Sensitivity Analysis and Uncertainty Quantification of Effective Models for Microwave Thermal Ablation of Liver Tumors

This paper analyzes the sensitivity to tissue properties of previously described effective biophysical models of microwave liver tumor ablations. Based on the results of the sensitivity analysis, the electrical conductivity and the blood perfusion rate are the most significant model parameters. These parameters are included in a probabilistic model for fast, approximate uncertainty quantification, derived using the mean- value first-order second-moment method (MVFOSM). The MVFOSM model can capture well the mean and standard deviation of the temperature and the thermal damage at the end of ablation but is affected by noise in the calculation of the ablation volume in time. The results of the MVFOSM model are validated against Monte Carlo predictions.