristics comparable to data from the parametric model [Gabriel 96b]. Due to the measurement principle which uses a greater amount of tissues with a well-defined interface at the open-ended line the dielectric properties obtained in the frequency range from 200 MHz to 20 GHz are more accurate. Taking into account the uncertainties of the 75-100 GHz measurement both measurements show consistent results. Only for fat tissue the 75-100 GHz measurement of the dielectric properties and the 200 MHz-20 GHz measurement show slightly inconsistent behavior. This may be caused by the great variety of water content of fat tissue reported in [Duck 90]. In order to use fresh tissue the different measurements were conducted using different (fresh) tissues.
80
60
40
Tissue: skin, T=27C rel. permittivity conductivity [S/m]
20
0 0,1
1 Frequency [GHz]
10
100
Fig. 21: Measured dielectric properties of skin. The temperature of the sample was T = 27C.
22
80
60
40
Tissue: skin, T=37C rel. permittivity conductivity [S/m]
20
0 0,1
1 Frequency [GHz]
10
100
80 Tissue: fat, T=27C rel. permittivity conductivity [S/m]
60
40
20
0 0,1
1 Frequency [GHz]
10
100
23
80 Tissue: fat, T=37C rel. permittivity conductivity [S/m]
60
40
20
0 0,1
1 Frequency [GHz]
10
100
120 100 80 60 40 20 0 0,1 Tissue: muscle, T=27C rel. permittivity conductivity [S/m]
1 Frequency [GHz]
10
100
Fig. 25: Measured dielectric properties of muscle tissue. The temperature of the sample was T = 27C.
24
120 100 80 60 40 20 0 0,1 Tissue: muscle, T=37C rel. permittivity conductivity [S/m]
1 Frequency [GHz]
10
100
Fig. 26: Measured dielectric properties of muscle tissue. The temperature of the sample was T = 37C. 3.3.2 Measurement of Dielectric Properties of Eye Tissue
80
60 Tissue: vitreous body, T=27C rel. permittivity conductivity [S/m]
40
20
0 0,1
1 Frequency [GHz]
10
100
Fig. 27: Measured dielectric properties of vitreous body. The temperature of the sample was T = 27C.
25
80
60 Tissue: vitreous body, T=37C rel. permittivity conductivity [S/m]
40
20
0 0,1
1 Frequency [GHz]
10
100
Fig. 28: Measured dielectric properties of vitreous body. The temperature of the sample was T = 37C. Fig. 27 and Fig. 28 shows the measurement results for a tissue temperature of 27C and 37C, respectively. For a sample temperature of 27C the results for the frequency range of 75100 GHz are included. As seen in the previous section, the results for both frequency ranges are consistent within measurement uncertainties and tissue variability. Taking into account this variability it can be stated that reliable data of the dielectric properties of the tissues under investigation has been collected for the following analysis of electromagnetic fields in the human skin and eye.
26
4 Analysis of the Field Strengths in Human Tissue

4.1 Models of the Objects under Investigation 4.1.1 Layered Model of the Human Skin The skin of an adult human being covers about 2 m2, and has a thickness between 1.5 mm and 4 mm. Underneath a fat layer (tela subcutanea) is situated, which is located above muscle tissue. The schematic structure of the human skin including the fat layer is shown in Fig. 29. The epidermis is the outer part of the skin. It is cornified, without vessels, multilayered and normally dry. The dermis, which represents the rest of the human skin except the fat layer, is the wet skin region.
1 mm
Fig. 29: Structure of the human skin. The fat layer begins in region 8. In the cm/mm wave range the penetration depth of electromagnetic fields in the human body is very small. A major part of the electromagnetic energy is absorbed in the surface of the human body. Because of the high frequency the electromagnetic field can be locally described by a planar wave. Therefore the field theoretical problem is reduced to a one dimensional investigation of the field distribution in a layered medium schematically shown in Fig. 30. For this investigation the field distribution in the human skin on the back is analyzed. This restriction is made because of the relative bad vascularity and the low density of sweat glands of the human skin on the back. The typical thickness of the different layers is taken from the literature [FSZ 1985][Lippert 1990]. These values are summarized in Table 3 together with the mass density of the layers [Dimbylow 1988][Dimbylow 1991].
27
Tissue
Mass density [103 kg/m3] 1.1 1.1 0.92 1.04
Thickness [mm]
epidermis dermis fat muscle
0.15 3.85 10 > 10
Table 3: Thickness and mass density of the different layers of the human skin on the back including typical values for fat and muscle (40 mm assumed in the simulation) tissue. For the problem shown in Fig. 30, the oblique incidence of a planar wave on a layered medium, an analytical solution exists in the literature (e.g. [Balanis 1989]). This algorithm was implemented and a post-processing was made to determine the SAR values according to the ANSI standard, the DIN and the ENV prestandard respectively.
epidermis dermis fat muscle
Fig. 30: A planar wave hitting the human skin on the back. 4.1.2 Model of the Human Eye The schematic structure of the human eye is shown in Fig. 31. The tissues of the eye are: retina (1), choroidea (2), sclera (3), cornea (4), tunica conjunctiva (5), iris (6), corpus ciliare (7), lens (8), camera anterior (9), camera posterior (10), pupilla (11), vitreous body (12), macula (13), discus nervi optici (14), and nervus opticus (15). For the numerical investigation of the field distribution a spherical voxel model of the human eye has been built up. The eye has a diameter of d = 20.8 mm and is embedded in muscle and skin tissue. Fig. 32 shows a cut-plane through this voxel model.
28
Fig. 31: Structure of the human eye [Lippert 1990].

