The effect of signal acquisition and processing choices on ApEn values: Towards a "gold standard" for distinguishing effort levels from isometric force records

Sarah M. Forrest, John H. Challis, Samantha L. Winter

Research output: Contribution to journalArticle

14 Scopus citations

Abstract

Approximate entropy (ApEn) is frequently used to identify changes in the complexity of isometric force records with ageing and disease. Different signal acquisition and processing parameters have been used, making comparison or confirmation of results difficult. This study determined the effect of sampling and parameter choices by examining changes in ApEn values across a range of submaximal isometric contractions of the first dorsal interosseus. Reducing the sample rate by decimation changed both the value and pattern of ApEn values dramatically. The pattern of ApEn values across the range of effort levels was not sensitive to the filter cut-off frequency, or the criterion used to extract the section of data for analysis. The complexity increased with increasing effort levels using a fixed '. r' value (which accounts for measurement noise) but decreased with increasing effort level when '. r' was set to 0.1 of the standard deviation of force. It is recommended isometric force records are sampled at frequencies >200. Hz, template length ('. m') is set to 2, and '. r' set to measurement system noise or 0.1. SD depending on physiological process to be distinguished. It is demonstrated that changes in ApEn across effort levels are related to changes in force gradation strategy.

Original languageEnglish (US)
Pages (from-to)676-683
Number of pages8
JournalMedical Engineering and Physics
Volume36
Issue number6
DOIs
StatePublished - Jun 2014

All Science Journal Classification (ASJC) codes

  • Biophysics
  • Biomedical Engineering

Fingerprint Dive into the research topics of 'The effect of signal acquisition and processing choices on ApEn values: Towards a "gold standard" for distinguishing effort levels from isometric force records'. Together they form a unique fingerprint.

  • Cite this