Evaluation of an ingestible telemetric temperature sensor for deep hyperthermia applications

We have investigated the potential of an ingestible thermometric system (ITS) for use with a deep heating system. The ingestible sensor contains a temperature-sensitive quartz crystal oscillator. The telemetered signal is inductively coupled by a radiofrequency coil system to an external receiver. The sensors, covered with a protective silicon coating, are 10 mm in diameter and 20 mm long and are energized by an internal silver-oxide battery. Experimental studies were carried out to investigate the accuracy of the system and the extent of reliable operation of these sensors in an electromagnetic environment. Different measurements were repeated for five sensors. Calibration accuracy was verified by comparison with a Bowman probe in the temperature range 30°C to 55°C. Linear regression analysis of individual pill readings indicated a correlation within ± 0.4°C at 95% prediction intervals in the clinical temperature range of 35°C to 50°C. Further work is required to improve this accuracy to meet the quality assurance guidelines of ± 0.2°C suggested by the Hyperthermia Physics Center. Response times were determined by the exponential fit of heat-up and cool-down curves for each pill. All curves had correlation coefficients greater than 0.98. Time (mean ± SE) to achieve 90% response during heat-up was 115 ± 8 sec. Time to cool-down to 10% of initial temperature was 11 4 ± sec. The effect of the external antenna and sensor spacing and the angle of orientation of the sensor relative to the antenna plane were also studied. Electromagnetic interference effects were studied by placing the sensor with a Bowman probe in a cylindrical saline phantom for the tests in an annular phase array applicator. Different power levels at three frequencies-80, 100, and 120 MHz-were used. Accurate temperature readings could not be obtained when the electromagnetic power was on because of interference effects with the receiver. However, the temperatures read with the ITS immediately after the electromagnetic power was switched off correlated well with the Bowman probe readings across the power categories and the three frequencies used. The phantom was heated to steady state, with a Bowman probe placed at the central axis of the cylinder used as control. During the heat-up period and the steady state, the mean difference (± SE) between the ITS and Bowman probe was 0.12°C (± 0.05°C). We have used ITS to measure temperatures in the human intestine within the hyperthermia field during deep-heating sessions of abdominal tumors. Laboratory and clinical data, and possible future applications of this technology in monitoring temperature during deep hyperthermia, are discussed.