4 mm sclera choroidea retina camera anterior cornea nervus opticus
lens iris vitreous body y skin muscle z x
Fig. 32: Horizontal cut-plane through the voxel model of the human eye.
Biological Effects in the cm/mm Wave Range, Part II/III 4.2 Simulation Methods 4.2.1 Analytical Method
29
The analytical method used for the calculation of the electromagnetic field in layered media is taken from [Balanis 1989]. For normal incidence of a plane wave the following terms provide the reflection coefficients rmn = Enr Z m Zn i = E n Z m + Zn , (2)
and transmission coefficients t mn = Ent 2Z m i = En Z m + Zn (3)
of multiple interfaces, as shown in Fig. 33, with the intrinsic impedances

Zm =
m . m
(4)
1 air r21
2 epidermis r32 r12 t21 t12 d2 x2 t32
3 dermis r43 r23 t43 t23 d3 x3 t34 r34
4 fat r54 r45 t54 t45 d4 x4
5 muscle
d5 x5
Fig. 33: Reflection and transmission coefficients in the layered model of skin. Introducing a phase and an attenuation term upon the traveling E-field of waves propagating in positive coordinate direction
+ + E m ( x m ) = Em ( xm = 0)e am xm e jm xm
(5)
and waves propagating in negative coordinate direction
30
E m ( x m ) = E m ( xm = d m )e am ( xm dm ) e jm ( xm dm ) (6)
the distribution of the electric field within the layers is calculated via a ray-tracing model, i.e. by superposition of the different propagating waves in the media. 4.2.2 The Finite Difference Time Domain Method The calculations of the electromagnetic fields inside the anatomical model of the human eye have been carried out using the finite difference time domain (FDTD) method. In 1966 Yee [Yee 1966] introduced this method, which has become one of the most popular numerical methods, because of the simplicity and stability of the algorithm. The FDTD is a purely numerically oriented method, which directly discretizes Maxwells equations in the time- and space-domain with second-order accuracy. According to the unit cell shown in Fig. 34 the electric field components E are positioned on the middle of the edges and the magnetic field components H are positioned on the middle of the surfaces. Time-stepping is done in a leapfrog way. The magnetic field H at the time (n+1/2)t is determined from the electric field E at the time nt, and afterwards the electric field E at the time (n+1)t is determined from the magnetic field H at the time (n+1/2)t.
E y (i,j+ 1 ,k+1) 2 H y(i+ 1 ,j,k+ 1 ) 2 2 H x(i,j+ 1 ,k+ 1 ) E (i,j,k+ 1 ) 2 2 z 2 1 1 H y (i- 2 ,j,k+ 2 ) H z(i+ 1 ,j+2 ,k) 2
E y (i,j+ 1 ,k) 2
1
E z(i,j+1,k+ 1 ) 2
H x(i,j- 1 ,k+ 1 ) 2 2
E x (i+ 1 ,j,k) 2
z x
y
Fig. 34: Lattice unit cell of the Yee-algorithm in Cartesian coordinates. An important aspect using the FDTD, especially for the solution of radiation problems, is the availability of an appropriate absorbing boundary condition (ABC), because in contrast to other numerical methods like the method of moments, the problem space of the FDTD is limited. To simulate free space conditions a special algorithm has to be evaluated at the outer grid planes of the FDTD mesh. In 1994 Berenger [Berenger 1994] proposed the perfectly matched layer (PML) absorbing boundary condition with improved performance in orders of magnitude compared to other ABCs. This absorbing boundary condition can be placed in the extreme near-field of the structures under investigation. Therefore this ABC is not only very accurate but also memory efficient.
Biological Effects in the cm/mm Wave Range, Part II/III 4.3 Simulation of the Field Distribution inside the Human Eye and Skin 4.3.1 Field Distribution in the Human Skin
31
In this section the SAR distribution inside the human skin according to the model introduced in Table 3 is analyzed in the frequency range from 3 GHz to 100 GHz. The incident electromagnetic field consists of a linearly polarized plane wave with a power density of 1 mW/cm2. The power density was chosen according to the derived reference level valid for frequencies higher than 3 GHz for uncontrolled environment in the European [ENV 50166] and German prestandard [DIN 0848 91]. The material parameters are taken from [Gabriel 96b]. Fig. 35 shows the frequency dependent SAR value in the human skin in comparison to the European [ENV 50166] and German prestandard [DIN 0848 91]. It can be seen, that the SAR value below 20 GHz has a relative complex frequency dependence. The reason for this is the existence of standing waves in the human skin. Above 20 GHz the SAR value is mainly influenced by the decreasing reflection coefficient of the boundary epidermis-air. This reflection coefficient is defined by 1 r (7) r= 1+ r and r the complex relative permittivity of epidermis. The absolute value is a linear decreasing function of frequency in the whole frequency range of interest (20 GHz: 0.69, 100 GHz: 0.54). Comparing the simulated SAR values with the basic restriction for the SAR it can be stated that there exist a 6 times minimum safety margin when the amplitude of the incident plane wave is chosen according to the derived reference level.
2,5 ENV and DIN standard 2,0

SAR [W/kg]
1,5 1,0 0,5 0,0 0 20
Human skin: SAR_10g
80
100
Fig. 35: Simulated SAR values resulting from an incident power density of 1 mW/cm2 inside the model of the human skin according to Table 3 in comparison to the European [ENV 50166] and German prestandard [DIN 0848 91].
32
The SAR value according to the ANSI standard [ANSI 1991] shows a very similar frequency dependence. Because of the lower averaging mass the simulated SAR values are higher than the SAR according to the European [ENV 50166] and German prestandard [DIN 0848 91]. Like stated before in section 2.5 the ANSI standard defines the specific absorption rate only for frequencies up to 6 GHz. For these frequencies a 4 times minimum safety margin exists for the SAR in the human skin. For the higher frequency range the incident power density S is the relevant reference level. According to Table 2 S is restricted to 100 W/m2 for frequencies higher than 15 GHz.
2,5 2,0 ANSI standard

