36
Electromagnetic Fields in Biological Systems
by assuming a plane wave incident on a human body. Specifically, FDTD calculations of
SAR distribution in anatomical models of children and adults exposed to uniform RF
field of both vertical and horizontal polarizations have been conducted at frequencies in
the cellular mobile communication band (900–3000 MHz).
In a study by Piuzzi et al. (2011), five different human body models were developed; all
were obtained by appropriately scaling the segmented VH data set along the three spatial directions while maintaining a 2-mm resolution. They include an adult model and
two 13-year-old and two 7-year-old child models. The two different child phantoms are
a child-scaled and a childlike model. The former was obtained by uniformly scaling the
VH phantom, whereas the latter was derived by applying a nonuniform scaling, which
yielded a more realistic gross anatomy.
The original VH phantom, having a height of 188 cm and a mass of 103 kg, was
scaled in its axial cross sections, 0.91 scaling factor for shoulder-to-shoulder (x-direction) scaling and 0.83 for front-to-back (y-direction) scaling, to arrive at a final mass
of about 80 kg, approximating the weight of a standard-mass adult male (CDC 2009).
The two child-scaled models were obtained using appropriate scaling factors along
the three directions of the VH model to reproduce a 50% mass-for-age and staturefor-age for this age group (WHO 2000). The final 13-year-old child model (0.71 scaling factor in the axial sections and 0.83 in the vertical direction) has a height of
156 cm and mass of 44 kg, whereas the 7-year-old child model (0.59 scaling factor in
the axial sections and 0.67 in the vertical direction) is 125 cm tall with a body mass
of 24 kg.
The scaling procedure, however, does not produce anatomically correct child models,
particularly with reference to the proportion between the head and the trunk. Indeed,
it is well-known that proportionally children tend to have larger heads than adults. For
this reason, two childlike models have also been derived by applying different scaling
factors to different body segments. In this case, the body is divided into four major segments: (1) trunk (feet to neck), (2) lower face (chin to the center of the mouth), (3) upper
face (center of the mouth to the center of the eyes), and (4) head (above the center of
the eyes). The scaling factors for the trunk were kept the same as those adopted for the
child-scaled models, whereas scaling factors for the face and head were based on anthropometric data (Leslie and Farkas 1981).
The adult model and the two scaled child models are shown in Figure 1.13. It can be
seen that the scaled child models appear different from normal children because of their
rather small heads. Anatomical realism is accomplished in the childlike model shown
in Figure 1.14, which gives a comparison between the 7-year-old child-scaled and childlike models. Note that although childlike models have realistic body shapes, they do not
ensure that the internal organs and other structures are accurately reproduced.
Figure 1.15 presents the computed whole-body average SAR (SAR WB ) for the adult
and the two scaled child models for two different polarizations using the FDTD algorithm. Note that child-scaled and childlike models of a given age essentially yield the
same SAR WB values. It can be seen that for a given incident power density (10 W/m 2 ),
SAR WB is higher for smaller body model sizes. Also, while SAR WB decreases as a function
of frequency for all body sizes, its variation with frequency differs for horizontal and
vertical polarizations; the horizontal polarization produces a higher absorption above