Australian Centre for RF Bioeffects Research
An NHMRC Centre of Research Excellence
You are here: Skip Navigation LinksHome   >>   Research   >>   RF Dosimetry
Skip Navigation Links
Human Neurobiology
Molecular Modelling
PublicationsExpand Publications
RF Dosimetry
Rodent Neurobiology
RF Dosimetry
Rigorous dosimetry is a key issue for ensuring the validity of results within a given RF study and consistency of results between RF studies. This requires that a well-characterized standard methodology and suite of tools be utilized for the provision of dosimetry to research programs that are run within the Centre. Further to this, concerns have been raised regarding the application of compliance techniques in RF human exposure standards, based on adult models, to children, and there are questions as to applicability of current compliance techniques to the rapidly changing new communications technologies. Therefore, apart from providing the necessary dosimetry for ACRBR research programs, we will also undertake an in depth study of fundamental dosimetry issues.

Current Projects


Specific Absorption rates (SAR) Compliance Methods


Project Leader: Mr. Ray McKenzie


NHMRC Funded staff: Maia Sauren (PhD student)


Background: RF transmitter devices are tested for compliance with exposure standards through the measurement of deposited RF power in fiberglass human phantoms filled with a tissue equivalent liquid.  Computational modelling is also rapidly developing into an alternative tool for human exposure analysis. 

Aims: To further advance knowledge in the identified areas and ensure the highest possible standard of RF dosimetry, this project seeks to test the accuracy and applicability of current and proposed exposure assessment techniques for humans of all shapes, sizes, and tissue content. 

The main topics considered will be:

(i)         evaluation of measurement protocols, as cited in regulations both nationally and internationally, given the protocols differ from each other in a number of ways;

(ii)        review of the influence of phantom shapes on energy deposition levels with a comparison of the Standard Anthropomorphic Mannequin (SAM) phantom cited in many protocols (eg CENELEC EN50361:2001) with others such as the phantom currently in use at TRL.  Further, analyse the influence of body shape and tissue content through the comparison of computational human body models such as the Visible Human (US National Library of Medicine), NORMAN (UK National Radiological Protection Board), and Utah Man (University of Utah);

(iii)      investigation of the scalability of results from adults to young teenagers and children (issues include the change of anatomical features and tissue dielectrics with age, and, for numerical models, the accuracy of results with changing voxel size);

(iv)      applicability of using homogeneous tissue phantoms compared with a complex multi-tissue human;

(v)       general sensitivity analysis of the effect of body size, shape and tissue type and the implications for standards setting.



Specific hypotheses tested

bullet What anatomical variations exist in human populations?
bullet Which anatomical variations affect the amount of RF energy absorbed by the human body, and how? What are the dielectric properties of human tissues and how important are they for SAR determination?
bullet What is an average human and can it be used as an accurate estimate for modelling purposes?
bulletIs there a difference in SAR between males and females? Between racial groups? Between children and adults? Between pregnant and non-pregnant females?
bullet Should different models exist for various population subgroups? How would such models be distinguished – by age, race, gender, other?
bulletHow is energy absorption in children different to adults? What's the threshold age between children & adults?
bulletCan children be modelled as small-scale adults?
bullet What approximations can be made in regards to population variations for mathematical and physical SAR modelling without sacrificing accuracy?
bullet Which is more important – average SAR or maximum SAR?
bulletAre current phantoms an accurate representation of humans for the purpose of compliance testing?
bulletAre whole-body models any better than head-only or head-and-shoulder ones?
bullet Should models be homogeneous, i.e. model the body/head as consisting of one tissue type, or heterogeneous? How much homogeneity can be considered reasonable?
bulletIs it appropriate to assume that the brain is a 3-sphere model for modelling purposes and research?
bulletHow does SAR correlate to temperature rise in various tissues?