SAR [W/kg]
1,5 1,0 0,5 0,0 0 20 40 60 Frequency [GHz]
Human skin: SAR_1g
80
100
Fig. 36: Simulated SAR values resulting from an incident power density of 1 mW/cm2 inside the model of the human skin according to Table 3 in comparison to the American standard [ANSI 1991]. The maximum SAR values in the different layers of the human skin are shown in Fig. 37-38. Because of the strong increase of the losses in the human skin as a function of frequency the highest SAR values are found in the outer layers of the human skin. In the fat and muscle region noticeable SAR values only exist for frequencies up to 20 GHz. On the other hand the strong losses lead to high SAR values at the surface of the human skin. As shown in Fig. 37 a maximum of 34 W/kg was found for 100 GHz. Finally the SAR distribution inside the model of the human skin is shown in Fig. 39-54 for five frequencies. For the two frequencies at the lower end of the frequency range of interest shown in Fig. 39 a typical standing wave behavior can be observed for the skin and fat region. Because of the low permittivity and conductivity of the fat layer a sharp discontinuity appears at the boundaries dermis-fat and fat-muscle. For 3 GHz the maximum SAR value is not found at the surface of the skin but at the boundary dermis-fat. The SAR values for the three higher frequencies depicted in Fig. 54 are linearly decreasing. The highest values are found at the surface and nearly all energy is absorbed in the outer skin region. For 77 GHz a decrease of the SAR of nearly 10 decades is observed from the epidermis layer to the fat layer.
33
35 30
SAR [W/kg]
25 20 15 10 5 0 0
Human skin: epidermis dermis
20
80
100
Fig. 37: Simulated maximum of the SAR values resulting from an incident power density of 1 mW/cm2 inside the outer layers of the human skin according to Table 3.
0,18 0,16 0,14

SAR [W/kg]
0,12 0,10 0,08 0,06 0,04 0,02 0,00 0 20 40 60 Frequency [GHz]
Human skin: fat muscle
80
100
Fig. 38: Simulated maximum of the SAR values resulting from an incident power density of 1 mW/cm2 inside the inner layers of the human skin according to Table 3.
34
0 -10
SAR [dB W/kg]
Frequency: 3 GHz 6 GHz
-20 -30 -40 -50 0 10 20 x [mm] 30 40
Fig. 39: SAR distribution inside the model of the human skin according to Table 3 resulting from an incident power density of 1 mW/cm2.
20 0
SAR [dB W/kg]
-20 -40 -60 -80
Frequency: 24 GHz 77 GHz 100 GHz
-100 0 5 10 x [mm] 15 20
Fig. 40: SAR distribution inside the model of the human skin according to Table 3 resulting from an incident power density of 1 mW/cm2.
35
4.3.2 Field Distribution in the Human Eye The FDTD-model of the human eye is shown in Fig. 32. In Table 4 the relative permittivity and the electrical conductivity of eye and skin tissue are summarized for a frequency of f = 77 GHz. For the excitation a plane wave described by the phasor of the electric field E = E 0 e z e j0 x (8)
with a power density of S = 1 mW/cm is applied. Fig. 41 shows the distribution of the specific absorption rate in the xy-plane, Fig. 42 in the xz-plane and Fig. 43 in the yz-plane. Due to the strong attenuation of the wave inside the eye and skin, most of the power is absorbed in the superficial tissues. The field distribution in Fig. 41 (xz-plane) and in Fig. 42 (yz-plane) are very similar. In addition this symmetry is visible in Fig. 43. In Table 5 and Fig. 44 the maximum values of the specific absorption rate (SAR) are displayed in the different tissues of the model as well as the 1g - averaged SAR value for all tissues. The maximum SAR value occurs in the cornea. r [S/m]
Tissue
Mass density [103 kg/m3] 1.04 1.1 1.06 1.035 1.006 1.06 1.006 1.058 1.1 1.035 1.1
muscle sclera choroidea retina vitreous body cornea camera anterior iris lens nervus opticus skin
19.84 22.49 19.84 8.41 10.33 5.82 5.08 19.84 14.25 19.84 11.67
106.22 76.76 106.22 54.60 40.9 56.27 50.02 106.22 29.44 106.22 55.60
Table 4: Dielectric parameters of model of the human eye for a frequency of f = 77 GHz. The mass density has been taken from [Dimbylow 1988][Dimbylow 1991][Flindt 1995] and [Geigy 1985]. r and are measured by IMST.
36
20 18 16 14
y [mm]
12 10 8 6 4 0 2 4 6 8 x [mm] 10 12
SAR in db W/kg 10+ -2,5 bis 10 -15 bis -2,5 -27,5 bis -15 -40 bis -27,5 -52,5 bis -40 -65 bis -52,5 -77,5 bis -65 -90 bis -77,5
Fig. 41: Distribution of the specific absorption rate in the xy-plane of the eye.
8 6 4 2
z [mm]
0 -2 -4 -6 -8 0 2 4 6 8 x [mm] 10 12
Fig. 42: Distribution of the specific absorption rate in the xz-plane of the eye.
37
10 8 6 4 2
z [mm]
0 -2 -4 -6 -8 -10 2 4 6 8 10 12 14 y [mm] 16 18 20 22
Fig. 43: Distribution of the specific absorption rate in the yz-plane (10 mm into the body from the surface of the skin).
100 10 1
SAR [W/kg]
0,1 0,01 0,001
0,0001 1E-005
le ra ea na dy ea ior iris lens cus kin s usc scle roid reti s bo corn nter pti m a u ho so c reo era rvu vit am ne c
Fig. 44:. Maximum local SAR values in the different tissues of the human eye model
38
Tissue muscle sclera choroidea retina vitreous body cornea camera anterior iris lens nervus opticus skin all eye tissue
SAR [W/kg] 2.31810-5 32.49 1.714 4.95810-2 1.09710-2 45.11 0.3657 0.921 1.5810-3 < 110-5 32.24 SAR_1g = 0.6588
Table 5: Maximum local SAR values in the different tissues of the eye model for plane wave excitation (normal incidence) with 1 mW/cm. The calculations show similar results for the layered skin model and for the three dimensional model of the human eye for a frequency of f = 77 GHz. In both cases the absorption is dominant in superficial tissues and the 1g-averaged SAR values are comparable: for the layered skin model the 1g-averaged SAR value is 0.6 W/kg and for the model of the human eye the local SAR value amounts to 0.6588 W/kg. The same applies to the maximum local SAR values: for the layered model of skin the maximum local SAR value is 27 W/kg and for the model of the human eye the local SAR value amounts to 45 W/kg in cornea tissue. Small differences in the calculations result from different material parameters which were used in the different models: the calculations in the layered model of skin are based upon dielectric parameters taken from [Gabriel 96c], because it provides data in a wide frequency range. The simulations of the field distribution in the human eye are based on measured dielectric parameters from IMST. Therefore, the maximum local SAR value in skin tissue is slightly higher and amounts to 32.24 W/kg. Due to the high absorption of the electromagnetic fields in superficial tissue the averaged values for higher frequencies are determined mainly by the parameters of the first layer of the model including the reflection coefficient of the interface air-skin.
39
5 Investigation of Thermal Effects

