Personal exposure to magnetic fields emitted by electronic article surveillance (EAS) systems was systematically assessed based on measurements of a representative sample of 19 different EAS devices. This sample included the two major EAS technologies currently on the market: acoustomagnetic (AM) and radio frequency (RF) systems. In addition to these measurements, numerical computations of the current densities and electric field strengths induced in body tissues were carried out for one representative AM-EAS device and several body models (adult male, adult female, child female, and a hand model) and reasonably foreseeable worst-case exposure scenarios using a real-valued magneto quasi-static solver, based on the Scalar Potential Finite Element (SPFE) method. The obtained measurement and computational results were compared to the different sets of exposure limits defined by the International Commission for Non-Ionizing Radiation Protection (ICNIRP) in 1998 and 2010. Our results demonstrated that current RF-EAS technology, which typically operates in the 8.2 MHz frequency range, does not conflict with the exposure limits, even under adverse exposure conditions. However, AM-EAS systems, which typically operate at 58 kHz, may lead to induced current densities in the central nervous system and induced electric field strengths in the peripheral nervous systems of adults and children that exceed the basic restrictions for the general public (up to a factor of 22.5 for the adult female bending in front of the antenna) and even for occupational exposure (up to a factor of 4.5 for the same scenario) according to the ICNIRP 1998 and 2010 guidelines under reasonably foreseeable exposure conditions, which are not covered by the assessment procedures defined in the present version of the applicable standard IEC EN 62369-1. Therefore, radiation protection and market regulatory authorities should have a close look and check presently installed and future AM-EAS technology with respect to their compliance to exposure limits.
Schneeweiss et al. (Fri,) studied this question.