Jin-Yuan Chen

A frequency-dependent finite-difference time-domain formulation for general dispersive media

A weakness of the finite-difference-time-domain (FDTD) method is that dispersion of the dielectric properties of the scattering/absorbing body is often ignored and frequency-independent properties are generally taken. While this is not a disadvantage for CW or narrowband irradiation, the results thus obtained may be highly erroneous for short pulses where ultrawide bandwidths are involved. In some recent publications, procedures based on a convolution integral describing D(t) in terms of E(t) are given for media for which the complex permittivity in *( omega ) may be described by a single-order Debye relaxation equation or a modified version thereof. Procedures are, however, needed for general dispersive media for which in *( omega ) and mu *( omega ) may be expressible in terms of rational functions, or for human tissues for which multiterm Debye relaxation equations must generally be used. The authors describe a new differential equation approach, which can be used for general dispersive media. In this method D(t) in terms of E(t) by means of a differential equation involving E, and their time derivatives. The method is illustrated for several examples. >

Currents Induced in a Human Being for Plane-Wave Exposure Conditions 0-50 MHz and for RF Sealers

The currents induced in a human being were measured for plane-wave exposure conditions 0-50 MHz. These are used to point out that very large SARs may be set up in the wet tissues for the cross section of the ankle for standing human beings exposed to electric fields suggested in the ANSI C95.1-1982 RF safety guide. For these exposure conditions, foot currents on the order of 627 mA are projected for a human adult of height 1.75 m for the frequency band 3-30 MHz with the value increasing to 780 mA for 40 MHz. The corresponding ankle-section SARs are 182 and 243 W/kg. Using electromagnetic scaling concepts, SARs as high as 371 and 534 W/kg are projected for ten- and five-year old children, respectively, for f = 50.7 and 62.5 MHz, E = 61.4 V/m (1 mW/cm2) recommended by the ANSI C95.1-1982 RF safety guide. The paper also gives currents induced by two sealer models and three industrial RF sealers for a human being under grounded and insulated conditions. Though smaller than those for plane-wave irradiation conditions, these currents may be substantial for high leakage fields that are quite typical.

Electromagnetic absorption in the human head from experimental 6-GHz handheld transceivers

We have used a new millimeter-resolution MRI-based model of the human body to calculate the electromagnetic absorption in the head and neck for three experimental Yagi antennas suggested for handheld transceivers of a proposed 6-GHz personal communication network (PCN) system. The SAR distributions are obtained with a resolution of 1.974/spl times/1.974/spl times/1.5 mm for transceivers that are held against the ears and tilted forward by 33/spl deg/. The finite-difference time-domain technique is used to calculate the EM fields and SARs for the transceiver, antenna, and head and neck coupled region that is divided into 158/spl times/84/spl times/188 or nearly 2.5 million cells. The highlights of the numerical calculations are verified by means of a head-shaped experimental model made of tissue-equivalent materials simulating the electrical properties (/spl epsiv//sub r/,/spl sigma/) of the skull, brain, muscle, eyes, and ears developed for use at 6 GHz. Because of the proximity to the antenna, the highest SARs are obtained for the upper part of the ear. For a planned radiated power of 0.6 W, the peak SARs averaged over any 1 g of tissue defined as a tissue volume in the shape of a cube are on the order of 0.5-1.0 W/kg for two of the proposed antennas and considerably higher (2.06 W/kg) for the third antenna using a narrower off-axis reflector. Low SARs for the first two antennas are likely due to the shielding provided by the relatively wider strip reflector plates used for these antennas.

Specific absorption rates and induced current distributions in an anatomically based human model for plane-wave exposures

We have previously reported local, layer-averaged, and whole-body-averaged specific absorption rates and induced currents for a 5,628-cell anatomically based model of a human for plane-wave exposures 20-100 MHz (Chen and Gandhi 1989). Using a higher resolution, 45,024-cell model of the human body, calculations have now been extended to 915 MHz using the finite-difference time-domain method. Because of the higher resolution of the model, it has been possible to calculate specific absorption rates for various organs (brain, eyes, heart, lungs, liver, kidneys, and intestines) and for various parts of the body (head, neck, torso, legs, and arms) as a function of frequency in the band 100-915 MHz. Consistent with some of the experimental data in the literature, the highest part-body-averaged specific absorption rate for the head and neck region (as well as for the eyes and brain) occurs at 200 MHz for the isolated condition and at 150 MHz for the grounded condition of the model. Also observed is an increasing specific absorption rate for the eyes for frequencies above 350 MHz due to the superficial nature of power deposition at increasing frequencies.

RF currents induced in an anatomically-based model of a human for plane-wave exposures (20-100 MHz)

The three-dimensional finite-difference time-domain (FDTD) method has been used to calculate local, layer-averaged and whole-body averaged specific absorption rates (SARs) and internal radiofrequency (RF) currents in a 5628-cell, anatomically-based model of a human for plane-wave exposures from 20-100 MHz. The conditions of exposure of the human considered are: 1) isolated from ground, and 2) feet in contact with ground. Also considered are various separations of the model from ground and the use of insulating, rubber-soled footwear close to the grounded resonance frequency of 45 MHz. The calculated results are in agreement with the experimental data of Hill and others. While the existence of large foot currents has been known previously, substantial RF currents (600-800 mA) induced over much of the body are obtained for E-polarized fields suggested in the 1982 ANSI RF safety guideline.

