2006 AIChE Annual Meeting
(602g) Multiscale Modeling of the Physiological Response of Skin to Focused High Power Microwave Exposure
Our long-term objective is to employ multiscale mathematical modeling to understand the complex response of skin to HPM exposure at a single cell resolution. To achieve this objective, we must first understand the spatial-temporal dynamics of the abnormal skin temperature and EM fields that together form the perturbation field driving physiological response. We propose the overall response of skin to AD exposure is the integration of single cell responses calculated by assuming each cell in a multicellular matrix can execute discrete probabilistic moves in response to perturbation. The probability of any particular move is a function of both intracellular and extracellular factors, e.g., the external temperature and energy field. Thus, models describing whole-body dynamics and energy transfer in the skin are key precursors to our investigation of the cellular response of skin to AD exposure. Whole-body Physiologically Based Pharmacokinetic (PBPK) models modified to account for body-region are used to predict the spatially averaged temperature distribution in the skin and changes in Skin Blood Flow rates (skBF) resulting from AD exposure. Parameters for the PBPK model are taken from the literature and the model is validated against previous studies where the relationship between skBF and cardiac output was probed by spraying supine men and women with 33C or 35C water pulses and measuring the forearm blood flow rate. The whole-body and shorter length scale skin models are coupled through boundary conditions at blood-vessel walls. To capture the spatial-temporal energy distribution of skin following AD exposure, the transient three-dimensional energy balance equation is formulated and solved in parallel using the Galerkin method and the PETSc library. We initially consider a cubic tissue section composed of epidermal and dermal layers containing blood vessels. Radiation transfer is assumed to possess a Gaussian intensity distribution and heat-transfer can occur at the blood-vessel walls and the skin surface. Algorithm scaling results are presented where wall-clock time is measured for a fixed spatial grid size with an increasing number of processors. In addition, experimental strategies for validation of the skin-energy model predictions, using the RF exposure chamber and single layer in-vitro skin model of Cho and coworkers (Dept. of Bioengineering, U. of Illinois), are presented.