Microwave ablation is one of the most common minimally invasive targeted cancer treatment technique. Since development phase of the technique resulted in several commercial ablation systems, current research topics include optimization of the current systems and practices. Recent studies have shown that outcomes of microwave ablation can vary patient to patient. To foresee the ablation results, investigation of treatment planning techniques has gained great interest recently. Simulations have been the main tool for the treatment planning process which requires accurate tissue models. In this work, I investigated new simulation approaches to increase microwave ablation simulation accuracy. Chapter 1 provides an introduction to the current state of microwave ablation and the variability of ablation outcomes with the current systems are presented. Potential sources of this variability were discussed. Moreover, the accuracy of microwave ablation simulations was evaluated with ex-vivo studies (experiments and simulations). The study results suggested that current dielectric property models need improvement. Therefore, following chapters focused on the dielectric tissue property model development for liver and lung tissues. In microwave ablation context, tissue properties can be divided into two groups: baseline and temperature dependent temporal changes during heating. Both depends on the constituents of tissues; therefore, dielectric mixture models were used in the development of liver and lung tissue models. Chapter 2 presents tissue models which were tissue water content dependent. It has been assumed that water content was responsible for the drastic changes in dielectric properties during heating. Chapter 3 outlines the techniques for application of the water content dependent dielectric mixture models in the microwave ablation setting. With this study, it was directly shown that vaporization was the most dominant mechanism in dielectric property change at high temperatures, dielectric mixture models can model vapor induced effects, and the proposed approach with the mixture model allowed more accurate dielectric property predictions than the current models in literature. Since this study focused the temporal changes of tissue properties. In the first 3 chapters, temporal changes in dielectric properties were studied. Chapter 4 presents experiments and simulations in phantoms for tissues with ectopic fat storage, which changes the baseline tissue properties. Differences in heating mechanisms in phantoms with varying fat contents were discovered. Moreover, a framework was introduced for imaging-based (magnetic resonance imaging) determination of baseline tissue dielectric properties. Chapter 5 describes lung tissue dielectric property models that have been built with the data that was collected. The aim of this study was to develop an air content tissue model for lung, therefore, concurrent measurements of dielectric properties with open-ended coaxial probe and air content with CT and microCT was included in the model development process. Chapter 6 describes the implementation of the tissue dielectric property model that was developed in the Chapter 2 and 3 in simulations. The results of the simulated ablation zones were compared with the ablation zones that were measured precisely by accounting for the contraction. The simulations with the proposed model had higher accuracy in the prediction of the ablation zone dimensions in ex-vivo tissues. Chapter 6 also included a study on thermal behavior of the liver perfusion. The literature review on perfusion in the thermal ablation area showed the need for a liver specific perfusion model. In-vivo liver ablation temperatures were collected and used as the training data for the optimization of a muscle perfusion model for liver tissue.