The Analytical Validity of Stride Detection and Gait Parameters Reconstruction Using the Ankle-Mounted Inertial Measurement Unit Syde

• Published in: Sensors (Basel, Switzerland) Vol. 24; no. 8; p. 2413
• Main Authors: Michaud, Mona, Guérin, Alexandre, Dejean de La Bâtie, Marguerite, Bancel, Léopold, Oudre, Laurent, Tricot, Alexis
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 10.04.2024; MDPI
• Subjects: Accelerometry - instrumentation; Accelerometry - methods; Adult; Algorithms; Analysis; analytical validation; Ankle - physiology; Biomechanical Phenomena - physiology; Cameras; Clinical trials; controlled environment; Data collection; Engineering Sciences; Exercise; Falls (Accidents); Female; Gait; Gait - physiology; Gait Analysis - instrumentation; Gait Analysis - methods; gait monitoring; Humans; IMU; Laboratories; Male; Measurement; Medical care; Medical research; Medicine, Experimental; Motion capture; Muscular dystrophy; Patients; Quality management; Sensors; Signal and Image processing; Software; Velocity; wearable; Wearable computers; Wearable Electronic Devices; Young Adult; France; IMU; wearable; gait monitoring; controlled environment; motion capture; analytical validation
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s24082413

The increasing use of inertial measurement units (IMU) in biomedical sciences brings new possibilities for clinical research. The aim of this paper is to demonstrate the accuracy of the IMU-based wearable Syde® device, which allows day-long and remote continuous gait recording in comparison to a reference motion capture system. Twelve healthy subjects (age: 23.17 ± 2.04, height: 174.17 ± 6.46 cm) participated in a controlled environment data collection and performed a series of gait tasks with both systems attached to each ankle. A total of 2820 strides were analyzed. The results show a median absolute stride length error of 1.86 cm between the IMU-based wearable device reconstruction and the motion capture ground truth, with the 75th percentile at 3.24 cm. The median absolute stride horizontal velocity error was 1.56 cm/s, with the 75th percentile at 2.63 cm/s. With a measurement error to the reference system of less than 3 cm, we conclude that there is a valid physical recovery of stride length and horizontal velocity from data collected with the IMU-based wearable Syde® device.