5.1 Infrared Thermography For the investigation of thermal effects in the human skin and in the porcine eye a high-speed thermal image system (Thermo Tracer TH 2111, NEC San-ei Instruments, Ltd.) is used. This system is a non-contact type infrared thermometer. All objects above the absolute zero (-273C) continuously radiate infrared energy. Therefore, infrared rays are closely related to the temperature of physical bodies. The detector unit of the thermal image system scans the surface of an object and collects the infrared energy by an infrared objective lens. After chopping this infrared energy with a reference temperature source it is converted to an electrical signal using an infrared HgCdTe-detector. The infrared detector remains cooled to -196C by liquid nitrogen and is capable of converting infrared energy with high sensitivity. For more information about IR thermography see [Cho 1992]. The main performance specifications of the system are as follows: Temperature resolution: 0.1C (for blackbody at 30C), 0.02C (in S/N improvement mode, for blackbody at 30C). Frame time: 20 frames / s. Detector unit: HgCdTe (liquid nitrogen cooling type), measurement wavelength: 8-13 m (half value width). 5.2 Human Skin 5.2.1 Experimental setup A generator based on a Gunn diode oscillator with output powers up to 38 mW was used as a source of millimeter electromagnetic irradiation. The frequency of the continuous wave (CW) signal is f = 77 GHz and the aperture of the rectangular horn antenna amounts to an area of 1.5 cm 1.1 cm. The gain of the horn antenna is G = 20 dBi. An estimation of the power density yields S = 1 mW/cm in a distance of d = 17.3 cm in front of the aperture and a power density of 10 mW/cm in a distance of d = 5.5 cm. (S = 10 mW/cm represents the maximum permissible exposure due to the ANSI Standard [ANSI 1991] and S = 1 mW/cm for the DINVDE prestandard [DIN 0848 91]). The antenna is aimed at the forearm of a volunteer and the time-dependent temperature field of the region of interest is recorded by the thermal image system. The schematics of the experimental setup are shown in Fig. 45. All experiments were conducted in an anechoic chamber. Before the measurement the volunteers had a time of rest in order to wait for thermal equilibrium in the human forearm. The thermophysiological response of the skin in-vivo were determined for different distances d between skin and aperture. The power was set to the maximum of P = 38 mW for all measurements. The ambient temperature during the different measurements was between 21C and 22C, but was constant (0.1C) for each measurement. The
40
relative humidity of the air was between 50% and 60%. There was essentially no air motion in the laboratory. Due to (a) small motion artifacts, (b) the resolution of the thermal image system and (c) slight changes in the ambient conditions the accuracy of the measurement is about 0.1-0.2C in the region of interest.
generator
horn antenna
d IR camera skin
Fig. 45: Block diagram of the experimental setup. 5.2.2 Measurement results The measured temperature changes in the skin of two volunteers are summarized in Fig. 46. The measurements show temperature changes up to 2.2C for a distance of d = 18 mm and a strong decay in temperature rise for greater distances. For d > 10 cm no significant temperature change occurred. Fig. 47 (A) and (B) show, as an example, the temperature field of the human forearm before and after seven minutes of irradiation for the female subject and a distance of d = 2 cm. Subtraction images of (A) and (B) are shown in Fig. 47 (C) and (D). To make the spatial extension of the temperature rise more clear subplot (D) contains four isothermal lines. The corresponding time course of the cursor temperature ( = region with maximum temperature rise) is depicted in Fig. 48. During irradiation the temperature shows an exponential increase and reaches steady state conditions after four to seven minutes of exposure. After irradiation the temperatures return to the equilibrium temperature.
41
2,5 2 d
T [C]
1,5 1 0,5 0 10
Maximum temperature change in the human skin (P = 38 mW) volunteer 1 (male) volunteer 2 (female)
20
30
40 50 Distance d [mm]
60
70
80
Fig. 46: Temperature changes in the human skin caused by RF irradiation.

T [C] 35.3 34.3 33.3 32.3 31.3 (A) before irradiation T [C] 35.3 34.3 33.3 32.3 31.3 (B) after 7 minutes of irradiation
T [C]
2.0 1.0 0.0 -1.0 -2.0 (C) subtraction image
T [C]
2.0 1.5 1.0 0.5 0.0 (D) subtraction image with isolines
Fig. 47: Temperature field (A) before and (B) after seven minutes of irradiation. (C) Subtraction image of (B) and (A). (D) Subtraction image with isothermal lines.
42
OFF 34,5
ON
OFF
d = 2 cm, skin
Temperature [C]
34
33,5
33
32,5
9 Time [min]
12
15
Fig. 48: Time course of cursor temperature for d = 20 mm and P = 38 mW. 5.2.3 Discussion The measurements revealed local changes in skin temperature up to 2.2C for irradiation in a distance of d = 18 mm from the radiation source. In order to assess the measured temperature changes some explanatory notes on skin temperature have to be made. The temperature of the human skin is determined by thermophysiological concerns of the human body, which depend on several factors, especially ambient conditions and clothing. For a naked subject an increase of the ambient temperature from 20C to 30C causes a linearly rise of the mean skin temperature from 30C to 34C. In the extremities this dependency is more obvious: For a naked subject an increase of the ambient temperature from 20C to 30C causes a linearly rise of the feet skin temperature from 23C to 33C [Aschoff 1971]. Despite of this dependency there is a shift in skin and core temperature due to diurnal changes of body temperature [Werner 1984][Aschoff 1971]. The amplitude of core temperature variation is up to 2C. In our study the measured changes in skin temperature were below the threshold for warmth sensation of the two volunteers. Considering the above mentioned facts, no adverse effects in the human skin are expected from the thermal point of view for distances d 2 cm. For comparison: the maximum permissible exposure due to [DIN 0848 91] and [ANSI 1991] is exceeded for d < 17.3 cm and d < 5.5 cm, respectively.
Biological Effects in the cm/mm Wave Range, Part II/III 5.3 Porcine Eye 5.3.1 Experimental setup
43
The experimental setup for the thermal investigation of porcine eyes is similar to that in the previous section. The schematics of the experimental setup are shown in Fig. 49. The eye is inserted into a polystyrene layer and positioned over a warm water bath (Twater = 37C). The experiment was repeated without the warm water bath in order to determine the influence of the physiological temperature range on thermal performance. The power of the radiation source was set to the maximum power of P = 38 mW for all measurements. The ambient temperature during the different measurements was between 21C and 22C, but was constant (0.1C) for each measurement. The relative humidity of the air was between 50% and 60%. Using this setup the maximum temperature rise for three different distances was investigated.
generator
horn antenna
d IR camera eye polystyrene warm water bath (optional)
Fig. 49: Block diagram of the experimental setup. 5.3.2 Measurement results Fig. 50 illustrates a temperature rise of up to 1.9C for a distance of d = 1.1 cm. With increasing distance the temperature rise decreases strongly. For d > 10 cm no significant temperature rise occurred. The curves for the eye with and without water bath show a similar behavior. The warm water did not affect the thermal performance of the temperature dynamic. Fig. 51 shows IR images of the porcine eye without warm water bath: (A) before irradiation, (B) after seven
44
minutes of irradiation. (C) and (D) show subtraction images of (A) and (B) and demonstrate the local heating effect.
1,5 d
T [C]
0,5
Maximum temperature change in a porcine eye (P = 38 mW) eye (without warm water bath) eye (with warm water bath)
0 10
20
30
40 50 Distance d [mm]
60
70
80
Fig. 50: Temperature changes in a porcine eye caused by RF irradiation.

