Measuring Coupling of Rhythmical Time Series Using Cross Sample Entropy and Cross Recurrence Quantification Analysis

John Mccamley, William Denton, Elizabeth Lyden, Jennifer M. Yentes

Research output: Contribution to journalArticlepeer-review

17 Scopus citations

Abstract

The aim of this investigation was to compare and contrast the use of cross sample entropy (xSE) and cross recurrence quantification analysis (cRQA) measures for the assessment of coupling of rhythmical patterns. Measures were assessed using simulated signals with regular, chaotic, and random fluctuations in frequency, amplitude, and a combination of both. Biological data were studied as models of normal and abnormal locomotor-respiratory coupling. Nine signal types were generated for seven frequency ratios. Fifteen patients with COPD (abnormal coupling) and twenty-one healthy controls (normal coupling) walked on a treadmill at three speeds while breathing and walking were recorded. xSE and the cRQA measures of percent determinism, maximum line, mean line, and entropy were quantified for both the simulated and experimental data. In the simulated data, xSE, percent determinism, and entropy were influenced by the frequency manipulation. The 1: 1 frequency ratio was different than other frequency ratios for almost all measures and/or manipulations. The patients with COPD used a 2: 3 ratio more often and xSE, percent determinism, maximum line, mean line, and cRQA entropy were able to discriminate between the groups. Analysis of the effects of walking speed indicated that all measures were able to discriminate between speeds.

Original languageEnglish (US)
Article number7960467
JournalComputational and Mathematical Methods in Medicine
Volume2017
DOIs
StatePublished - 2017
Externally publishedYes

ASJC Scopus subject areas

  • Modeling and Simulation
  • General Biochemistry, Genetics and Molecular Biology
  • General Immunology and Microbiology
  • Applied Mathematics

Fingerprint

Dive into the research topics of 'Measuring Coupling of Rhythmical Time Series Using Cross Sample Entropy and Cross Recurrence Quantification Analysis'. Together they form a unique fingerprint.

Cite this