Artefacts in intracavitary temperature measurements during regional hyperthermia

For adequate hyperthermia treatments, reliable temperature information during treatment is essential. During regional hyperthermia, temperature information is preferably obtained non-invasively from intracavitary or intraluminal measurements to avoid implant risks for the patient. However, for intracavitary or intraluminal thermometry optimal tissue contact is less natural as for invasive thermometry. In this study, the reliability of intraluminal/intracavitary measurements was examined in phantom experiments and in a numerical model for various extents of thermal contact between thermometry and the surroundings. Both thermocouple probes and fibre optic probes were investigated. Temperature rises after a 30 s power pulse of the 70 MHz AMC-4 hyperthermia system were measured in a tissue-equivalent phantom using a multisensor thermocouple probe placed centrally in a hollow tube. The tube was filled with (1) air, (2) distilled water or (3) saline solution that mimics the properties of tissue, simulating situations with (1) bad thermal contact and no power dissipation in the tube, (2) good thermal contact but no power dissipation or (3) good thermal contact and tissue representative power dissipation. For numerical simulations, a cylindrical symmetric model of a thermocouple probe or a fibre optic probe in a cavity was developed. The cavity was modelled as air, distilled water or saline solution. A generalised E-Field distribution was assumed, resulting in a power deposition. With this power deposition, the temperature rise after a 30 s power pulse was calculated. When thermal contact was bad (1), both phantom measurements and simulations with a thermocouple probe showed very high temperature rises (>0.5 degrees C), which are artefacts due to self-heating of the thermocouple probe, since no power is dissipated in air. Simulations with a fibre optic probe showed almost no temperature rise when the cavity was filled with air. When thermal contact was good, but no power was dissipated in the tube (2), artefacts due to self-heating were not significant and the observed temperature rises were very low ( approximately 0-0.1 degrees C). For the situation, with tissue representative power dissipation (3), a temperature rise of approximately 0.23 degrees C was observed for both measurements and simulations. A clinical example of a regional hyperthermia treatment of a patient with a cervix uteri carcinoma showed that the artefacts observed in the case of bad thermal contact also affect the steady-state temperature measurements. Good tissue contact must be assured for reliable intraluminal or intracavitary measurements.