Elsevier

Gait & Posture

Volume 28, Issue 3, October 2008, Pages 351-357
Gait & Posture

ESMAC 2007 Best Paper Award
The gait deviation index: A new comprehensive index of gait pathology

https://doi.org/10.1016/j.gaitpost.2008.05.001Get rights and content

Abstract

This article describes a new multivariate measure of overall gait pathology called the Gait Deviation Index (GDI). The first step in developing the GDI was to use kinematic data from a large number of walking strides to derive a set of mutually independent joint rotation patterns that efficiently describe gait. These patterns are called gait features. Linear combinations of the first 15 gait features produced a 98% faithful reconstruction of both the data from which they were derived and 1000 validation strides not used in the derivation. The GDI was then defined as a scaled distance between the 15 gait feature scores for a subject and the average of the same 15 gait feature scores for a control group of typically developing (TD) children. Concurrent and face validity data for the GDI are presented through comparisons with the Gillette Gait Index (GGI), Gillette Functional Assessment Questionnaire Walking Scale (FAQ), and topographic classifications within the diagnosis of Cerebral Palsy (CP). The GDI and GGI are strongly correlated (r2 = 0.56). The GDI scales with FAQ level, distinguishes levels from one another, and is normally distributed across FAQ levels six to ten and among TD children. The GDI also scales with respect to clinical involvement based on topographic CP classification in Hemiplegia Types I–IV, Diplegia, Triplegia and Quadriplegia. The GDI offers an alternative to the GGI as a comprehensive quantitative gait pathology index, and can be readily computed using the electronic addendum provided with this article.

Introduction

Comprehensive measures of gait pathology are useful in clinical practice. They allow stratification of severity, give an overall impression of gait quality, and aid in objective evaluation of treatment outcome. There are many ways to gauge overall gait pathology. Parent report questionnaires such as the Gillette Functional Assessment Walking Scale (FAQ), observational video analysis schemes like the Edinburgh Gait Score, or rating systems such as the Functional Mobility Scale (FMS), can provide a general picture of gait impairment [1], [2], [3]. While parent and caregiver assessments are useful and practical, they lack the precision and objectivity provided by three-dimensional quantitative gait data.

Gait data can be used to assess pathology in a variety of ways. For example, stride parameters such as walking speed, step length, and cadence provide an overall picture of gait quality. These parameters are especially useful when non-dimensionalized to account for differences in stature [4]. It is possible, however, to walk with adequate stride parameters and still have significantly atypical joint motions and orientations. This suggests a need for three-dimensional gait data in assessing overall gait pathology. Interpreting three-dimensional gait data in a global sense is not a simple task. Difficulties arise from the complexity of gait, and from the interdependent nature of gait data. For example, to assess the motions of the lower extremities during a single stride requires the analysis of multiple joints and body segments in multiple planes at multiple instants of time. Furthermore, these motions are coupled across joints, planes, and time. Motions of one joint affect the motions of adjacent or remote joints. Motions of a joint in one plane are coupled to motions in other planes. Finally, positions of a joint at one time affect positions at a later instant. Combining these effects, it can be surmised that the motion of a joint in a given plane at one instant can affect the position of a different joint, in a different plane, at a different instant. It is clear, therefore, that some method for dealing with this complexity and interdependence is necessary to gain an overall sense of gait pathology.

A number of multivariate statistical methods have been developed for dealing with the complexity and interdependence of gait data [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20]. While some of these methods focus primarily on identifying gait patterns and relationships among variables, several aim to develop either joint-specific or overall indexes of gait pathology [7], [8], [10], [12], [13], [14], [19]. Among these, the Gillette Gait Index (GGI) appears to be the most extensively validated, commonly cited (based on a

™ citation search), and is widely used in clinical gait research and practice [3], [12], [13], [21], [22], [23], [24], [25]. While the GGI has been shown to be useful, a number of limitations have also been noted [26], [27]. These include the arbitrary, unbalanced, and incomplete nature of the 16 univariate parameters that comprise the index, uncertainty surrounding principal component scaling, non-normality of the index, lack of physical meaning for the multivariate components, and difficulties in implementation—including excessive sensitivity to lab-specific control data.

This article describes a new measure of overall gait pathology—the Gait Deviation Index (GDI). Face and concurrent validity data for the GDI are presented through comparisons with the GGI, FAQ, and topographic classifications within the diagnosis of Cerebral Palsy (CP).

Section snippets

Motivation

The method used in constructing the GDI was motivated by a biometric method used for face identification—the so-called “eigenface” method [28]. In the eigenface method, a large collection of faces is digitized and the resulting arrays of grayscale values are converted to vectors. This collection of vectors is then subjected to principal component analysis. A small number of the extracted eigenvectors (called eigenfaces) that account for a large percentage of the information in the original

Order of reconstruction

Examination of VAF and Φ showed that 15 features accounted for 98% of the total variation and produced a 98% faithful gait vector as measured by the mean Φ (Fig. 1). Further examination showed that 99% of all subjects exhibited a Φ > 95%. The difference in reconstruction efficiency between native and non-native data was trivial (<0.1%) at m = 15. Based on these results, 15 features (mcrit = 15) was deemed to provide a “sufficiently” faithful reconstruction of the native and non-native data.

A typical

Discussion

A new measure of overall gait pathology, the Gait Deviation Index (GDI), has been introduced along with concurrent and face validity data. The GDI scales with overall gait function, is well behaved statistically, and can be implemented easily using the Supplemental data provided with this article.

The GDI was strongly correlated (r2 = 0.56) with the previously validated and widely used GGI. This suggests that the GDI and GGI are both measures of the same underlying construct, though the large

Conflict of interest

None of the authors had any financial or personal relationships with other people or organizations that could inappropriately influence this work.

References (30)

  • K.J. Deluzio et al.

    Principal component models of knee kinematics and kinetics: normal vs. pathological gait patterns

    Human Movement Science

    (1997)
  • L.M. Schutte et al.

    An index for quantifying deviations from normal gait [see comment]

    Gait and Posture

    (2000)
  • A. Daffertshofer et al.

    PCA in studying coordination and variability: a tutorial

    Clinical Biomechanics

    (2004)
  • M. Romei et al.

    Use of the normalcy index for the evaluation of gait pathology

    Gait and Posture

    (2004)
  • H.K. Graham et al.

    The functional mobility scale (FMS)

    Journal of Pediatric Orthopedics

    (2004)
  • Cited by (523)

    View all citing articles on Scopus
    View full text