A comparative study on algorithms for optimizing Debye model of tissue dielectric properties

Abstract

In order to simulate the interaction of tissues and electromagnetic fields, the dielectric properties of tissues are usually described by the multi-pole Debye model which is converted to the time domain easily instead of Cole-Cole model. This paper presents a comparative study of several algorithms used in optimizing the multi-pole Debye model of tissues dielectric properties. The algorithms are: graphical method, genetic algorithm (GA), simulated annealing algorithm (SA) and hybrid genetic algorithm (hybrid-GA). The accuracy and cost time of these algorithms in optimizing the multi-pole Debye model based on the data of literature and actual measurement are investigated. The results show the fit effects of these algorithms on frequency band and number of Debye poles for the optimizing model.