AbstractIn human life, respiration serves as a crucial physiological signal. Continuous real‐time respiration monitoring can provide valuable insights for the early detection and management of several respiratory diseases. High‐sensitivity, noninvasive, comfortable, and long‐term stable respiration devices are highly desirable. In spite of this, existing respiration sensors cannot provide continuous long‐term monitoring due to the erosion from moisture, fluctuations in body temperature, and many other environmental factors. This research developed a wearable thermal‐based respiration sensor made of cubic silicon carbide (3C‐SiC) using a microfabrication process. The results showed that as a result of the Joule heating effect in the robustness 3C‐SiC material, the sensor offered high sensitivity with the negative temperature coefficient of resistance of approximately 5,200ppmK‐1, an excellent response to respiration and long‐term stability monitoring. Furthermore, by incorporating a deep learning model, this fabricated sensor can develop advanced capabilities to distinguish between the four distinct breath patterns: slow, normal, fast, and deep breathing, and provide an impressive classification accuracy rate of ≈ 99.7%. The results of this research represent a significant step in developing wearable respiration sensors for personal healthcare systems.
Read full abstract