The tumorous tissues prone to have higher electric conductivity than the normal tissue. When the RF electric field is applied, the more electric current is drawn by the tumor cell resulting in the increased temperature around the tumor [1]. This process can be carried out more safely with accuracy if we can monitor the treatment process such as the body temperature [2]. We have performed the phantom and animal experiments using an MR compatible RF hyperthermia system developed for a 3.0 T MRI.

A. Phantom study

A tissue-mimic phantom has been constructed with a simplified structure with liver and tumor part. The dielectric properties of the tumor and normal liver tissues are distinctly different, by the factors of 6-7.5 in the conductivity and 2-5 in the permittivity at 10 to 20 MHz [3]. The liver mimic region (agar, 15 g/L; NaCl, 2 g/L; CuSO4, 1 g/L) and tumor mimic region (agar, 15 g/L; NaCl, 16 g/L) are constructed, where dielectric properties are measured using a DAK system (SPEAG, Switzerland). The thermal conductivity and specific heat capacity are calculated as in [4].

Simulation is performed to calculate the effect of RF electric field and resulting heat generation in the tissue mimic agar-water gel phantom. Using the FDTD method, the SAR map in the phantom is calculated. This SAR information is used as a map of thermal energy source. The power is adjusted to 100 W at 13.56 MHz. The total simulation and experiment time was 2 hour. Electromagnetic and thermal simulation were performed using Sim4Life V2.0 (SPEAG, Switzerland).

We have designed, constructed, and tested experimentally the MR compatible RF hyperthermia system working at 13.56 MHz without affecting MR image acquisition at 128 MHz [4]. The mimetic diagram of the MR compatible RF hyperthermia system is shown in Fig. 1 and the whole system arrangement is shown in Fig. 2. The phantom in the MRI bore is properly heated as expected (13.56 MHz, 100 W) for 2 hours. And MR temperature images of the phantom were obtained every 20 minutes using a 3.0 T Achieva MRI (Philips, Netherlands). PRFs methods (FFE sequence, TE = 15 msec, TR = 300 msec) was used for MR thermometry [5].

B. Animal study outside the MRI room

Prior to MRI experiment, we validated the performance of hyperthermia system and temperature rise is monitored. The pig (weight: 20 kg) is heated for 100 minutes (13.56 MHz, 100 W). The temperature is monitored and the sensor is placed on the top and bottom surfaces of the abdominal part of the pig.

C. Animal study with MR thermometry

The pig (weight: 20 kg) is properly heated as expected (13.56 MHz, 100 W, experiment time: 1 hour) using the installed hyperthermia system in the MRI bore. The breath-hold pulse sequence is used to reduce the motion artifact. PRFs methods (FFE sequence, TE = 10 msec, TR = 150 msec) was used for MR thermometry.

A. Phantom study

Figure 3A shows the thermal simulation result. In the tumor region, temperature rise is about 5 °C/hour and the difference in temperature between tumor and liver region goes up to about 2 °C as the thermal simulation continues. The experimental result based on MR thermometry measurements is shown in Fig. 3B. It can be seen that the temperature rise and the temperature difference in MRI are in high correlation with the simulation result.

B. Animal study outside the MRI room

Figure 4 shows the temperature changes of the pig skin when the RF field is applied. The initial temperature on the top and bottom surfaces are 31.3 °C and 28.9 °C, respectively. The temperature rise was about 10 °C during the heating.

C. Animal study with MR thermometry

The experimental results based on MR thermometry measurements are shown in Fig. 5. The initial temperature of pig skin was 34 °C. In the skin and water bolus region, the temperature rise is almost 4 °C.

Phantom and animal experiments have been conducted and the results compared with the simulations results for tumor and tissue model. They are in very good coincidence with each other. The simulation and experimental results confirm the utility and feasibility of the MR compatible RF hyperthermia system with capacitive driving. A novel switching circuits worked effectively for MR-compatible electrodes for hyperthermia. In the animal experiment, breath-hold imaging technique was applied but motion artifact still exist. New MR sequence are currently being developed to reduce this problem.