6.5.3 Determining Dosimetric Quantities
Dosimetric quantities generally cannot be measured directly within the tissues of a person undergoing ordinary RF exposure. They are therefore determined using a combination of experimental measurement, numerical calculation, and, where appropriate, measurements of the external field. The chosen method must represent the operating conditions and geometry relevant to the assessment, including frequency, source output, modulation, duty cycle, polarization, separation, orientation, posture, and the required spatial and temporal averaging.
Experimental dosimetry uses physical models, known as phantoms, that reproduce relevant electromagnetic properties of the body. A phantom may contain a tissue-equivalent liquid, gel, or solid material with specified permittivity and conductivity. In standardized tests of a device used close to the head or body, the device is operated in defined configurations and a calibrated electric-field probe scans the field within the phantom. Local SAR is derived from the measured internal field together with the conductivity and density of the phantom material, and is then averaged over the mass specified by the applicable standard. IEC/IEEE 62209-1528 provides one widely used example of this approach.
Physical phantoms deliberately simplify human anatomy. Standardized shapes and materials allow different laboratories and devices to be compared under repeatable, conservative conditions, while more specialized phantoms can be used to investigate particular exposure geometries. The result nevertheless applies to the test model and configuration rather than predicting the exact absorption in every individual. Probe dimensions and calibration, phantom material properties, source positioning, output-power control, and reflections from the test environment are among the factors that can affect the result.
Computational dosimetry determines internal quantities by solving Maxwell's equations for a numerical representation of the source, the surrounding environment, and the body. Common techniques include the finite-difference time-domain (FDTD) method and the finite element method (FEM). Body models range from simple homogeneous shapes to anatomically detailed voxel models in which individual tissues are assigned frequency-dependent electrical properties. The calculated fields can be used to determine induced electric-field strength, whole-body and localized SAR, specific energy absorption, and absorbed power or energy density. The IEC/IEEE 62704 series specifies procedures for FDTD and FEM calculations of localized SAR, including verification, validation, and uncertainty assessment.
Numerical methods are especially valuable where internal fields cannot be probed without disturbing the exposure, where whole-body or highly non-uniform exposure must be examined, or where many frequencies, postures, source positions, or body models must be compared. Their results depend on the accuracy of the source and body models, tissue-property data, spatial resolution, boundary conditions, and averaging procedure. Fine resolution is particularly important where fields vary rapidly, such as close to an antenna or near the body surface at higher frequencies.
Measurement and calculation are complementary rather than competing methods. Computational models are checked against analytical solutions, standardized benchmarks, or measurements, while measurement systems are calibrated and validated using reference sources. Every assessment also requires an uncertainty analysis. Relevant contributions can include transmitter output, probe calibration and positioning, phantom properties, numerical discretization, anatomical variability, and the spatial or temporal averaging process. Conservative assumptions are commonly used where uncertainty might otherwise lead to an underestimate.
A practical compliance assessment often begins with external quantities. If the measured or calculated exposure is below the applicable reference levels, compliance with the corresponding basic restrictions is demonstrated. If a reference level is exceeded, or if it is unsuitable for a strongly non-uniform near-field or close-body exposure, experimental or computational dosimetry may be used to assess the relevant basic restriction directly. Product and installation standards specify the source configurations, models, averaging rules, and uncertainty requirements appropriate to particular applications.
Whatever method is used, the output of RF dosimetry remains a set of internal physical quantities: fields induced in tissue and power or energy absorbed by the body. These quantities describe the physical conditions from which a biological response might arise; they do not by themselves establish whether a response occurs or whether it is adverse. Determining that relationship is the subject of Chapter 7.