Radiofrequency ablation is an important surgical method for eliminating cancer; however, the lack of adequate technology to image the internal organ temperature profile forces surgeons to often guess at the ablation margin. If a sufficient temperature is not reached and all of the cancerous tissue is not destroyed, a recurrence is likely. Therefore, we propose to develop a numerical electromagnetic and thermal model of radiofrequency ablation that will be used in future surgical planning. The model is based on the finite element method and couples the electromagnetic and thermal models by considering the electric fields as the heat source. Furthermore, the two physical phenomena are coupled through temperature-dependent material properties.
To verify our models, we compare them to experiments conducted on excised bovine liver. Internal temperatures are measured with thermocouples and lesion shape and size are compared after ablation. At the same time, we attempt to predict surface temperature during ablation in order to investigate the possibility of correlating surface temperature to internal temperatures. During the experiments, surface temperature was measured with an infrared camera.
Over the course of three experiments, we found that internal temperatures are predicted with good accuracy (within 2 deg C) when the ablation ground plane is placed more than 8 cm away from the electrode. If the ground plane is closer, then some error is introduced into our approximate model. Also, we found that the lesion shape and size predicted by the simulation are similar to the lesion observed after ablation. Finally, the simulation predictions for surface temperature were mixed. In one case, the temperature values were predicted closely but the distribution was somewhat different. In the other case, the isothermal contours were very similar but the simulated temperatures were as much as 25 deg C above what was measured.