242
Electromagnetic Fields in Biological Systems
percentile value of the volume-averaged in situ electric field in the nerve tissue was more
stable than that of the maximum value for different averaging volume.
The draft ICNIRP guidelines can use the in situ electric field as a measure of the basic
restriction (ICNIRP 2009). There is a growing consensus that the in situ electric fields
should be used as basic restriction instead of induced current densities. This change
requires reevaluation of the rationale for the establishment of basic restrictions from
the standpoint of both numerical dosimetry and human health protection. ICNIRP
issued new exposure guidelines for time-varying electric and magnetic fields from 1 Hz
to 100 kHz on November 10, 2010 (ICNIRP 2010). In these new exposure guidelines,
the physical quantity used to specify the basic restriction is the internal electric field
strength. From the results obtained from high-resolution calculations of induced electric field with voxel sizes below 4 mm, it recommends that the determination of induced
electric field is the vector summation of the average of the electric field in a small volume
2 × 2 × 2 mm 3 . For specific tissue, the 99th percentile value of the electric field is the
relevant value to be compared with the basic restriction.
4.5.4 Validation of Methods
Numerical dosimetry in bioelectromagnetics has been developed gradually from using
simple human models during the 1970s through about 1-cm anatomical cross-section
human models in the 1980s to 1- to 10-mm MRI-based and CAD-meshed human
models from the end of 1990s until now.
The early human model represented the body by simple objects such as cylinders,
spheres, and ellipsoids, and the conductivity of the body was assumed to be homogeneous. The calculations of induced current densities and the electric fields were carried
out for these simple human models exposed to a uniform ELF electric or magnetic fields
by simple analytical methods. The effects of nonuniform ELF magnetic field sources such
as appliances have been also studied. The point is the accuracies of calculation methods
and models used in numerical dosimetry. The accuracy of the calculation results has
been tested in several ways. Historically, comparisons between the calculated results
of the numerical method and measured values in mannequin-like human models were
made. The human body model was assumed to be a homogeneous structure, indicating that the conductivity values within models are constant although experimentally
obtained conductivity values have considerable uncertainty. The second method compares the solutions in simple models obtained by different methods. The other compares
the results of the numerical method to measured values in experimental models using
animal and other saline-filled models. From these tests, it can be said that the calculation methods seem to be accurate. On the other hand, the data on biological values such
as tissue conductivity are still uncertain in the ELF region. This leads to uncertainty in
the assignment of tissue conductivity to cubes or voxels.
It has been led to the method that the Maxwell’s equations are solved in term of individual cubes or voxels with assigned conductivities in harmonizing with development of
computational algorithms and high-speed computers. These methods have advantages
in both modeling a complex shape, anatomical, MRI-based body taking the regional
variations in conductivity into consideration and the estimation of the electric fields and