Uncertainty in hyperthermia treatment planning: the need for robust system design.

摘要

Hyperthermia treatment planning (HTP) is an important tool to improve the quality of hyperthermia treatment. It is a practical way of designing new hyperthermia systems and can be used to optimize the phase and amplitude settings to achieve optimal heating. One of the main challenges to be dealt with however is the uncertainty in the modeling parameters. The role of dielectric and combined dielectric and perfusion uncertainty on optimization was investigated by means of HTP for six different systems: the 70 MHz AMC-4 (AMC: Academic Medical Center) and AMC-8 system, a 130 MHz version of the AMC-8 system, a three-ring AMC-12 system operating at 130 MHz, the BSD SigmaEye applicator and a dipole applicator with three rings each containing six dipole pairs operated at 150 MHz. For five patients with cervix uteri carcinoma, a patient model was created based on a hyperthermia planning CT. Variation of tissue parameters resulted in 16 dielectric models for every patient. In addition, four thermal models were created to study the combined effect of perfusion and dielectric uncertainty. The impact of dielectric uncertainty on optimization is found to be clearly dependent on the number of channels and increased from 0.5 degrees C for four channels to 1.5 degrees C for the 18-channel system. As a result, the potential gain relative to the AMC-4 system for the 70 MHz AMC-8 system was found to be largely compromised, while for the remaining systems a robust improvement in T(90) was observed. The dipole applicator showed the best target heating for two out of five patients, while for three others heating efficacy was comparable to the 130 MHz AMC-12 system or the 130 MHz AMC-8 system (one patient). Considering the increase in complexity when the number of channels is increased from 12 to 18, the AMC-12 system is considered as a good compromise between heating efficacy and robustness while still being a manageable heating system in clinical practice.