Identifying coordinative structure using principal component analysis based on coherence derived from linear systems analysis

Xinguang Wang, Nicholas O'Dwyer, Mark Halaki, Richard Smith

Research output: Contribution to journalArticlepeer-review

9 Citations (Scopus)

Abstract

Principal component analysis is a powerful and popular technique for capturing redundancy in muscle activity and kinematic patterns. A primary limitation of the correlations or covariances between signals on which this analysis is based is that they do not account for dynamic relations between signals, yet such relationsâ€Â'such as that between neural drive and muscle tensionâ€Â'are widespread in the sensorimotor system. Low correlations may thus be obtained and signals may appear independent despite a dynamic linear relation between them. To address this limitation, linear systems analysis can be used to calculate the matrix of overall coherences between signals, which measures the strength of the relation between signals taking dynamic relations into account. Using ankle, knee, and hip sagittal-plane angles from 6 healthy subjects during overground walking at preferred speed, it is shownthat with conventional correlation matrices the first principal component accounted for ∼50% of total variance in the data set, while with overall coherence matrices the first component accounted for > 95% of total variance. The results demonstrate that the dimensionality of the coordinative structure can be overestimated using conventional correlation, whereas a more parsimonious structure is identified with overall coherence.
Original languageEnglish
Pages (from-to)167-179
Number of pages13
JournalJournal of Motor Behavior
Volume45
Issue number2
DOIs
Publication statusPublished - 2013

Fingerprint

Dive into the research topics of 'Identifying coordinative structure using principal component analysis based on coherence derived from linear systems analysis'. Together they form a unique fingerprint.

Cite this