Step detection using multi- versus single tri-axial accelerometer-based systems

• Published in: Physiological measurement Vol. 36; no. 12; pp. 2519 - 2535
• Main Authors: Fortune, E, Lugade, VA, Amin, S, Kaufman, K R
• Format: Journal Article
• Language: English
• Published: England IOP Publishing 01.12.2015
• Subjects: accelerometer; Accelerometry - instrumentation; Adult; Algorithms; body-worn sensors; Female; Humans; Male; Middle Aged; movement analysis; Posture; sensor location; step detection; Walking; Young Adult
• ISSN: 0967-3334; 1361-6579; 1361-6579
• DOI: 10.1088/0967-3334/36/12/2519

Multiple sensors are often considered necessary for increased step count accuracy. However, subject adherence to device-wear increases using a minimal number of activity monitors (AMs). The study aims were to determine and compare the validity of using multiple AMs versus a single AM to detect steps by comparison to video using a modification of an algorithm previously developed for a four-accelerometer AM system capable, unlike other algorithms, of accurate step detection for gait velocities as low as 0.1 m s−1. Twelve healthy adults wore ankle, thigh and waist AMs while performing walking/jogging trials at gait velocities from 0.1-4.8 m s−1 and a simulated free-living dynamic activities protocol. Nineteen older adults wore ankle and waist AMs while walking at velocities from 0.5-2.0 m s−1. As little as one AM (thigh or waist) accurately detected steps for velocities  >0.5 m s−1. A single ankle AM accurately detected steps for velocities   0.1 m s−1. Only the thigh AM could not accurately detect steps during the dynamic activities. Only the thigh-ankle combination or single waist AM could accurately distinguish between walking and jogging steps. These laboratory-based results suggest that the presented algorithm can accurately detect steps in a free-living environment using only one ankle or waist AM.