| Popis: |
In this chapter we would like to provide an overview of our own research carried out regarding the modeling of human skin in the framework of projects for assessing the suitability of transcutaneous monitoring of specific physiological changes, the measurement of dielectric changes in cell suspensions and by impedance flow cytometry. Different kinds of models of human skin are compared and applied to scenarios where a sensor is placed on top of human skin represented by a complex heterogeneous and partly anisotropic multilayer structure. These models are based on a rigorous analysis of all components present in human skin at different length scales. The aim was to identify the relevant features and structures that have to be taken into account in order to reliably reproduce e.g. measured dielectric spectroscopy data in the frequency range from MHz up to the virtually few GHz. Focusing on the relevant dispersion-dominated frequency region of 1MHz-100MHz models including very different levels of details are developed: First, a model including sublayers obtained from two-phase mixtures, second, including three-phase mixtures of cell-like core-shell ellipsoids and finally, including multi-phase mixtures obtained from numerical models of single cells with flexible shapes using a surface parameterization based on superquadrics. All skin models are numerically evaluated using the Finite-Element Method (FEM) where the frequency response is retrieved from a fringing field sensor on top of the multilayer system serving as an impedance probe. Using these models the achievable sensitivity and specificity is evaluated. Furthermore, measurements with a planar sensor probing skin in vivo were carried out. The validity of the models was tested by measuring the spectral response of the skin before and after removing the stratum corneum, which stands for the uppermost skin layer, and therefore for the most influential one. It was found that at least a three-phase mixture (with the constituents extracellular medium, cell membrane and cytoplasm) is required to qualitatively reproduce the measured frequency response of the truncated skin model if any a priori knowledge of the underlying material dispersion (e.g. a Cole-Cole fit to measured data) is lacking. Consequently, microstructural features of tissue are an essential part of any accurate skin model in the MHz region. |