T [C] 22.5 20.5 18.5 16.5 14.5 (A) before irradiation T [C] 22.5 20.5 18.5 16.5 14.5 (B) after 7 minutes of irradiation
T [C]
4.0 2.0 0.0 -2.0 -4.0 (C) subtraction image (D) subtraction image
T [C]
2.87 1.88 0.87 -0.12 -1.13
Fig. 51: Temperature field (A) before and (B) after seven minutes of irradiation. (C) Subtraction image of (B) and (A). (D) Subtraction image with reduced temperature range.
45
OFF 17,5
ON
OFF
d = 11 mm
17
Temperature [C]
16,5 16 15,5 15
9 12 Time [min]
15
18
Fig. 52: Time course of cursor temperature for d = 11 mm and P = 38 mW. Fig. 52 shows the time plot of the cursor temperature during exposure of the eye without warm water bath. 5.3.3 Discussion It is well known that microwave radiation can cause injury to the eye. In the past several experimental investigations have been carried out to determine time and power thresholds for cataractogenesis [Rosenthal 1976]. Our measurements show no significant temperature rise for distances d between radiation source and eye greater than 10 cm. Therefore, this distances are uncritical. Distances d < 10 cm need further investigation from the biological point of view. The maximum permissible exposure due to [DIN 0848 91] and [ANSI 1991] is exceeded for d < 17.3 cm and d < 5.5 cm, respectively. 5.4 Simulation of Thermal Effects 5.4.1 Mathematical Model of Bio-Heat-Transfer In 1948 H. Pennes proposed a mathematical model for heat transfer processes in blood perfused tissue. Although more complex models for heat transfer processes have been developed Pennes approach has been refined and is still being used today [Cho 1992]. Pennes model describes the effect of blood flow on tissue temperature on a continuum basis. Therefore a heat source/sink term is introduced in the heat equation:
46
T = ( T ) + MR + SAR + b cb (Ta T ) . t (9) The parameters of this equation are: mass density of tissue, cp specific heat capacity of tissue, thermal conductivity, MR heat generation rate according to metabolic processes ( = metabolic rate), perfusion rate, b mass density of blood, cb specific heat capacity of tissue, Ta arterial temperature, and SAR the specific absorption rate. c p
On the surface of the body Cauchy boundary conditions are applied in order to account for the heat loss to the environment: & n T q = ( T Tu ) , (10)
& with the heat transfer coefficient , the heat flow q , the ambient temperature Tu, and the outward unit normal vector n. The heat transfer coefficient contains the four heat loss mechanisms: radiation, convection, conduction, and evaporation. The differential equation with initial and boundary conditions is solved using Finite-Element method (FEM) [Gustrau 1997].
5.4.2 Layered Model of Skin Fig. 53 shows a Finite-Element model for the one dimensional analysis of the heat transfer processes in the skin. The thickness of the layers is chosen according to Table 3. The specific absorption rate SAR displayed in Fig. 38 is scaled to an incident power density of S = 10 mW/cm (according to [ANSI 1991]) and applied to the thermal model. Table 6 shows the thermal parameters of the three different tissues. The high perfusion rate of muscle identifies the isothermal core region of the body. The additional parameters are: Ta = 36C, cb = 3350 J/(kgK), and b = 1000 kg/m. Thermoregulatory effector mechanisms, i.e. changes in physiological parameters like perfusion rate caused by the thermal impact, are not considered.
muscle fat dermis epidermis
y x z
47
Fig. 53: Layered Finite-Element model of the human skin for the one-dimensional numerical investigation. Epidermis cp [J/(kgK)] [W/(mK)] [kg/m] MR [W/m] [m/(sm)] [W/(mK)] 3350 0.5 1000 0 0 12 dermis 3350 0.5 1000 200 0.0002425 0 fat 3350 0.5 1000 200 0.0002425 0 muscle 3350 0.5 1000 200 10000 0
Table 6: Thermal properties of skin tissue. 5.4.3 Results Fig. 54 shows the time plot of the superficial temperature change in the layered model.
0,9 0,8 0,7 0,6

