A linear, time-invariant model for cancerous and normal breast tissue

Electrical properties of biological cells and human tissue have been used to characterize cancerous vs. normal tissue. Commonly, measured dielectric spectra (permittivity and conductivity) are fitted into an empirical function and the best-fit parameters of the function are considered as a tool to differentiate various types of tissue; however, these parameters do not provide any explanation for the underlying molecular structure. In this work, we modeled the frequency dependence of impedance data collected from human breast tissue using a high-order, linear, time-invariant filter. The results show that the parameters of the filter not only can be used to classify tissue types, they may provide meaningful information about the properties and structure of tissues.