**Auteurs**

Catarinucci L, Tarricone L. **On the use of advanced numerical models for the evaluation of dosimetric parameters and the verification of exposure limits at workplace.** Radiat Prot Dosimetry. Oct 21, 2009. Ahead of print.

**Background**

Methods of accurately measuring human exposure to electromagnetic (EM) fields are often debated. This issue has come to the forefront recently in the European Community with the publication of the Directive 2004/40/EC, which makes the employer responsible for exceeding EM exposure safety limits at the workplace. Within this context, it is important for the scientific community to decide on correct and practicable methods to evaluate risks of EM exposure.

**Objective**

The objective of this paper was to discuss the options available and to illustrate the issues associated with accurate measuring of human exposure to EM fields through three examples.

**Methods**

The science of measuring exposure to EM energy, including radiofrequency (RF) fields, is called dosimetry. The dose of EM energy absorbed by biological tissues is measured as specific absorption rate (SAR), in W Kg-1. Methods of measuring SAR can be broadly classified as theoretical, experimental, and numerical. Complex exposure problems (such as human exposure) are often tackled by the latter two. In experimental dosimetry, artificial human body models, called phantoms, are used. Phantoms tend to be homogenous, representing at most only a few kinds of different tissues. Numerical dosimetry relies on very accurate heterogeneous numerical body models, to which different sources of EM radiation can be exposed using computer simulations.

In the first example, the authors discuss how SAR values from a numerical dosimetric study can depend on the type of heterogeneous phantom chosen and whether a homogeneous approximation of the phantom can provide accurate results. In the second example, action values and exposure limits measured by numerical dosimetry on a heterogeneous phantom are compared. In the third example, different mathematical formulas used in numerical dosimetry to calculate the SAR are compared.

**Results**

In the first example, six phantoms were studied (two heterogeneous, four homogeneous approximations of the heterogeneous models) at various distances from the exposure source. The authors found that differences up to 40% were found between the two heterogeneous models. Differences were smaller when each heterogeneous model was compared to its two corresponding homogeneous models. In the second example, a heterogeneous phantom exposed to EM fields (92 V m-1) higher than the ICNIRP action value (88 V m-1) had a SAR of 1.4 W kg-1, much lower than the exposure limit (10 W kg-1). In the third example, depending on how the mathematical formulas specified the volume, the calculated value of SAR varied.

**Interpretation and Limitations**

Heterogeneous models should not be simplified into homogeneous models. Exceeding the action values does not necessarily mean that the exposure limit has been exceeded, and correlation between the two values is not straightforward. The most sophisticated mathematical formulas provide the best estimates of SAR.

**Conclusions**

The authors present some critical aspects of numerical dosimetry that should be considered when compliance with safety limits is being determined.