T [C]
0,5 0,4 0,3 0,2 0,1 0 -0,1 -5 0 5
Skin temperature change without fat layer with 1 cm fat layer
Start of irradiation
10 t [min]
15
20
25
Fig. 54: Time plot of the superficial skin temperature change for the layered model of skin. Simulation was carried out with and without fat layer.
48
For comparison the fat layer has been omitted in a second calculation. It turns out that the isolation of the fat layer influences both, the amplitude and the dynamic of the heat balance process. For the first calculation a steady-state temperature change of T = 0.84C is achieved after 15 minutes of exposure, for the second calculation a temperature rise of T = 0.4C is achieved after 3 minutes. 5.4.4 Discussion This simulation results for a power density of S = 10 mW/cm have to be compared with the measurement of skin temperature changes for a distance d = 5.5 cm. The interpolation of the measurement data (see Fig. 46) shows a temperature change of T 0.7C and the thermal simulations predict a value between 0.4C and 0.84C depending on the consideration of a fat layer. In view of the natural span of measurement data due to inter- and intraindividual differences of the volunteers and in view of the simplicity of the model which considers only onedimensional heat transfer the results of the thermal measurements and simulations provide consistent results for the assessment of thermal effects of mm-wave irradiation.
6 Summary
This investigation, which was carried out on behalf of the FGF at IMST, quantified the specific absorption rate in skin and eye tissue and the resultant thermal effects of cm/mm wave radiation. Additionally, it contains the determination of the dielectric parameters of human tissue in the frequency range from 200 MHz up to 20 GHz and from 75 GHz up to 100 GHz. The tissues under investigation included muscle, fat, skin (dermis), cornea, retina, lens, sclera, vitreous body, and liquid from camera anterior. First of all a literature study was made concerning the dielectric parameters of human tissue. In a next step the unknown dielectric parameters were measured using porcine tissue with material measurement systems in the frequency range of interest. The measured material parameters and data from human tissue reported in literature showed a good agreement. The measured material parameters and additional data from literature were applied to the analysis of field strengths in the human eye and skin. The exposure that was realized was a linearly polarized plane wave. For the human skin an analytical method was used for the calculation of the electromagnetic field in a layered model of skin in the frequency range 3 100 GHz. Besides epidermis and dermis, fat and muscle tissue were distinguished. The highest values of the specific absorption rate were found in the outer layers of the human skin because of the strong increase of the losses in the human skin as a function of frequency. For 77 GHz a decrease of the SAR of nearly 10 decades was observed from the epidermis layer to the fat layer. For a power density of 1 mW/cm2 the strong losses led to a maximum SAR of 34 W/kg at 100 GHz and 27 W/kg at 77 GHz. The 10g-averaged SAR-values were about 0.27 W/kg and therefore more than 7 times below the maximum permissible SAR value of 2 W/kg [DIN 0848 91]. Due to the strong absorption of the electromagnetic fields in superficial tissue the averaged values for higher frequencies were determined mainly by the parameters of the first layer of the model including the reflection coefficient of the interface air-skin.
49
The investigation of the human skin has been carried out using material parameters from the parametric model of Gabriel [Gabriel96b] because it provides a closed form over the whole frequency range of interest. Of course, the choice of dielectric properties influences the computational results. This effect can be estimated by the transmission coefficient of the interface air-skin. The simulations show, that the highest SAR value occurs at the surface of the skin for a frequency of 100 GHz. For this frequency increasing the complex permittivity by 50 % leads to a 11 % higher maximum SAR value. Decreasing the complex permittivity by 50 % leads to a 16 % lower maximum SAR value. The human eye was modeled as a rotary body with the tissues mentioned above. The calculation was carried out using the FDTD method at a frequency of 77 GHz and plane wave exposure with a power density of 1 mW/cm2. At this frequency the calculations of the three dimensional model of the human eye and the layered skin model showed similar results. In both cases the absorption was dominant in superficial tissues and the 1g-averaged SAR values were comparable: for the layered skin model the 1g-averaged SAR value was 0.6 W/kg and for the model of the human eye this value amounted to 0.6588 W/kg. The same applied to the maximum local SAR values: for the layered model of skin the maximum local SAR value was 27 W/kg and for the model of the human eye the local SAR value amounted to 45 W/kg in cornea tissue. This allows the conclusion that at high frequencies f > 30 GHz it is not necessary to model the whole eye in great detail but it is sufficient to consider its surface region. For the investigation of thermal effects an infrared thermography system was used. A radiation source with a horn antenna was aimed at the skin of the forearm of two volunteers and at a porcine eye in vitro. For this radiation source a power density of 1 mW/cm2 [DIN 0848 91] corresponds to a distance of 17.3 cm and a power density of 10 mW/cm2 to a distance of 5.5 cm. The measured temperature changes in the skin showed temperature changes up to 2.2C for a distance of d = 18 mm between aperture and skin and a strong decay in temperature rise for greater distances. The temperature changes were beneath the warmth sensation thresholds of the volunteers. The measured temperature changes in the eye showed temperature changes up to 1.9C for a distance of d = 11 mm between aperture and eye and a strong decay in temperature rise for greater distances. For d > 10 cm no significant temperature change occurred for skin and eye. Finally the calculated distribution of the specific absorption rate was introduced in a thermal model. The thermal calculations were based on the Finite Element Method. Two skin models were analyzed: the layered model presented before and a model without fat layer. The optional isolating fat layer influenced the dynamic and amplitude of the heat balance process. The simulations showed temperature changes in the same order of magnitude as the measurement: for a power density of 10 mW/cm2 [ANSI 1991] the temperature rise was about 0.7C in the measurement and between 0.4C and 0.84C in the simulation. The investigation of thermal effects showed a temperature rise in the order of 0.7C for a power density of 10 mW/cm2 [ANSI 1991]. For the more restrictive German and European standards a permissible power density of 1 mW/cm2 showed temperature rises in the order of 0.1C. In real world applications the typical power density is up to 60 W/cm2 (e.g. automotive radar), which is far below the amplitudes mentioned above.
