Finite-element modeling of needle electrodes in tissue from the perspective of frequent model computation.

Information about electric field distribution in tissue is very important for effective electropermeabilization. In heterogeneous tissues with complex geometry, finite-element (FE) models provide one of alternative sources of such information. In the present study, modeling of needle electrode geometry in the FE model was investigated in order to determine the most appropriate geometry by considering the need for frequent FE model computation present in electroporation models. The 8-faceted needle electrode geometry proposed--determined on a model with a single needle electrode pair by means of criteria function--consisted of the weighted sum of relative difference between measured and computed total current, the relative difference in CPU time spent on solving model, and the relative difference in cross section surface of electrodes. Such electrode geometry was further evaluated on physical models with needle arrays by comparison of computed total current and measured current. The agreement between modeled and measured current was good (within 9% of measurement), except in cases with very thin gel. For voltage above 50 V, a linear relationship between current and voltage was observed in measurements. But at lower voltages, a nonlinear behavior was detected resulting from side (electrochemical) effects at electrode-gel interface. This effect was incorporated in the model by introducing a 50-V shift which reduced the difference between the model and the measurement to less than 3%. As long as material properties and geometry are well described by FE model, current-based validation can be used for a rough model validation. That is a routine assay compared with imaging of electric field, which is otherwise employed for model validation. Additionally, current estimated by model, can be preset as maximum in electroporator in order to protect tissue against damage.