Monte-Carlo Based Numerical Dosimetry in Reverberation Chamber Exposure Systems Employed for in-Vivo Rodent Bioassays

A Monte-Carlo based computational approach for the statistical characterization of the whole-body specific absorption rate (wbSAR) variability in large cohorts of rodents exposed to radio-frequency (RF) energy in reverberation chambers (RCs) is applied to adult male rat exposures illustrative of those in a US National Toxicology Program (NTP) cancer bioassay. A large number of 3D electromagnetic field realizations fulfilling Rayleigh fading properties were generated within an electrically-large volume representative of an ideal RC, yielding granular wbSAR distributions for an ensemble of 96 homogeneous rodent models with random mass distribution, postures, positions and orientations. Two case studies were addressed: a "momentary exposure " with each rat fixed in posture, position and orientation, and a "day-long exposure " in which the position, orientation and posture were varied randomly for each subsequent Rayleigh field realization. Over 500 and 2500 field realizations or "snapshots ", respectively, the rats' instantaneous wbSARs, as well as their individual time-averaged wbSARs, were found to be well fit by lognormal distributions. The large variability in instantaneous wbSARs in the cohort was due in part to the inherent Rayleigh field variability in RCs (70-80%) and in part to weight, posture and position variations (20-30%), while the effect of cage location was found to be small over day-long exposures. Averaging the exposure over field realizations substantially reduces the range of wbSARs in the cohort. Hence, when RF-induced thermal effects are studied, the relevant exposure metric (wbSAR averaged over appropriate times) features a narrower range than instantaneous wbSAR, which is the relevant metric in studies dealing with non-thermal effects. Compared to previous studies, the present approach was found to be computationally more efficient enabling thus a Monte-Carlo analysis by varying concurrently the incident field and the animals posture, position, and orientation. In practice, it can inform the choice of wbSAR targets in rodent bioassay, allowing to identify possible dose-effect trends while avoiding undue thermal stress.