Cells were exposed to TETRA RF fields in a perfusion chamber mounted on the stage of an inverted microscope within a stripline parallel-plate transmission line. The exposure system that is depicted with measures in a figure, was similar to that described in the reference article, but adapted to fit the stage of the inverted microscope. TETRA RF waveforms were provided by a function generator and an amplifier. The output waveform from the signal generator comprised an allocated frequency of 380.8875 MHz modulated according to the ETSI standard, to simulate that of in-service hand-sets. A directional coupler and power meter were used to monitor forward and reverse power. For exposures, the RF signal was fed into the transmission line and terminated (50 W termination). For sham exposures, the RF signal was fed directly into the termination, thus bypassing the transmission line. Cell cultures were randomly assigned to exposed or sham groups.
In order to establish a relationship between the input power of the generic TETRAsignal and the SAR in the cell culture, the exposure system (including the microscope objective) was modelled using the finite integral method code, CST Microwave Studio. The electric field strength was determined along the central line of the transmission line and the calculated field strengths were then used to determine the SAR inside the perfusionmedium volume. No distinction was made between the cellular component within the perfusionmedium volume and the bulk perfusionmedium. Dielectric properties of the cell perfusion liquid were measured at 34°C using an open ended coaxial probe and a computer controlled network analyser in the frequency region of 130 MHz-20 GHz. Using measured values for the perfusion solution, the distribution of SAR within the perfusion fluid volume was calculated. In the vicinity of the microscope objective, where the cells were recorded, the SAR varied by ±20% across the field of view, i.e. the SAR values for an average exposure of 400 mW/kg ranged from 320-480 mW/kg. Taking into account the temperature dependence, the whole volume average SAR at 30°C was determined to be 7.88% lower, and that at 37°C 3.17% higher, than that at 34°C. The input values used for the computational modelling of the transmission line were confirmed by measurement at the National Physical Laboratory. The measurements of impedance, voltage standing wave ratio and E field were in good agreement with the computed values.