Electromagnetic-based Correction of Bio-Integrated RFID Sensors for Reliable Skin Temperature Monitoring

Abstract

Bio-integrated wireless sensors in the form of conformable plaster, based on the Radiofrequency Identification (RFID) communication, have been recently proposed for the battery-less measurement of the human skin temperature. However, the response of the Integrated Circuit (IC) transponder is sensitive to the strength of the interrogating power. Indeed, high power produces artifacts on the sampled temperature up to 2 °C when the mutual position between reader and sensors, as well as the emitted power, can not be carefully controlled. Hence, a reliable adoption of this technology in real cases is challenging and still in question. A combined macro-scale electromagnetic-thermal model is here introduced to predict and correct the above artifact so that the temperature measurement becomes insensitive to the RF power collected by the IC. The method is based on the new generation RFID ICs with on-chip temperature sensor that are also capable to give back the strength of the collected RF power. The model is validated in controlled conditions and then applied for different skin temperature measurements on human body. An average accuracy of ±0.25 °C, compared with a reference calibrated thermocouple, was demonstrated in the considered tests.