Uncertainty and Sensitivity Analysis for a Tissue Laser-Irradiation Tissue Model

Abstract

The modeling and control of laser-irradiated tissue is a challenging problem due to the non-linear behavior of the tissue when heated. This is compounded by the fact that in-vivo tissue parameters are not well known. In order to properly model and design control methods for LITT, it is necessary to quantify the uncertainties in the model parameters and their effect on the variability of the final output. For this purpose, a non-linear LITT model has been developed and an uncertainty and sensitivity analysis of the model was performed. This was also done to identify the parameters which have the largest contribution to the uncertainty in the output. The uncertainty analysis revealed that uncertainties in the model parameters can result in a variance of up to 42 °C (40% of the mean) in the predicted temperature. The sensitivity analysis showed that thermal parameters have a larger effect on the predicted temperature and thermal dose than optical parameters. The analysis also showed that changes in the specific heat and mass density had the largest effect on the model output early in a treatment, while thermal conductivity had the largest effect later in the treatment. These results are being used to develop a framework for controlling LITT.

Uncertainty and sensitivity analysis for anisotropic inhomogeneous head tissue conductivity in human head modelling

“…Mesh generation provides 13732, 4711, 5123 and 77966 elements for different tissue layers, respectively. The details of the realistic head model construction are described in our other study [15]. We then assign either isotropic homogeneous or anisotropic inhomogeneous (uncertain) conductivity to each head tissue.…”