Modelling fields induced in humans by 50/60 Hz magnetic fields: reliability of the results and effects of model variations

This paper presents a comparison of anatomically realistic human models and numerical codes in the dosimetry of power frequency magnetic fields. The groups at the University of Victoria and the National Radiological Protection Board have calculated the induced electric fields in both their 'UVic' and 'NORMAN' models using independently developed codes. A detailed evaluation has been performed for a uniform magnetic field at 60 Hz. Comparisons of all dosimetric metrics computed in each particular model agree within 2% or less. Since in situ measurements cannot be performed in humans, and achievable accuracy of measurements in models and animals is not likely to be better than 10-15%, the comparisons presented should provide confidence limits on computational dosimetry. An evaluation of the effect of model size, shape and resolution has also been performed and further illuminated the reasons for differences in induced electric fields for various human body models.