Relevance to RF Bioeffects Research:

This study will allow faster, more efficient, more accurate determination of compliance of radiofrequency technology with existing radiation protection safety standards.


community benefits:

bullet known exposure levels from radio communications equipment use
bullet public confidence in compliance processes and requirements


industry benefits:

bullet better guidelines for compliance testing of radio communications equipment

more accurate compliance testing

bullet less money and resources spent on compliance testing of radio communications equipment



Methods: Investigations will initially be limited to SAR inside the human head.  Due to the nature of available computational models based on either complex realistic models such as Visible Human (Brooks Air Force) or the simple single tissue SAM phantom model (specific Anthropomorphic Mannequin; IEEE) this issue is not easily explored since the identified parameters are not easily varied within the models.  To overcome this deficiency, we have proposed an alternative compromise model which includes a reduced set of tissues in a semi-homogeneous, simplified geometry for which the key parameters may be varied parametrically.   A review of the current literature is used to obtain an estimate of variations in a set of key anatomical parameters affecting SAR - tissue dielectric properties, thickness, relative location and tissue size. This information is used to vary the model at the 5th, 50th and 95th percentiles of human anatomical range for five key tissues: skin, skull, brain, eye and ear. 


Plane wave excitation is used as the source.  The study may be expanded at a later stage to include more tissues and other forms of excitation.  Mathematical modelling is performed using commercially available FDTD (finite-difference time domain) and methods of moments software packages. Phantom studies will be included for validation purposes. These will take place at the Telstra Research Laboratories (TRL) using SAM head phantoms; the TRL whole body phantom, which contains partitions for head, torso, arms and legs; and an IEEE P1528 compliant flat torso phantom.


Results: A simplified model of the human head has been constructed. The model contains the five tissues of interest previously noted, the properties of which will be varied over the range of values for the human population as previously described.  The shapes in this model may be varied parametrically, greatly reducing the effort required to undertake this extensive modelling task.  We have reviewed some of the available literature on the relevant tissue properties which we have used to populate the new models.  A preliminary study has been conducted on the effect of adult cranial thickness on SAR (see Sauren et al 2005 and 2006), and work is now proceeding on the effects of skin thickness on SAR.  With further work and validation, it is hoped to develop a more efficient and accurate model for determining compliance with SAR based exposure levels for a range of RF devices.



Dosimetry support to other ACRBR studies


Project Leader: Mr. Ray McKenzie


NHMRC funded staff: Teddy Kurniawan (PhD student)


Aims: To provide Telstra research Laboratory (TRL) dosimetry support to current ACRBR investigations



Rodent Stream

In Vitro

This work involves the exposure of cells and tissue cultures using the TRL designed co-axial RF exposure chamber.  Dosimetry support in this case involves the provision of signal generation and transmission equipment, chamber refinements and ongoing dosimetry analysis.  An additional study will consider analytic modelling solutions to determine SAR in very thin layers such as cell membrane interfaces.


In Vivo

At least some of this work will involve the exposure of animals (mice or rats) using the Motorola designed RF exposure carousel currently located at IMVS or other apparatus.  Modification of this apparatus for exposure of different animals or the use of different exposure apparatus or regimes will require the provision of some signal generation and transmission equipment and ongoing dosimetric design and analysis.


Human Stream

RF-Sensitivity and Hypersensitivity

This work requires the use of modified mobile phone handsets featuring remote control of power output by proprietary software, provided by TRL.  Exposure assessment will require computational dosimetry analysis, and, most likely, validation by physical SAR measurement in one of TRL's phantoms.


RF-Effect Consistency

As for above


Heat Shock Proteins

Molecular Modelling of RF/HSP Interaction

To validate the molecular modeling results empirically, exposures of proteins in vitro at known SAR's will be required to reproduce any potential Heat shock protein (HSP) interactions predicted by the Resonant recognition model (RRM) models.  This will require provision of and dosimetry design for exposure apparatus and ongoing dosimetric analysis.



'Morpheous' Mobile phone effects in teenage children

This study involves the investigation of end points in subjects exposed from normal use of their own handsets (i.e. not a controlled laboratory situation).  To estimate exposure for this study, analysis of actual handset typical exposures will be conducted on the combined basis of self-reported usage, subscriber records and handset type.  Dosimetry analysis will be undertaken to determine typical exposure from at least a few handset classes which may reasonably represent the actual handsets used by subjects in this study.  Additional account will be taken of relevant network parameters relating to location of use etc.  Further validation using the Motorola Hardware and Software Modified Phones (SMP, HMP) which function as a normal phone but provide additional data which can be used for RF exposure assessment will also be considered.


Results: Expected December 2008


© 2008 Australian Centre for RF Bioeffects Research
Website Built By Kane Elfman