Millimeter-Wave Radar-Based Tracheal Respiratory Sound (rTS) Monitoring

Abstract

Conventional bio-radar systems rely solely on gross body motion estimation to understand respiratory activity. However, such methods are fundamentally limited at any given moment; the processed phase signal cannot be reliably mapped to a specific phase of the breathing cycle. Furthermore, involuntary body movements introduce motion artifacts that compromise the integrity of the estimated respiratory patterns, hindering their potential diagnostic usage. In this work, we extend the capabilities of traditional bio-radar systems by introducing tracheal respiratory sound monitoring. By leveraging high-frequency acoustic features embedded in the radar signal, we resolve the inherent phase ambiguity, enabling precise identification of inhalation and exhalation phases. Additionally, we demonstrate, for the first time, accurate estimation of the respiratory flow waveform from radar signals. This unified approach, using a single millimeter-wave (mmWave) radar sensor, enables the extraction of multiple respiratory biomarkers—including breathing rate, motion pattern, breathing phase, and airflow dynamics.