Modeling Temperature Effects of RF Thermoablation in a Human Liver using the bioheat formulation in CST DE

Approximately 70% of people who contract cancer of the liver are unable to undergo surgical treatment and are considered to have inoperable cancer. Among the most successful minimally invasive treatments is the direct application of energy using treatments such as Radio Frequency Thermoablation (RFT).

The goal of RF thermoablation is to deposit RF energy in a tumour. RF electrodes are inserted into the tumour and under the guidance of either ultrasound, CT or MRI, RF energy is deposited into the tumour. The subsequent heating of the region destroys the tumour and tissue in a small circumferential region around the tip of the electrode used for depositing the RF energy. The amount of deposited energy is carefully controlled in order to minimize damage to healthy tissue.

Determining the temperature distribution in the liver by means of EM simulation. The finite integration technique is very well suited to simulating the EM interaction between thin metallic radiators and heterogeneous voxel based anatomical models. In order to determine the temperature distribution in the liver due to the electromagnetic fields in the body, a co-simulation was performed using CST MWS and CST EMS. The simulation set-up consists of an electrode, a voxel model of the human body and a reference (grounding) plate. 

This set-up is demonstrated in Figure 1, where a cut view of the model shows the organs and the location of the inserted electrode. The electrode is modeled according to the dimensions of the Elektrotom Hitt 106 device (www.integra-ls.com), and consists of a perfect electric conductor (PEC) and lossless Teflon. Two different body models are used for comparative analysis; these are the HUGO body model, based on anatomical data of the Visible Human Project®, and also a heterogeneous anatomical model (UAq ALES) developed at UAq EMC Laboratory, University of L'Aquila, Italy (http://orlandi.ing.univaq.it/UAq_Laboratory/index.html).

Since the heating is localized, only a section of the torso was simulated. Some of the details of the catheter (electrode) are shown in Figure 2. A 40 W, Gaussian signal with center frequency 375 MHz was used as excitation.

An animation of the Electric field inside the HUGO model is shown in Figure 3. Due to the dielectric and conductive properties of the human tissue, the fields are concentrated around the catheter and the skin touching the groundplane. In Figure 4, the power loss density inside the HUGO model (a) and the UAq model (b) is shown.

The slight differences in the distribution of the power loss density are considered negligible, since the two models are different from each other and it is impossible to have identical cross sections for both models.

Using co-simulation with CST EM STUDIO™ (CST EMS), it is possible to calculate the heating in the tissue, due to an induced current, which includes cooling effects from blood flow as described by the bioheat equation. In Figure 5, a cross section of the temperature distribution in the liver of the HUGO model is shown.

In this study, a co-simulation using CST MWS and CST EMS was performed in order to determine the localized heating inside a tumor when exposed to an RF signal. 

Due to the heterogeneous, voxel nature of the model, a time domain solver was used. In the solution of heterogenous body models, the time domain solution requires significantly less memory, than a frequency domain method. In order to verify results, the simulation was performed on two different body models. The results were found to be in good agreement.