In vivo evaluation of two-dimensional temperature variation in perirenal fat of pigs with B-mode ultrasound

Abstract

Benefiting from their minimally or noninvasive nature, thermal therapies are becoming increasingly important in tumor treatment, in which real-time monitoring of in vivo temperature based on ultrasonic imaging has shown great promise. In this work, an improved dynamic frame selection algorithm and a modified adaptive filtering method were combined with a thermal expansion model, and in vivo temperature monitoring with improved accuracy was achieved. The ultimate aim being the use of thermometry in the thermal treatment of hypertension, experiments targeting the perirenal fat of living pigs were carried out, in which microwaves were applied as a heat source at different power levels. By comparing the echo shift of the ultrasound (US) and the temperature—sensed via a thermocouple—a constant temperature evaluation coefficient was determined. As the tissue was raised to 6.4, 9.8, and 19.3 °C above its base temperature, the root-mean-square evaluation error (ɛrms) was about 0.3, 0.5, and 0.8 °C, respectively. High precision and a high signal-to-noise ratio can help US thermometry play a more important role in monitoring the application of thermal therapies.