Development of accurate multi-pole debye functions for electromagnetic tissue modelling using a genetic algorithm

The development of anatomically and dielectrically representative tissue models is key to the design and refinement of electromagnetic based diagnostic and therapeutic technologies. An important component of any such model are accurate and efficient Debye models which allow for the incorporation of the frequency dependent properties of biological tissues. The establishment of multi-pole Debye models is often a compromise between accuracy and computational cost. Furthermore, some finite difference time domain schemes impose constraints on the minimum Debye pole time- constant. In this study, the authors have developed an optimised genetic algorithm to establish Debye coefficients with minimal yet sufficient Debye poles for several different biological tissues. These Debye coefficients are fitted to existing Cole-Cole models and their accuracy is compared to previously fitted Debye models.