Validity and Consistency of Concurrent Extraction of Gait Features Using Inertial Measurement Units and Motion Capture System

Anwary, Arif Reza; Yu, Hongnian; Callaway, Andrew; Vassallo, Michael

Abstract

Conditions causing gait abnormalities are very common and their treatment requires the detailed assessment of gait. Currently such assessments are carried out in gait laboratories and require the use of complex and expensive equipment. To increase availability and use at home and clinics, we design and develop an affordable, user friendly, wireless, portable automatic system to extract spatiotemporal features of gait that can be used indoors and outdoors. This study determines the concurrent validity of extracted gait features from Inertial Measurement Units (IMUs) against ‘gold standard’ Motion Capture System (MoCap) using a hybrid gait features extraction method. The analysis of the proposed method is based on minimum prominence and abrupt transition points in the IMU signals. We also compare the degree of agreement for mean spatiotemporal gait features. The concurrent data from synchronized IMUs and MoCap are collected from 18 subjects. We validate our proposed system using two experiments; 1) IMU and MoCap with self-selected (free) walking and 2) IMU and MoCap at various walking speeds. Interclass correlations, Lin’s concordance correlation coefficients and Pearson’s correlation coefficients (r) are applied to determine the correlation between extracted gait features from IMU and MoCap measurements. Bland-Altman plots are also generated to evaluate any unknown bias between the mean extracted features. The experiments show that spatiotemporal features of gait extracted from IMUs are highly valid. Our methods facilitate gait assessment in clinics and at home including the possibility of self-assessment.

• Journal Article Type: Article
• Acceptance Date: Sep 1, 2020
• Online Publication Date: Sep 3, 2020
• Publication Date: 2021-01
• Deposit Date: Jan 14, 2021
• Publicly Available Date: Jan 14, 2021
• Journal: IEEE Sensors Journal
• Print ISSN: 1530-437X
• Electronic ISSN: 1558-1748
• Publisher: Institute of Electrical and Electronics Engineers
• Peer Reviewed: Peer Reviewed
• Volume: 21
• Issue: 2
• Pages: 1625-1634
• DOI: https://doi.org/10.1109/jsen.2020.3021501
• Keywords: Inertial measurement unit (IMU), accelerometer, gyroscope, feature extraction, wearable sensors, gait analysis
• Public URL: http://researchrepository.napier.ac.uk/Output/2715115