Human movement analysis as a measure for fatigue: A hidden markov-based approach

Michelle Karg, Gentiane Venture, Jesse Hoey, Dana Kulic

Research output: Contribution to journalArticleResearchpeer-review

26 Citations (Scopus)

Abstract

Fatigue influences the way a training exercise is performed and alters the kinematics of the movement. Monitoring the increase of fatigue during rehabilitation and sport exercises is beneficial to avoid the risk of injuries. This study investigates the use of a parametric hidden Markov model (PHMM) to estimate fatigue from observing kinematic changes in the way the exercise is performed. The PHMM is compared to linear regression. A top-level hidden Markov model with variable state transitions incorporates knowledge about the progress of fatigue during the exercise and the initial condition of a subject. The approach is tested on a squat database recorded with optical motion capture. The estimates of fatigue for a single squat, a set of squats, and an entire exercise correlate highly with subjective ratings.

Original languageEnglish
Article number6716986
Pages (from-to)470-481
Number of pages12
JournalIEEE Transactions on Neural Systems and Rehabilitation Engineering
Volume22
Issue number3
DOIs
Publication statusPublished - 1 May 2014
Externally publishedYes

Keywords

  • Fatigue
  • linear regression
  • parametric hidden Markov model

Cite this