Background: Electroporation-based cancer treatments are minimally invasive, nonthermal interventional techniques that leverage cell permeabilization to ablate the target tumor. However, the amount of permeabilization is susceptible to the numerous uncertainties during treatment, such as patient-specific variations in the tissue, type of the tumor, and the resolution of imaging equipment. These uncertainties can reduce the extent of ablation in the tissue, thereby affecting the effectiveness of the treatment. Purpose: The aim of this work is to understand the effect of these treatment uncertainties on the treatment outcome for irreversible electroporation (IRE) in the case of colorectal liver metastasis (CRLM). Understanding the nature and extent of these effects can help us identify the influential treatment parameters and build better models for predicting the treatment outcome. Methods: This is an in silico study using a static computational model with a custom applicator design, spherical tissue, and tumor geometry. A nonlinear electrical conductivity, dependent on the local electric field, is considered. Morris analysis is used to identify the influential treatment parameters on the treatment outcome. Seven treatment parameters pertaining to the relative tumor location with respect to the applicator, the tumor growth pattern, and the electrical conductivity of tissue are analyzed. The treatment outcome is measured in terms of the relative tumor ablation with respect to the target ablation volume and total ablation volume. Results: The Morris analysis was performed with 800 model evaluations, sampled from the seven dimensional input parameter space. Electrical properties of the tissue, especially the electrical conductivity of the tumor before ablation, were found to be the most influential parameter for relative tumor ablation and total ablation volume. This parameter was found to be about 4-15 times more influential than the least influential parameter, depending on the tumor size. The tumor border configuration was identified as the least important parameter for treatment effectiveness. The most desired treatment outcome is obtained by a combination of high healthy liver conductivity and low tumor conductivity. This information can be used to tackle worst-case scenarios in treatment planning. Finally, when the safety margins used in the clinical applications are accounted for, the effects of uncertainties in the treatment parameters reduce drastically.
Conclusions:The results of this work can be used to create an efficient surrogate estimator for uncertainty quantification in the treatmentThis is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.