Influence of motion artifacts on a smart garment for monitoring respiratory rate

Wearable devices are gaining large acceptance in the continuous monitoring of vital signs. Among others, respiratory rate (fR) can be used to detect physiological abnormalities and health status changes.The purpose of this work was to investigate how the output of a smart garment used for respiratory monitoring is influenced by walking and running. This garment consists of three bands, each one embeds two piezoresistive elements sensitive to strain.Experimental trials were carried out on a volunteer who worn the three bands at the level of upper thorax, inferior thorax and abdomen during three different activities (i.e., static pose, walking and running). A treadmill was used to set specific speeds (i.e., from 1.6 km•h-1 to 8.8 km•h-1). The fR values estimated by the proposed garment were compared to the ones monitored by a reference system (i.e., a flowmeter).The analysis in the frequency-domain demonstrated differences up to 3 bpm between the average fR estimated by the two systems. The mean absolute error (MAE) was used to investigate the performances of the garment against the reference device in estimating the instantaneous fR. MAE increased with speed (it reached 1.8 bpm during running). Bland-Altman analysis showed a bias of -0.02±2.02 bpm when all the data of walking and running were considered.The garment based on 6 sensing elements provides good performances for estimating both average and instantaneous fR values during activities of walking and running.