A Monte Carlo Computational Framework Yielding Optimal RF-Dose Bounding for Rodent Bioassays Conducted in Reverberation Chambers

Large-scale lifetime rodent bioassays conducted in reverberation chambers (RCs) rely on automated radiofrequency (RF) power control feedback, driven by a sparse set of field probes, to regulate cohort-wide RF energy absorption. As moving rodents load the RC differently from calibration phantoms, the relationship between probe readings and the realized whole-body specific absorption rate (wbSAR) shifts. Compounded by inevitable instrumentation and thermal calibration errors, sparse field sampling may introduce a wider, yet undetected, exposure variance. To quantify and mitigate these effects, a Monte Carlo (MC) dosimetry framework emulating the calibration-to-live experiment chronology, including field-sensing-driven RF-power control, is presented for an RC loaded with 96 male rat models at 900 MHz. Feedback control schemes based on net RF input power and total RF energy density metrics are compared in terms of realized RF dose. Contextualized within historical bioassay parameters, the computational results indicate that compounding sparse field sampling at arbitrary RC locations with said errors induces an unknowable, unrecoverable cage-wise exposure spread, potentially causing dose-group overlaps and overexposures. To identify optimal probe locations and types, an MC-based selection criterion is demonstrated to substantially mitigate dosimetric spread, bounding cohort-wide exposures within known envelopes and providing a reliable dosimetric design framework for future rodent bioassays.