Currents induced in an anatomically based model of a human for exposure to vertically polarized electromagnetic pulses

The finite-difference time-domain (FDTD) technique is used to calculate the internal fields and the induced current densities in anatomically based models of a human using 5628 or 45024 cubical cells of dimensions 2.62 and 1.31 cm, respectively. A layer of dielectric constant of epsilon /sub r/=4.2 and having a thickness of 2.62 cm is assumed under the feet to simulate a human wearing rubber-soled shoes. The total induced currents for the various sections of the body and the specific absorptions for several organs are given for two representative electromagnetic pulses. The calculated results for the induced currents are in excellent agreement with the data measured for a human subject. The FDTD method is ideally suited for exact representation of the pulse shapes and offers numerical efficiency to allow detailed modeling of the human body and the various organs. >

Electromagnetic deposition in an anatomically based model of man for leakage fields of a parallel-plate dielectric heater

An anatomically based 5628-cell model of a human being was used to calculate local, layer-averaged, and whole-body-averaged specific absorption rates and internal RF currents at 27.12 and 40.68 MHz for spatially variable electromagnetic fields of a parallel-plate applicator representative of RF dielectric heaters used in industry. The conditions of exposure of the man model considered are: isolated from ground, shoe-wearing condition, feet in contact with ground, and an additional grounded top plate 13.1 cm above the head to simulate screen rooms that are occasionally used for RF heaters. Since peak E fields as high as 1000-2700 V/m have been measured at locations typically occupied by the operator, significant internal RF currents on the order of 0.5-2.3 A are projected for the operators. Measurements of the foot currents at 27.12 and 40.68 MHz for a human subject are in reasonable agreement with the calculated values for the various conditions of exposure. >

SAR and induced current distributions for operator exposure to RF dielectric sealers

The finite-difference time-domain method is used to calculate local, layer-averaged, and whole-body averaged specific absorption rates (SARs) and induced current distributions in a 16-tissue, anatomically based, 5628-cell model of a human to assess operator exposure to RF sealers. Industrially relevant shapes and dimensions of commonly used RF dielectric heaters using parallel-plate and bar-type electrodes are considered. Realistic postures of the human operators are used for the calculations, including extending arms to simulate working conditions or an operator sitting on a wooden or metallic stool. Due to the high-intensity leakage fields in proximity to the RF applicators, some of the highest induced currents and SARs are calculated for the hands and the ankles and, in the sitting position, the knees. It is concluded that steps should therefore be taken either to reduce the leakage fields or shield the hands and the knees if it is necessary for them to be in high leakage field regions. >

Absorption and Distribution Patterns of RF Fields

Over the past five to ten years, magnetic resonance imaging (MRI) has become an important tool for medical diagnostic applications. There are several aspects of the intense magnetic fields that require close scrutiny from the point of view of patient safety. These include larger static magnetic fields, large spatial gradients of the magnetic fields, rapidly switched magnetic fields, and higher and higher frequency RF magnetic fields. In this report, we will examine the absorption and distribution patterns of radio-frequency (RF) magnetic fields, which in emerging techniques may be of frequencies as high as 200 MHz. Highly simplified homogeneous spherical, cylindrical, and disk models have in the past been used to obtain power depositions due to RF magnetic fields.12 Because only homogeneous models have been used, these models, naturally, are incapable of providing information on distributions of RF absorption (specific absorption rates or SARs) . For RF dosimetry, we have developed an anatomically based model of the human body‘ and a new numerically efficient method called the impedance method that we and our colleagues have used to obtain S A R distributions for frequencies up to 63 In the impedance method, we represent the body or the parts thereof by a threedimensional network of impedances, each of which represents a certain subregion of the body typically about 1-1.5 cm in length. Durney et al.’ have recently generalized the impedance network formulation using the complete set of Maxwell’s equations, which should therefore be usable at any frequency. This generalization involves inclusion of inductances in addition to the resistance-capacitance equivalent impedances that were used previously.M

Induced current and SAR distributions for a worker model exposed to an RF dielectric heater under simulated workplace conditions.

We have used the finite-difference time-domain method to calculate the distributions of absorbed energy for a 1.34 x 1.34 x 1.4 cm resolution anatomically based model of the human body for exposure to leakage electromagnetic fields of a radiofrequency dielectric heater operating at 40.68 MHz. To simulate workplace conditions, the dielectric heater is assumed to be placed in a screen room and different operator postures, such as standing or sitting on a wooden or metal stool with hands on sides or extended toward the radiofrequency heater, are considered. To obviate the problem of having to model a fairly large volume of the screen room of assumed dimensions 2.13 x 3.05 x 2.13 m, we have used a uniform finer grid of points for the finite-difference time-domain method for the closely coupled region consisting of the front region of the heater and the human model, while a newly developed expanding-grid finite-difference time-domain formulation is used elsewhere. This results in a saving of both the memory and computation times by almost a factor of four. The average rates of energy absorption are given for the whole body, selected parts of the body, and various organs. As expected, the foot currents and the rates of energy absorption are higher with the screen room for sitting postures where the upper parts of the body are in higher electromagnetic fields, and for hands extended toward the heater.