Experimental characterization of the miniannular phased array as a hyperthermia applicator

Noninvasive temperature imaging using diffusion MRI

Efficacy and safety considerations for cancer therapy with hyperthermia require accurate temperature measurements throughout the heated volume. We report the use of molecular diffusion, whose temperature dependence is well known. A dedicated hyperthermia applicator was built, combining a MRI gradient coil and a rf coil. Diffusion and derived temperature images were obtained with a 1 x 2 mm pixel size on a polyacrylamide gel phantom using a clinical 1.5-T whole body MRI system. Temperatures determined from these images using 1 cm2 regions of interest were found to be within 0.2 degrees C of those recorded from the thermocouples and fiber-optic probes placed inside the gel.

A ring capacitor applicator in hyperthermia: energy distributions in a fat-muscle layered model for different ring electrode configurations

The energy deposition pattern within a radially layered fat-muscle phantom, diameter 135 mm, heated by a novel ring capacitor applicator has been determined experimentally as well as theoretically. Good to excellent agreement is found between measured and predicted energy distributions. For the specific absorption rate in the muscle tissue the differences are in general smaller than 6%. When the ring electrodes are placed directly on the phantom surface both measured and predicted energy distributions show the presence of superficial hot spots located within the fat layer at the site of the ring electrodes. The theoretical distributions showed that the radial component of the E-field contributes for more than 90% to the energy absorption at the hot spot in the fatty tissue in front of the ring electrodes. Introducing a small air gap (10 mm) between the phantom surface and the ring electrode results in a decrease of the energy absorption within the fatty tissue at the hot spot location by 30%. Further theoretical analysis of the energy distribution within the inhomogeneous model showed that the intensity of the hot spots at the ring electrodes can be controlled by adjustment of the applicator configuration. Independent of the size of the electrode to phantom gap the specific absorption rate values predicted in the fat-muscle model show a more favorable distribution at a frequency of 27.12 MHz than at 13.56 MHz. For a similar electrode to phantom gap the specific absorption rate within the fatty tissue is approximately two times lower at 27.12 than at 13.56 MHz. For the model calculations performed the best ratio of fat to muscle SAR (0.2) is obtained with distilled water as bolus medium in the gap.