Citation

Angelini L, Buckley E, Bonci T, Radford A, Sharrack B, Paling D, Nair KPS, Mazzà C. IEEE Trans. Biomed. Eng. 2021; ePub(ePub): ePub.

Copyright

(Copyright © 2021, Institute of Electrical and Electronic Engineers)

DOI

10.1109/TBME.2021.3061998

PMID

unavailable

Abstract

OBJECTIVE: Mobility assessment is critical in the clinical management of people with Multiple Sclerosis (pwMS). Instrumented gait analysis provides a plethora of metrics for quantifying concurrent factors contributing to gait deterioration. However, a gait model discriminating underlying features contributing to this deterioration is lacking in pwMS. This study aimed at developing and validating such a model.

METHODS: The gait of 24 healthy controls and 114 pwMS with mild, moderate, or severe disability was measured with inertial sensors on the shanks and lower trunk while walking for 6 minutes along a hospital corridor. Twenty out of thirty-six initially explored metrics computed from the sensor data met the quality criteria for exploratory factor analysis. This analysis provided the sought model, which underwent a confirmatory factor analysis before being used to characterize gait impairment across the three disability groups.

RESULTS: A gait model consisting of five domains (rhythm/variability, pace, asymmetry, and forward and lateral dynamic balance) was revealed by the factor analysis, which was able to highlight gait abnormalities across the disability groups: significant alterations in rhythm/variability-, asymmetry-, and pace-based features were present in the mild group, but these were more profound in the moderate and severe groups. Deterioration in dynamic balance-based features was only noted in pwMS with a moderate and severe disability.

CONCLUSION: A conceptual model of gait for disease-specific mobility assessment in pwMS was successfully developed and tested. SIGNIFICANCE: The new model, built with metrics that represent gait impairment in pwMS, highlighted clinically relevant changes across different disability levels, including those with no clinically observable walking disability. This shows the clear potential as a monitoring biomarker in pwMS.

Language: en