Variable porosity-based bioheat model vs variable perfusion-based Pennes’ equation: A comparison with in vivo experimental data

Abstract

• Variable porosity and variable perfusion based bioheat models developed. • Modelling in vivo thermal ablation of a tumor liver tissue bounded by healthy tissue. • Comparison between Pennes’ equation and porous media-based model. • Results in terms of temperature fields and final ablation zones diameters. • Accurate prediction of ablation zones to improve both medical protocols and devices. Thermal ablation has a primary role in cancer treatment due to its minimal invasiveness, and the improvement of this technique is fundamental in order to avoid incomplete ablation or healthy tissues ablation. In this work, modified bioheat models based on Pennes’ or LTNE bioheat models are employed in a simplified two-layer spherical model made up by a hepatocarcinoma tumor bounded by a healthy liver tissue. Improvements are done by including variable porosity or perfusion into either LTNE-based or Pennes-based models, accounting for local properties variations. Other modifications in common to both models, like blood flow arrest induced by necrosis, or employing an equivalent external thermal dose as the heat source term, are introduced too. Numerical results are presented in terms of temperature fields, thermal damage evolution, and the resulting ablation size. The latter, that represent the main outcome from a thermal ablation procedure, is compared with experimental measurements performed on an in vivo clinical study referred to hepatocellular carcinomas in human beings. From these comparisons, it is shown that the modified porous media-based model provides a very good agreement, with lesion sizes differences which are no higher than 9.1%; on the other hand, the modified Pennes’ equation allows to a slight underestimation of the achieved temperatures due to a too much high heat sink effect. The information derived from the present study would be very helpful to improve thermal ablation modeling by employing a LTNE bioheat porous model that accounts for aspects like variable porosity and other model improvements.