50
7 References
[Aschoff 1971] J. Aschoff, B. Gnther, K. Kramer: Energiehaushalt und Temperaturregulation, in Gauer, Kramer, Jung (Hrsg.), Physiologies des Menschen, Vol. 2, pp. 43-116, Urban und Schwarzenberg, Mnchen, 1971. ANSI C95.1: IEEE Standard for Safety Levels with Respect to Human Exposure to Radio Frequency Electromagnetic Fields, 3 kHz to 300 GHz, Inst. of Electrical and Electronics Engineers, Inc., 1991. C. A. Balanis: Advanced Engineering Electromagnetics, Chapter 5, New York, John Wiley & Sons, pp. 180-253, 1989. J.-P. Berenger: A Perfectly Matched Layer for the Absorption of Electromagnetic Waves, J. Comput. Phys., Vol. 114, pp. 185-200, 1994.
[ANSI 1991]
[Balanis 1989] [Berenger 1994]
[BImSchV 1996] Sechsundzwanzigste Verordnung zur Durchfhrung des Bundes-Immissionsschutzgesetzes: Verordnung ber elektromagnetische Felder - 26. BImSchV, 1996. [Bohrmann 1993] S. Bohrmann, R. Pitka, H. Stcker, G. Terlecki: Physik fr Ingenieure, Verlag Harri Deutsch, Frankfurt am Main, 1993. [Cho 1992] Y.I. Cho, J.P. Hartnett, T.F. Irvine (Eds): Bioengineering Heat Transfer, Advances in Heat Transfer, Vol. 22, Academic Press, San Diego, 1992.
[Dimbylow 1988] P. J. Dimbylow: The calc1ulation of induced currents and absorbed power in a realistic, heterogeneous model of the lower leg for applied electric fields from 60 Hz to 30 MHz, Phys. Med. Biol., Vol. 33, pp. 1453-1468, 1988. [Dimbylow 1991] P. J. Dimbylow and O. P. Gandhi: Finite-difference time-domain calculations of SAR in a realistic heterogeneous model of the head for plane-wave exposure form 600 MHz to 3 GHz, Phys. Med. Biol., Vol. 36, pp. 1075-1089, 1991. [DIN 0848 91] [Duck 1990] [Edrich 1976] Deutsche Norm (Entwurf): Sicherheit in elektromagnetischen Feldern, Schutz von Personen im Frequenzbereich von 30 kHz bis 300 GHz, DIN VDE 0848 Teil 2, Oktober 1991. F.A. Duck: Physical properties of tissue, Academic Press, San Diego, 1990. J. Edrich and P. C. Hardee: Complex permittivity and penetration depth of muscle and fat tissues between 40 and 90 GHz. IEEE Trans. Microwave Theory Tech., MTT-24, pp. 273275, 1976. European Prestandard ENV 50166-2: Human exposure to electromagnetic fields - High frequency (10 kHz to 300 GHz), CENELEC, January 1995. Federal Communications Commission: Report and order: Guidelines for evaluating the environmental effects of radiofrequency radiaton, Tech. Rep. FCC 96-326, FCC, 1996. R. Flindt: Biologie in Zahlen, Fischer Verlag, Stuttgart, 4. ed., 1995. K. Fleischhauer, J. Staubesand and W. Zenker: Benninghoff Anatomie 3, Chapter 23 and 27, Urban+Schwarzenberg, Mnchen, pp. 485-571, 1985. S. Gabriel and R. W. Lau and C. Gabriel: The dielectric properties of biological tissue: II. Measurements in the frequency range 10 Hz to 20 GHz. Phys. Med. Biol., Vol. 41, pp. 2251-2269, 1996. S. Gabriel and R. W. Lau and C. Gabriel: The dielectric properties of biological tissue: III. Parametric models for the dielectric spectrum of tissues. Phys. Med. Biol., Vol. 41, pp. 2271-2293, 1996.
[ENV 50166] [FCC 1996]
[Flindt 95] [FSZ 1985] [Gabriel 96a]
[Gabriel 96b]
51
[Gabriel 96c] S. Gabriel and R. W. Lau and C. Gabriel: The dielectric properties of biological tissue: I. Literature survey. Phys. Med. Biol., Vol. 41, pp. 2231-2249, 1996. [Geigy 1985] [Gustrau 1997] Ciba Geigy: Wissenschaftliche Tabellen, Geigy, 8. ed., 1985. Frank Gustrau: Ein Beitrag zur numerischen Berechnung der thermischen Belastung von Patienten durch hochfrequente elektromagnetische Felder whrend Magnet-ResonanzTomographie-Untersuchungen, Dissertation, Fakultt fr Elektrotechnik, Ruhr-Universitt Bochum, Bochum, 1997. IRPA Guidelines: Guidelines on Limits of Exposure to Radiofrequency Electromagnetic Fields in the Frequency Range from 100 kHz to 300 GHz, Health Physics, Vol. 54, No. 1, pp. 115-123, 1988. A. S. Duchene and J. R. A. Lakev: The IRPA guidelines on protection against non-ionizing radiation, The collected publication of the IRPA, Chapter 5, New York: Pergamon Press, pp. 72-82, 1991. H. Lippert: Lehrbuch Anatomie, Chapter 1 and 6, Urban+Schwarzenberg, Mnchen, pp. 85541, 1990.
[IRPA 1988]
[IRPA 1991]
[Lippert 1990]
[Rosenthal 1976] S.W. Rosenthal, et al.: Effects of 35 and 107 GHz CW Microwaves on the Rabbit Eye, pp. 110-128, 1976. [Schwan 57] [Schwan 80] [SSK 1993] [Werner 1984] [WHO 1993] [Yee 1966] H. P. Schwan: Electric properties measured with alternating currents; body tissues, Handbook of biological data, edited by W. S. Spector, W B Saunders Co, Philadelphia, 1957. H. P. Schwan and K. R. Foster: RF-field interactions with biological systems: Electric properties and biophysical mechanisms, Proceedings of the IEEE, pp. 104-113, 1980. Bundesamt fr Strahlenschutz: Schutz vor elektromagnetischen Feldern - Empfehlungen und Stellungnahmen der Strahlenschutzkommission, SSK/6-93/EMF, 1993. J. Werner, Regelung der menschlichen Krpertemperatur, W. de Gruyter, Berlin, 1984. WHO/IRPA task group on electromagnetic fields: Environmental Health Criteria 137, Electromagnetic Fields (300 Hz to 300 GHz), WHO, Geneva, 1992. K. S. Yee: Numerical Solution of Initial Boundary Value Problems Involving Maxwells Equations in Isotropic Media, IEEE Trans. Antennas Propagat., AP-14, pp. 302-307, 1966.
52
8 Appendix
8.1 Comparison of Measured Dielectric Properties and Data from Literature Fig. 55 57 show a comparison between measured dielectric properties of skin, fat, and muscle tissue as well as data from literature [Gabriel 96b]. The temperature of the probe was T = 37C during the measurement. Taking into account the natural variability of dielectric properties of tissue the measured values correspond well with data from Gabriel in the frequency range from 200 MHz 20 GHz. In the higher frequency range from 75 GHz 100 GHz the differences increase, especially for fat and muscle tissue. The deviation of the conductivity of fat tissue from the parametric model may result from the variable water content of fat tissue [Duck 1990]. However, it has to be noted that in the high frequency range (f > 40 GHz) the parametric model of Gabriel is based only on very few experimental data.
100 80 60 40 20 0 0,1
Tissue: skin rel. permittivity (Measurement) conductivity [S/m] (Measurement) rel. permittivity (Gabriel) conductivity [S/m] (Gabriel)
1 Frequency [GHz]
10
100
Fig. 55: Comparison of measured dielectric properties of skin (T = 37C) and data from literature [Gabriel 96b].
53
60 Tissue: fat rel. permittivity (Measurement) conductivity [S/m] (Measurement) conductivity [S/m] (Gabriel) rel. permittivity (Gabriel)
40
20
0 0,1
1 Frequency [GHz]
10
100
Fig. 56: Comparison of measured dielectric properties of fat (T = 37C) and data from literature [Gabriel 96b].
120 100 80 60 40 20 0 0,1
Tissue: muscle rel. permittivity (Measurement) conductivity [S/m] (Measurement) rel. permittivity (Gabriel) conductivity [S/m] (Gabriel)
1 Frequency [GHz]
10
100
Fig. 57: Comparison of measured dielectric properties of muscle (T = 37C) and data from literature [Gabriel 96b].
54
IMST GmbH, Institut fr Mobil- und Satellitenfunktechnik Carl-Friedrich-Gau-Strae 2 D-47475 Kamp-Lintfort Tel. +49- 2842-981 370 Fax +49- 2842-981 398 email: [email protected]

