Bioelectromagnetic Uncertainty Analysis Using Geometrically Stochastic FDFD Method

In this work, a novel frequency-domain stochastic numerical method named the geometrically stochastic finite-difference frequency domain (GS-FDFD) is proposed to study the variation in electromagnetic fields caused by random geometric changes in the model. The delta method approximation is employed to extract the 1-D and 2-D GS-FDFD equations. Two bioelectromagnetic examples are studied. First, a 1-D multilayered body tissue with size variation is analyzed to calculate electromagnetic field variations. Second, the influence of size variation in a 2-D slice of an anatomical head model on specific absorption rate (SAR) is investigated. In both examples, the GS-FDFD results compare well with the benchmark Monte Carlo results.