MTHR - Mobile Telecommunications and Health Research

Joint Session on Standard Human Exposure System

Chair: Professor L J Challis

A Joint session had been arranged between the Human Volunteer Study and Dosimetry Workshop participants to discuss progress and problems with the development of the Standard Human Exposure System.

Dr Philip Chadwick gave a short presentation on the exposure system he has been developing for the programme in consultation with a Working Group (comprising members of the Programme Management Committee and invited external experts) and the project teams.

The underling philosophy had been to produce an exposure system that would be easy to use, compatible with double blinded study designs, and that would provide exposures that were standardised, reproducible, quantifiable and that realistically simulated exposure to a mobile phone. Exposure from real phones are typically very localised around the antenna and sometimes also the body of the phone. This localised exposure means that it is important to ensure a standard position for the phone. The localised exposure will be achieved by using a self-contained generic handset based on the Nokia 1610 phone and reproducibility of position will be ensured by mounting the phone on a headband. The system is capable of producing either a CW carrier frequency, or a single modulated signal, (simulating either a GSM 900 mobile phone or a TETRA emergency services personal radio). All phones can operate at one of three power settings, representing 100%, 50% and 25% power, although at this stage it is anticipated that all studies will use the highest output only. In addition, the output of the unit can be directed to the antenna (exposure), an internal load (sham exposure condition), or an external monitoring port that can be used for simple tests of phone functionality. The different output power, exposure type, and modulation options can be combined to give 18 different emission modes. These are selected via two hexadecimal switches that give a possible 256 coded output options. The 18 emission modes are randomly and multiply assigned between these settings.

The handsets are designed to produce maximum peak powers of either 1W (TETRA) or 2 W (GSM), corresponding to maximum mean powers of 250mW. Dosimetry has been checked in a calibrated test facility with UKAS accreditation using the CENELEC standard assessment position. Data obtained indicate that energy deposition is strongly localised to the area surrounding the antenna. At 100% output the system produces a maximum SAR adjacent to the antenna of 1.4 W/kg averaged over 10g; the SAR dose to the body of the phone is around 0.4 W/kg. For comparison the sham exposure setting gives a maximum SAR of less that 0.0001 W/kg. It is expected that the phone will be capable of running for up to an hour on one set of batteries.

In subsequent discussion it was agreed that the comfort of subjects expected to wear the phones for an hour would have to be determined by trial. The phones had not been designed for exposure of childrens’ heads, and dosimetry would be an important issue if this was ever required. There was also some discussion of the problems associated with exposure of the left and right sides of the head. This arises because, like a real phone, the antenna is not central. The Working Group had considered this issue and decided that the phone would be offset when used on the left side of the head, in order to ensure that the antenna position, which defines the location of the peak SAR, will be in the same position on both sides of the head. It was accepted that this would not exactly replicate the situation with a real phone, but then different models of phones have the antennas in different positions anyway. The important issue was to standardise exposures.

There was discussion about the blinding of the system. There was some concern that it was possible to defeat the blinding of the exposure mode settings by working through them. To prevent this, some settings are not assigned. In addition, the phone also has a capability to simulate discontinuous transmission (DTX) and this is also randomly assigned to the exposure mode settings. Another important issue was whether subjects could pick up cues from the phone. It had been found that the phone generates a very quiet noise in one exposure mode, although this is only detectable in a very quiet room. This arises from an electrical effect in the battery and could be overcome by using an external power supply or masked with white noise; volunteers in the blood pressure study will watch videos anyway. One possible problem would be interference of the phone with equipment in the room. For example, real TETRA handsets are known to produce interference. However initial tests had not identified any problems with the exposure system ‘phones’. If a problem is identified by research groups a solution would have to be developed.

Professor Tony Barker (Royal Hallamshire Hospital) gave a presentation on preliminary temperature measurements on the exposure systems that were undertaken to determine whether subjects could obtain exposure cues from the phone temperature. Using an infra-red imaging system it had been found that the phones start to heat up after around 8 minutes of operation and reach thermal equilibrium in around 35 minutes. At this stage the bulk temperature of the phone will have increased by around 7-10oC depending on the type of phone in use. The main driver for the work was concern that directing output power into the dummy load (sham exposure condition) may cause heating of the load that would be detectable by the subject.

The temperature of the dummy load was not found to be dependent on the sham or exposure mode selected. However, it had been noted that the CW mode of operation resulted in greater heating in the vicinity of the processing chip in the middle of the phone. This effect resulted in an increase in bulk temperature of 4-5o°C for GSM handsets and 1.5-3°C for TETRA handsets. For any single modulation mode (GSM, CW, TETRA) there is no difference between the sham and exposure modes.

In subsequent discussion it was recognised that where the design of a study did not require consecutive exposure to CW and unmodulated signals (on the Sheffield study, for example, these are carried out a week apart) the difference may not be recognised. Nevertheless, it was felt that the effect was unacceptably large and that the phone would require modification (this has since been successfully completed). There was also concern that even if the subjects were unable to distinguish the different exposure modes, it may still be possible for the investigator to detect this and so break the blinding code. Again this would depend on the study design. If the phone was not switched on until fitted to the subject, it is unlikely that the investigator would have knowledge of the temperature.

There was some discussion about differences in the extremely Low Frequency (ELF) fields that would be associated with the different modes of operation. This had been previously discussed by the Standard Human Exposure System Working Group, but the decision reached would now be reviewed in light of the new information. In any event the first priority would be to establish whether there was any difference between sham and exposed conditions. If there was, then it would be important to determine whether this was due to the RF or ELF field components.

There was also discussion about whether the measurements reflected the situation in normal use where the phone would be radiating into the head rather than free space. It was felt to be unlikely that the different mismatch would have a significant effect. In support of this it was noted that no difference was observed between the sham and exposure conditions although the former directed the RF to a matched load and the latter did not.

The different modes of operation did not appear to have a significant effect on battery life.

Summaries of the closed sessions of the MTHR Research Seminar - 11th of November 2002