Identification of Material Parameters from Temperature Measurements in Radio Frequency Ablation

Abstract

The mathematical simulation of the method of radio frequency ablation (RFA) offers an opportunity to improve the success of the RFA. The results of the RFA depend highly on the experience of the radiologist. A simulation will offer a prediction of the results which can be used to adapt the setting and enable a complete destruction of the tumor, e.g. by adapting the probe's position. A good simulation needs as much information of the reality as possible. Especially the material properties pose a challenge since they vary from patient to patient, they can not be measured in vivo and they additionally change during the ablation. The aim of this thesis is to develop a mathematical model for the identification of the material parameters from temperature measurements and apply it to appropriate data sets. At first a minimization problem is formulated, where the difference between the measured temperature and the calculated temperature is minimized with respect to the material parameters. The temperature distribution is calculated with a coupled system of partial differential equations. Different approaches are considered which depend on the diverse modeling of the material parameters. The parameters are modeled as constant values as well as temperature dependent, tissue dependent and also spatially distributed. The advantages and disadvantages of the diverse models are illustrated by the numerical results for the identification with artificial temperature distributions as well as real temperature measurements.