1999.01. Final Report Biological Effects in The MM Wave Range

Uploaded by

Copyright:

Available Formats

1999.01. Final Report Biological Effects in The MM Wave Range

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1999.01. Final Report Biological Effects in The MM Wave Range

Uploaded by

Copyright:

Available Formats

Final Report

14. January 1999

Biological Effects in the cm/mm Wave Range, Part II/III

3.2.1 3.2.2 3.3

4.1.1 4.1.2 4.2

4.2.1 4.2.2 4.3

Biological Effects in the cm/mm Wave Range, Part II/III

5.3.1 5.3.2 5.3.3 5.4

5.4.1 5.4.2 5.4.3 5.4.4 6 7 8

Biological Effects in the cm/mm Wave Range, Part II/III

2 Radiofrequency Field Exposure Standards

Biological Effects in the cm/mm Wave Range, Part II/III

Biological Effects in the cm/mm Wave Range, Part II/III

Biological Effects in the cm/mm Wave Range, Part II/III

SAR limit [W/kg] 2.0

[DIN 0848 91]

Biological Effects in the cm/mm Wave Range, Part II/III

Eeff [V/m] Heff [A/m]

[DIN 0848 91] [ENV 50166]

f / 0.15 [ANSI 1991] 100 10 [IRPA 1991]

3 Determination of the Relevant Material Parameters

Biological Effects in the cm/mm Wave Range, Part II/III

Biological Effects in the cm/mm Wave Range, Part II/III

Tissue: skin (wet) rel. permittivity conductivity [S/m]

Fig. 2: Dielectric properties of skin (wet) as a function of frequency.

Tissue: skin (dry) rel. permittivity conductivity [S/m]

Fig. 3: Dielectric properties of skin (dry) as a function of frequency.

Biological Effects in the cm/mm Wave Range, Part II/III

Tissue: fat (average infiltrated) rel. permittivity conductivity [S/m]

Fig. 4: Dielectric properties of fat (average infiltrated) as a function of frequency.

Tissue: fat (not infiltrated) rel. permittivity conductivity [S/m]

Fig. 5: Dielectric properties of fat (not infiltrated) as a function of frequency.

Biological Effects in the cm/mm Wave Range, Part II/III

10 7 Tissue: muscle 10 6 rel. permittivity 10 conductivity [S/m] 5 10 4 10 3 10 2 10 1 10 0 10 -1 10 -2 10 1 2 3 4 5 6 7 8 9 10 11 10 10 10 10 10 10 10 10 10 10 10 Frequency [Hz]

Biological Effects in the cm/mm Wave Range, Part II/III

tissue sample (liquid)

Biological Effects in the cm/mm Wave Range, Part II/III

Tissue: skin, T=27C conductivity [S/m] rel. permittivity

Biological Effects in the cm/mm Wave Range, Part II/III

Tissue: skin, T=37C conductivity [S/m] rel. permittivity

Tissue: fat, T=27C conductivity [S/m] rel. permittivity

Biological Effects in the cm/mm Wave Range, Part II/III

Tissue: fat, T=37C conductivity [S/m] rel. permittivity

Biological Effects in the cm/mm Wave Range, Part II/III

60 Tissue: vitreous body, T=27C conductivity [S/m] rel. permittivity

Biological Effects in the cm/mm Wave Range, Part II/III

40 Tissue: cornea, T=27C conductivity [S/m] rel. permittivity

80 Tissue: lens, T=27C conductivity [S/m] rel. permittivity

Biological Effects in the cm/mm Wave Range, Part II/III

40 Tissue: retina, T=27C conductivity [S/m] rel. permittivity

100 80 60 40 20 0 75 Tissue: sclera, T=27C conductivity [S/m] rel. permittivity

Biological Effects in the cm/mm Wave Range, Part II/III

Biological Effects in the cm/mm Wave Range, Part II/III

Tissue: skin, T=27C rel. permittivity conductivity [S/m]

Biological Effects in the cm/mm Wave Range, Part II/III

Tissue: skin, T=37C rel. permittivity conductivity [S/m]

80 Tissue: fat, T=27C rel. permittivity conductivity [S/m]

Biological Effects in the cm/mm Wave Range, Part II/III

80 Tissue: fat, T=37C rel. permittivity conductivity [S/m]