Respiratory rate is a relevant parameter to be monitored to prevent adverse cardiopulmonary situations. Radar sensing represents a promising alternative to wearable sensors, since it gives the opportunity to remotely monitor a subject, while maintaining privacy and fairness. In this work, radar is used to enable a single sensor to track and extract the respiratory rate of a freely moving subject. To validate performance, the subject first walks in a straight line and then moves randomly within a room. As a reference, they are instructed to control their breathing pace using a metronome. A state of art deep learning algorithm, previously developed for person detection purposes and called mmDetect, is applied to localize and track the subject. The developed processing pipeline allows the extraction of the beat signal phase shift associated with chest displacement at each location, which is then used to reconstruct a phase signal linked to the respiratory phenomenon. A resulting measurement error of 1 breath per minute (BPM) is demonstrated.
Indoor Tracking for Motion-Robust Respiration Rate Extraction with FMCW Radar / Raimondi, M., Nocera, A., Ciattaglia, G., Senigagliesi, L., Gambi, E.. - (2025). (30th IEEE Symposium on Computers and Communications, ISCC 2025 Bologna, Italy 2 - 5 July 2025) [10.1109/ISCC65549.2025.11326268].
Indoor Tracking for Motion-Robust Respiration Rate Extraction with FMCW Radar
Raimondi M.Primo
;Nocera A.;Ciattaglia G.;Gambi E.Ultimo
2025-01-01
Abstract
Respiratory rate is a relevant parameter to be monitored to prevent adverse cardiopulmonary situations. Radar sensing represents a promising alternative to wearable sensors, since it gives the opportunity to remotely monitor a subject, while maintaining privacy and fairness. In this work, radar is used to enable a single sensor to track and extract the respiratory rate of a freely moving subject. To validate performance, the subject first walks in a straight line and then moves randomly within a room. As a reference, they are instructed to control their breathing pace using a metronome. A state of art deep learning algorithm, previously developed for person detection purposes and called mmDetect, is applied to localize and track the subject. The developed processing pipeline allows the extraction of the beat signal phase shift associated with chest displacement at each location, which is then used to reconstruct a phase signal linked to the respiratory phenomenon. A resulting measurement error of 1 breath per minute (BPM) is demonstrated.| File | Dimensione | Formato | |
|---|---|---|---|
|
Raimondi_Indoor-Tracking-Motion-Robust_2025.pdf
Solo gestori archivio
Tipologia:
Versione editoriale (versione pubblicata con il layout dell'editore)
Licenza d'uso:
Tutti i diritti riservati
Dimensione
8.25 MB
Formato
Adobe PDF
|
8.25 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


