A Machine Learning Pipeline for Gait Analysis in a Semi Free-Living Environment

• Published in: Sensors (Basel, Switzerland) Vol. 23; no. 8; p. 4000
• Main Authors: Jung, Sylvain, de l’Escalopier, Nicolas, Oudre, Laurent, Truong, Charles, Dorveaux, Eric, Gorintin, Louis, Ricard, Damien
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 14.04.2023; MDPI
• Subjects: Algorithms; change point detection; Data mining; Falls; Feedback; free-living; Gait; Gait Analysis; graphical feedback; Health care; Human Activity Recognition; Human acts; Human behavior; Humans; IMU; Life Sciences; Locomotion; Longitudinal studies; Machine Learning; Pathology; Pipe lines; Signal processing; Visualization (Computers); Wearable Electronic Devices; wearable sensor; France; IMU; change point detection; wearable sensor; Human Activity Recognition; free-living; graphical feedback; C; Ricard; E; Jung; L; L. Truong; S. de l'Escalopier; E. Gorintin; D. A Machine free-living wearable sensor IMU graphical feedback change point detection Human Activity Recognition; N; S; L. Ricard; C. Dorveaux; de l'Escalopier; Oudre; Gorintin; Truong; D. A Machine free-living; N. Oudre; Dorveaux
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s23084000

This paper presents a novel approach to creating a graphical summary of a subject’s activity during a protocol in a Semi Free-Living Environment. Thanks to this new visualization, human behavior, in particular locomotion, can now be condensed into an easy-to-read and user-friendly output. As time series collected while monitoring patients in Semi Free-Living Environments are often long and complex, our contribution relies on an innovative pipeline of signal processing methods and machine learning algorithms. Once learned, the graphical representation is able to sum up all activities present in the data and can quickly be applied to newly acquired time series. In a nutshell, raw data from inertial measurement units are first segmented into homogeneous regimes with an adaptive change-point detection procedure, then each segment is automatically labeled. Then, features are extracted from each regime, and lastly, a score is computed using these features. The final visual summary is constructed from the scores of the activities and their comparisons to healthy models. This graphical output is a detailed, adaptive, and structured visualization that helps better understand the salient events in a complex gait protocol.