Three identical rectangular waveguide chambers made of aluminium were used in the study. The exposure chamber consisted of a straight rectangular section (24.8 cm x 20 cm) and of two waveguide-to-coaxial adapters. The total length of the chamber with the adapters was 190 cm. The mice were kept in small cylindrical acrylic restrainers (i.d. 32 mm, length adjustable) preventing them from aligning their longitudinal axis parallel to the electric field. The restrainers were placed 25 at a time on a Styrofoam holder so that each mouse was in the center of the cross section of the waveguide with its longitudinal axis perpendicular to the electric field and the direction of propagation. The distance between adjacent mice was 5 cm. The mice that were missing were replaced with mouse phantoms made of plastic cylinders partly filled with a solution simulating the dielectric properties of muscle at 900 MHz.
Modified and computer-controlled mobile phones were used as signal sources. The signals were amplified by boosters manufactured for vehicle use. RF power meters were used for measuring the input, reflected and output power of the exposure chamber. The output power was absorbed by a coaxial 10 W termination. The SAR was controlled daily by adjusting the input power.
The animals were randomized into four equal groups: NMT exposure, GSM exposure, sham exposure and cage control. The sham exposed group was kept in an unenergised waveguide chamber. The mice in all groups except the cage-control group were exposed to ionizing radiation at the beginning of the study. The total-body dose was 4 Gy delivered as three equal fractions of 1.33 Gy at 1-week intervals with linear accelerators. The dose rate was 0.51 Gy/min. To ensure a uniform irradiation of the whole body, a Plexiglas scatterer was placed 15 cm above the animals.
The whole-body-averaged SAR of mice was determined by the measurement of RF power absorbed in mice and a corrected conventional absorption equation. The SAR based on the power measurements was verified by measurements of the temperature increase of the phantoms for each phantom location. The two methods agreed within 2%. The spatial variation of the SAR due to reflections and power attenuation in the waveguide was estimated by electric-field measurements with a dipole probe. The maximum variation from the average SAR was about ± 30% for a 27-g mouse, and it decreased with the mass of the mouse. To ensure identical long-term average exposure, the order of animals in the Styrofoam holder was randomized for each exposure session. The estimated uncertainty of the whole-body-averaged SAR of an individual mouse during the whole exposure period was ± 15%. The measured SARs were compared with values obtained from FDTD calculations revealing that the ratio of the local maximum SAR inside the mouse to the whole-body-averaged SAR was from 4 to 8 and decreased with increasing mass of the mouse.
Shirai T et al.
Lack of promoting effects of chronic exposure to 1.95-GHz W-CDMA signals for IMT-2000 cellular system on development of N-ethylnitrosourea-induced central nervous system tumors in F344 rats.