Characterizing the shape of dose-response curve is an essential part of the quantitative assessment of risk to human health from man-made chemicals. Ideally, this characterization is based on chemical-specific mechanistic data for people. In the absence of such data, the dose-response curve may be based on default assumptions and/or on extrapolation from exposure of laboratory animals.Mathematical modeling of biological structures provides a way for a hypothesis testing and a meaningful extrapolation between species.

Anna Georgieva has developed a mathematical model that links the airflow-driven deposition of formaldehyde with regional formation of DNA-protein cross-links in the rat nasal passages. This is done by integrating a three-dimensional, anatomically-accurate computational fluid dynamics (CFD) model of the rat nasal airflow and gas uptake and a phramacokinetic model for formaldehyde distribution in the nasal lining. Because of good agreement with laboratory measurements, the model's parameters will be used in a benchmark dose study to extrapolate fodmaldehyde-induced cancer in rats to potential cancer risk in humans.

In parallel, Anna Georgieva is studying the effect of cyclic breathing on airflow patterns and formaldehyde uptake using CFD simulations in the human nasal passages. Cyclic breathing is modeled by the time-dependent Navier-Stokes equations for airflow, using physiological breathing rates, with the appropriate periodic boundary condition at the nostrils. In addition, she is exploring the use of computer-aided design to obtain meaningful ways for validation of the computer results in a laboratory experiment.

Figures below show a comparison between experimental streaklines in a stereolithography reconstruction of the human nasal passages with simulated streamlines in the human CFD model. The point of entry at the nostril's surface,marked on the left, and the direction of the flow are in very good agreement.