Parallel Factor Analysis of gait waveform data: A multimode extension of Principal Component Analysis

Nathaniel E. Helwig, Sungjin Hong, John D. Polk

Research output: Contribution to journalArticle

Abstract

Gait data are typically collected in multivariate form, so some multivariate analysis is often used to understand interrelationships between observed data. Principal Component Analysis (PCA), a data reduction technique for correlated multivariate data, has been widely applied by gait analysts to investigate patterns of association in gait waveform data (e.g., interrelationships between joint angle waveforms from different subjects and/or joints). Despite its widespread use in gait analysis, PCA is for two-mode data, whereas gait data are often collected in higher-mode form. In this paper, we present the benefits of analyzing gait data via Parallel Factor Analysis (Parafac), which is a component analysis model designed for three- or higher-mode data. Using three-mode joint angle waveform data (subjects × time × joints), we demonstrate Parafac's ability to (a) determine interpretable components revealing the primary interrelationships between lower-limb joints in healthy gait and (b) identify interpretable components revealing the fundamental differences between normal and perturbed subjects' gait patterns across multiple joints. Our results offer evidence of the complex interconnections that exist between lower-limb joints and limb segments in both normal and abnormal gaits, confirming the need for the simultaneous analysis of multi-joint gait waveform data (especially when studying perturbed gait patterns).

Original languageEnglish (US)
Pages (from-to)630-648
Number of pages19
JournalHuman Movement Science
Volume31
Issue number3
DOIs
StatePublished - Jun 1 2012

Fingerprint

Principal Component Analysis
Gait
Statistical Factor Analysis
Joints
Lower Extremity
Multivariate Analysis
Extremities

Keywords

  • Locomotion
  • Multivariate analysis
  • Parallel Factor Analysis
  • Principal Component Analysis
  • Walking

ASJC Scopus subject areas

  • Biophysics
  • Orthopedics and Sports Medicine
  • Experimental and Cognitive Psychology

Cite this

Parallel Factor Analysis of gait waveform data : A multimode extension of Principal Component Analysis. / Helwig, Nathaniel E.; Hong, Sungjin; Polk, John D.

In: Human Movement Science, Vol. 31, No. 3, 01.06.2012, p. 630-648.

Research output: Contribution to journalArticle

@article{6461ac31818243a28e552190377d678c,
title = "Parallel Factor Analysis of gait waveform data: A multimode extension of Principal Component Analysis",
abstract = "Gait data are typically collected in multivariate form, so some multivariate analysis is often used to understand interrelationships between observed data. Principal Component Analysis (PCA), a data reduction technique for correlated multivariate data, has been widely applied by gait analysts to investigate patterns of association in gait waveform data (e.g., interrelationships between joint angle waveforms from different subjects and/or joints). Despite its widespread use in gait analysis, PCA is for two-mode data, whereas gait data are often collected in higher-mode form. In this paper, we present the benefits of analyzing gait data via Parallel Factor Analysis (Parafac), which is a component analysis model designed for three- or higher-mode data. Using three-mode joint angle waveform data (subjects × time × joints), we demonstrate Parafac's ability to (a) determine interpretable components revealing the primary interrelationships between lower-limb joints in healthy gait and (b) identify interpretable components revealing the fundamental differences between normal and perturbed subjects' gait patterns across multiple joints. Our results offer evidence of the complex interconnections that exist between lower-limb joints and limb segments in both normal and abnormal gaits, confirming the need for the simultaneous analysis of multi-joint gait waveform data (especially when studying perturbed gait patterns).",
keywords = "Locomotion, Multivariate analysis, Parallel Factor Analysis, Principal Component Analysis, Walking",
author = "Helwig, {Nathaniel E.} and Sungjin Hong and Polk, {John D.}",
year = "2012",
month = "6",
day = "1",
doi = "10.1016/j.humov.2011.06.011",
language = "English (US)",
volume = "31",
pages = "630--648",
journal = "Human Movement Science",
issn = "0167-9457",
publisher = "Elsevier",
number = "3",

}

TY - JOUR

T1 - Parallel Factor Analysis of gait waveform data

T2 - A multimode extension of Principal Component Analysis

AU - Helwig, Nathaniel E.

AU - Hong, Sungjin

AU - Polk, John D.

PY - 2012/6/1

Y1 - 2012/6/1

N2 - Gait data are typically collected in multivariate form, so some multivariate analysis is often used to understand interrelationships between observed data. Principal Component Analysis (PCA), a data reduction technique for correlated multivariate data, has been widely applied by gait analysts to investigate patterns of association in gait waveform data (e.g., interrelationships between joint angle waveforms from different subjects and/or joints). Despite its widespread use in gait analysis, PCA is for two-mode data, whereas gait data are often collected in higher-mode form. In this paper, we present the benefits of analyzing gait data via Parallel Factor Analysis (Parafac), which is a component analysis model designed for three- or higher-mode data. Using three-mode joint angle waveform data (subjects × time × joints), we demonstrate Parafac's ability to (a) determine interpretable components revealing the primary interrelationships between lower-limb joints in healthy gait and (b) identify interpretable components revealing the fundamental differences between normal and perturbed subjects' gait patterns across multiple joints. Our results offer evidence of the complex interconnections that exist between lower-limb joints and limb segments in both normal and abnormal gaits, confirming the need for the simultaneous analysis of multi-joint gait waveform data (especially when studying perturbed gait patterns).

AB - Gait data are typically collected in multivariate form, so some multivariate analysis is often used to understand interrelationships between observed data. Principal Component Analysis (PCA), a data reduction technique for correlated multivariate data, has been widely applied by gait analysts to investigate patterns of association in gait waveform data (e.g., interrelationships between joint angle waveforms from different subjects and/or joints). Despite its widespread use in gait analysis, PCA is for two-mode data, whereas gait data are often collected in higher-mode form. In this paper, we present the benefits of analyzing gait data via Parallel Factor Analysis (Parafac), which is a component analysis model designed for three- or higher-mode data. Using three-mode joint angle waveform data (subjects × time × joints), we demonstrate Parafac's ability to (a) determine interpretable components revealing the primary interrelationships between lower-limb joints in healthy gait and (b) identify interpretable components revealing the fundamental differences between normal and perturbed subjects' gait patterns across multiple joints. Our results offer evidence of the complex interconnections that exist between lower-limb joints and limb segments in both normal and abnormal gaits, confirming the need for the simultaneous analysis of multi-joint gait waveform data (especially when studying perturbed gait patterns).

KW - Locomotion

KW - Multivariate analysis

KW - Parallel Factor Analysis

KW - Principal Component Analysis

KW - Walking

UR - http://www.scopus.com/inward/record.url?scp=84863836941&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84863836941&partnerID=8YFLogxK

U2 - 10.1016/j.humov.2011.06.011

DO - 10.1016/j.humov.2011.06.011

M3 - Article

C2 - 21925756

AN - SCOPUS:84863836941

VL - 31

SP - 630

EP - 648

JO - Human Movement Science

JF - Human Movement Science

SN - 0167-9457

IS - 3

ER -