Smart Vest And Adaptive Algorithm For Vital Signs And Physical Activity Monitoring: A Feasibility Study

Abstract

Monitoring physiological and biomechanical variables plays an important role in assessing human health in several scenarios, including sports. In this context, there is an increasing demand for wearable devices that can potentially provide heterogeneous information about the athlete’s health state, with the challenge of meeting the athlete’s needs in terms of comfort.In this paper, we propose a low-cost and unobtrusive wearable smart vest and a related adaptive algorithm to estimate both physiological variables such as heart rate (HR) and respiratory rate (RR) and the pace cadence. The sensing part of our smart vest consists of a single Inertial Measurement Unit (IMU) embedding both an accelerometer (ACC) and a gyroscope (GYR) positioned on the thorax of the subject. Our device was tested on four volunteers during both at-rest postures, walking and running at different speeds. Results show that the GYR outperforms the ACC for both HR and RR estimation during at-rest postures taken up before performing the running exercise. Average Mean absolute error (MAE) values of 1.22 bpm and 0.39 bpm have been achieved for ACC and GYR in HR monitoring in the pre-exercise phase; Average MAE of 3.59 breaths/min and 0.36 breaths/min have been achieved for ACC and GYR in RR monitoring in pre-exercise phase. Also, after the exercise protocol, the results are very promising, with average MAE of 0.24 bpm and 0.25 bpm using ACC and GYR for HR estimation and up to 1.60 breaths/min and 0.15 breaths/min for ACC and GYR in RR monitoring, respectively. Moreover, the pace cadence estimated by our system matches all the protocol phases required for the volunteer. The obtained results support the feasibility of estimating HR, RR, and cadence by using the sensors inside the smart vest in daily life and sports activities.