Statistical Outliers in Voxel SARs and Their Effect of Whole‐Body Average SARs in Pregnant Woman and Child for Far‐Field Exposure

Abstract

SUMMARYThe World Health Organization (WHO) recommended the necessity of electromagnetic (FM) dosimetry evaluation of pregnant women with fetus and of children as an urgent research subject in 2006, with emphasis on studies of whole‐body average specific absorption rates (WBA‐SARs) in various numerical models of pregnant woman and children for the purpose of determining the safety limits of WBA‐SARs. The current safety limits were determined on the basis of behavior abnormalities in healthy adult animals in radio‐frequency FM exposure and FM absorption characteristics at resonant frequencies, but not in experiments using pregnant and young animals. In this investigation, we calculated the voxel SARs and WBA‐SARs in anatomically detailed models of pregnant woman and 3‐year‐old children at their resonant frequencies. The histograms and cumulative relative distributions of the voxel SARs were also derived to determine statistical outliers in the voxel SARs for pregnant woman and 3‐year‐old child models. We found that the mean voxel SARs agree with the WBA‐SARs, and that median voxel SARs in the pregnant woman and 3‐year‐old child models are 47% and 55% of their means, respectively, and the peak voxel SARs are 70 times the mean in both cases. This suggests the possibility that finite‐difference time‐domain (FDTD)–calculated WBA‐SARs may be overestimated due to the existence of statistical outliers. It was also found that although the total number ratio of voxel SARs for the outliers is 0.36% for the pregnant woman model and 0.34% for the 3‐year‐old child model, WBA‐SARs excluding these outliers are less than those in the pregnant woman and 3‐year‐old child models by 6.8% and 5.7%, respectively.