Your browser doesn't support javascript.
loading
Empirical Mode Decomposition-Derived Entropy Features Are Beneficial to Distinguish Elderly People with a Falling History on a Force Plate Signal.
Chou, Li-Wei; Chang, Kang-Ming; Wei, Yi-Chun; Lu, Mei-Kuei.
Afiliação
  • Chou LW; Department of Physical Medicine and Rehabilitation, China Medical University Hospital, Taichung City 40402, Taiwan.
  • Chang KM; Department of Physical Medicine and Rehabilitation, Asia University Hospital, Asia University, Taichung City 41354, Taiwan.
  • Wei YC; Department of Physical Therapy and Graduate Institute of Rehabilitation Science, China Medical University, Taichung City 40402, Taiwan.
  • Lu MK; Department of Medical Research, China Medical University Hospital, China Medical University, Taichung City 40402, Taiwan.
Entropy (Basel) ; 23(4)2021 Apr 16.
Article em En | MEDLINE | ID: mdl-33923557
Fall risk prediction is an important issue for the elderly. A center of pressure signal, derived from a force plate, is useful for the estimation of body calibration. However, it is still difficult to distinguish elderly people's fall history by using a force plate signal. In this study, older adults with and without a history of falls were recruited to stand still for 60 s on a force plate. Forces in the x, y and z directions (Fx, Fy, and Fz) and center of pressure in the anteroposterior (COPx) and mediolateral directions (COPy) were derived. There were 49 subjects in the non-fall group, with an average age of 71.67 (standard derivation: 6.56). There were also 27 subjects in the fall group, with an average age of 70.66 (standard derivation: 6.38). Five signal series-forces in x, y, z (Fx, Fy, Fz), COPX, and COPy directions-were used. These five signals were further decomposed with empirical mode decomposition (EMD) with seven intrinsic mode functions. Time domain features (mean, standard derivation and coefficient of variations) and entropy features (approximate entropy and sample entropy) of the original signals and EMD-derived signals were extracted. Results showed that features extracted from the raw COP data did not differ significantly between the fall and non-fall groups. There were 10 features extracted using EMD, with significant differences observed among fall and non-fall groups. These included four features from COPx and two features from COPy, Fx and Fz.
Palavras-chave

Texto completo: 1 Base de dados: MEDLINE Tipo de estudo: Prognostic_studies Idioma: En Ano de publicação: 2021 Tipo de documento: Article

Texto completo: 1 Base de dados: MEDLINE Tipo de estudo: Prognostic_studies Idioma: En Ano de publicação: 2021 Tipo de documento: Article