Towards an optimal combination of input signals and derived representations for gait classification based on ground reaction force measurements

The application of machine learning in clinical gait analysis has gained popularity in recent years, e.g. for automatic identification of gait patterns [1]. The main purpose of these methods is to support clinicians during medical decision-making processes. In practice, classification performance is strongly influenced by various factors, such as the employed input signals and the selection of derived signal representations. There is currently no best-practice recommendation available on this topic.
© Copyright 2010 Gait and Posture. Elsevier. All rights reserved.

Bibliographic Details
Subjects:
Notations:biological and medical sciences
Published in:Gait and Posture
Language:English
Published: 2010
Online Access:https://doi.org/10.1016/j.gaitpost.2018.06.155
Volume:65
Issue:S 1
Pages:249-250
Document types:article
Level:advanced