Stochastic dosimetry of a three compartment head model

Abstract

The deterministic three-compartment head model based on the hybrid finite element/boundary element method (FEM/BEM) formalism is coupled with the nonintrusive stochastic collocation method (SCM) in this paper. The SCM based on sparse grid (SG) interpolation is used to assess the stochastic moments of the electric field induced in the head exposed to high frequency (HF) plane wave. The conductivity and relative permittivity of the scalp, skull and brain tissue, respectively, are modelled as random variables with uniform distribution. The stochastic mean and variance of the internal field in three tissues are computed and the sensitivity analysis of the tissue parameters is carried out, as well. The convergence of the SCM method is shown to be satisfactory. Presented approach provides an insight into the behaviour of the model output with respect to input parameters variation. The analysis showed that the highest impact pertains to scalp permittivity, while skull conductivity impact can be considered rather negligible. The results obtained using the three-compartment head model confirm that both brain permittivity and conductivity are the parameters most significantly influencing the variance of the induced field inside the brain.