Numerical dosimetry ELF: accuracy of the method, variability of models and parameters, and the implication for quantifying guidelines

In situ electric fields and current densities are investigated by numerical simulations for exposure to ELF electric and magnetic fields. Computations are based on the finite-difference time-domain method (FDTD). The computational uncertainty is determined by comparison of analytical and numerical results and amounts to a worst-case expanded uncertainty (95% confidence interval) of +/-9.89 dB for both dosimetric quantities (E, J). Detailed investigations based on the Visible Human body model with a resolution of 2 mm show a strong influence of the tissue boundaries on the simulation results, which is caused by the numerical method. For the tissue specific in situ electric field and current density changes in excess of 10 dB are observed when comparing the results with and without evaluation of the dosimetric quantities at tissue boundaries. Moderate sensitivities with respect to tissue boundaries are observed only for low conductivity tissues when evaluating the in situ electric field whereas this behavior is observed for high conductivity tissues when evaluating the current density. For exposure to a 50 Hz magnetic field corresponding to the ICNIRP reference level, the simulated current density for central nervous system (CNS) tissue is in compliance with the ICNIRP guidelines. Exposure to a 50 Hz electric field may exceed the ICNIRP basic restriction for CNS tissue at least in a worst-case scenario (grounded human body, vertical electric field, tissue boundaries included for the evaluation of the current density). The in situ electric field is the more stable dosimetric quantity with respect to changes of the tissue conductivity of the Visible Human body model. The maximum conductivity sensitivity coefficient amounts to +122% for the current density whereas the maximum sensitivity coefficient for the in situ electric field is -20%. For electric field exposure the in situ electric field remains comparable (-6% to -4%), the averaged current density change ranges from -57% to -16% for the tissues under investigation. Magnetic field exposure of a scaled model of a five year old child leads to a decrease of the dosimetric quantities (J: -74% to -45%, E: -42% to -23%) compared to the Visible Human results.

Exposure setups for laboratory animals and volunteer studies using body-mounted antennas

For two different in vivo exposure setups body-mounted antenna systems have been designed. The first setup is designed for investigation of volunteers during simulated mobile phone usage. The setup consists of a dual-band antenna for GSM/WCDMA with enhanced carrying properties, which enables exposure for at least 8 h a day. The 10 g averaged localised SAR--normalised to an antenna input power of 1 W--measured in the flat phantom area of the SAM phantom amounts to 7.82 mW g(-1) (900 MHz) and 10.98 mW g(-1) (1966 MHz). The second exposure setup is used for a laboratory behavioural study on rats. The design goal was a localised, well-defined SAR distribution inside the animals' heads at 900 MHz. To fulfil the biological requirements, a loop antenna was developed. For tissues around the ears, a localised SAR value of 50.12 W kg(-1) averaged over a mass of 2.2 g for an antenna input power of 1 W is obtained.