286
Electromagnetic Fields in Biological Systems
magnitude and cannot provide any phase information about the measured field. Future
applications will require development of more advanced scanners providing not only
amplitude but also phase information (Su et al. 2006).
5.3.3 Uncertainty Assessment
Basic information regarding uncertainty evaluation in experimental dosimetry can
be found in (Iskra, McKenzie, and Cosic 2010; Vulevic and Osmokrovic 2010). In a
more generic sense, the uncertainty components to be considered can be grouped into
three main categories: the measurement uncertainty, the phantom uncertainty, and the
source uncertainty.
The measurement uncertainty comprises the uncertainty in the calibration procedures as well as other measurement factors that will affect the overall uncertainty. These
include temperature and drift effects, resolution of the display, perturbation of measurement by people or objects present in the surrounding environment, and the degree of
repeatability.
The type and positioning of the measurement probes also insert some sort of uncertainty. For example, performance tests conducted on several commercially available
probes equipped with orthogonally positioned sensors showed deviations from isotropy
from ±1.5 up to ±3.4 dB (Kuster 2002), while in the studies of Manning and Gabriel
(2000) the way that the probe is articulated through the entry penetration at the top of a
head model was found to highly influence the SAR values.
Furthermore, the assessment of a dosimetric quantity, such as the SAR, starting from
an unknown field distribution requires field measurements in several hundred points,
data processing, and extrapolation/interpolation between the measured points. Such a
task incorporates many different error components, all of which must be thoroughly
analyzed in such a manner that is valid for all evaluations.
Phantom uncertainty includes the uncertainty with respect to the maximum exposure occurring in realistic conditions. For example, an assessment of the uncertainties
in the estimation of the absolute SAR values for several allowances for the upright geometry of head phantoms has been performed in Manning and Gabriel (2000). In the case
of personal monitoring systems, the perturbation of the impressed field by the phantom
may result in considerable uncertainty. For instance, the field strength recorded by a
body-worn instrument may differ from that recorded by the same instrument in the
same position with the phantom absent by up to 10–15 dB close to body resonance frequencies (few 10s of MHz), depending on the direction of incidence and the polarization
of the radiation (ICNIRP 2009).
Finally, source uncertainty incorporates deviations of the device tested from those
of mass production, manufacturing tolerances, and device position with respect to the
human body.
Repeatable experimental dosimetry is essential to reduce the uncertainties. For example, in the case of exposure assessment to the radiation of cellular phones, the large
standard deviations observed between different samples of phones indicate that the
uncertainty can only be assessed by evaluating a sufficient number of randomly selected
samples. Intercomparison of several experimental methodologies is also required.
Electromagnetic Fields in Biological Systems
magnitude and cannot provide any phase information about the measured field. Future
applications will require development of more advanced scanners providing not only
amplitude but also phase information (Su et al. 2006).
5.3.3 Uncertainty Assessment
Basic information regarding uncertainty evaluation in experimental dosimetry can
be found in (Iskra, McKenzie, and Cosic 2010; Vulevic and Osmokrovic 2010). In a
more generic sense, the uncertainty components to be considered can be grouped into
three main categories: the measurement uncertainty, the phantom uncertainty, and the
source uncertainty.
The measurement uncertainty comprises the uncertainty in the calibration procedures as well as other measurement factors that will affect the overall uncertainty. These
include temperature and drift effects, resolution of the display, perturbation of measurement by people or objects present in the surrounding environment, and the degree of
repeatability.
The type and positioning of the measurement probes also insert some sort of uncertainty. For example, performance tests conducted on several commercially available
probes equipped with orthogonally positioned sensors showed deviations from isotropy
from ±1.5 up to ±3.4 dB (Kuster 2002), while in the studies of Manning and Gabriel
(2000) the way that the probe is articulated through the entry penetration at the top of a
head model was found to highly influence the SAR values.
Furthermore, the assessment of a dosimetric quantity, such as the SAR, starting from
an unknown field distribution requires field measurements in several hundred points,
data processing, and extrapolation/interpolation between the measured points. Such a
task incorporates many different error components, all of which must be thoroughly
analyzed in such a manner that is valid for all evaluations.
Phantom uncertainty includes the uncertainty with respect to the maximum exposure occurring in realistic conditions. For example, an assessment of the uncertainties
in the estimation of the absolute SAR values for several allowances for the upright geometry of head phantoms has been performed in Manning and Gabriel (2000). In the case
of personal monitoring systems, the perturbation of the impressed field by the phantom
may result in considerable uncertainty. For instance, the field strength recorded by a
body-worn instrument may differ from that recorded by the same instrument in the
same position with the phantom absent by up to 10–15 dB close to body resonance frequencies (few 10s of MHz), depending on the direction of incidence and the polarization
of the radiation (ICNIRP 2009).
Finally, source uncertainty incorporates deviations of the device tested from those
of mass production, manufacturing tolerances, and device position with respect to the
human body.
Repeatable experimental dosimetry is essential to reduce the uncertainties. For example, in the case of exposure assessment to the radiation of cellular phones, the large
standard deviations observed between different samples of phones indicate that the
uncertainty can only be assessed by evaluating a sufficient number of randomly selected
samples. Intercomparison of several experimental methodologies is also required.