A Wireless Body Sensor Network for Cardiorespiratory Monitoring During Cycling

In this work, a completely wearable Wireless Body Sensor Network (WBSN) for cardiorespiratory monitoring is presented. The WBSN consists of three Inertial Measurement Units (IMUs) for respiratory signal detection and a pulse oximeter for pulse rate measurements. The devices communicate using the ANT protocol, with a smartphone serving as the master node. The system was preliminarily validated against gold standard measurements using an experimental protocol involving cycling exercises. The results showed good agreement between the IMU-based system and the gold standard methods for both respiratory rate (RR) and pulse rate (PR) measurements, i.e., the COSMED K5 metabolic cart and a Garmin chest strap, respectively. The RR estimation had a mean difference of −0.85, with limits of agreement of [−6.76; +5.06]. The PR estimation had a mean difference of 1.80, with limits of agreement of [−8.80; +10.96]. The ad hoc filtering algorithm used for RR estimation based on the IMU data contributed to the accuracy of the system. While the validation results were positive, further studies with larger sample sizes and diverse activities are needed to confirm the validity of the system.