Letter to the editor regarding “Application of principal component analysis in clinical gait research” by Federolf and colleagues.
Refereed Original Article
The recent paper by Federolf, Boyer, and Andriacchi (2013) proposes a very interesting approach to 2 analysing gait biomechanics which aims to address a number of challenges, namely: the widespread 3 use of discrete point analysis which utilises only an extremely small percentage of available data (<2 4 %), the coordination and interdependence of movement kinematics or/and kinetics within body and 5 environmental constraints (Bernstein 1967), and the limitation of small sample sizes. The authors 6 propose a new approach based around principal component analysis (PCA) which combines all of the 7 kinematic and kinetic measures into one large vector, rather than (a) examining specific preselected 8 discrete measures, or (b) examining all of the measures but only for one variable at a time (e.g. 9 Harrison, Ryan, and Hayes 2007; Richter et al. 2013). Overall we highly commend the authors for 10 attempting to address these major challenges, and we strongly support the use of PCA to analyse 11 biomechanical data (Richter et al. 2013); however, we believe that the proposed approach contains 12 a number limitations that could preclude its use as applied in the present study.
Digital Object Identifer (DOI):
Date Accepted for Publication:
Tuesday, 8 April, 2014
Journal of Biomechanics
Dublin City University (DCU)
Open access repository: