Assessment of Gait Symmetry and Gait Normality Using Inertial Sensors: In-Lab and In-Situ Evaluation - Biomedical Engineering Systems and Technologies - page 216

The present study concerns the development of a inertial sensor system for GA that
can be successfully deployed in the clinical setting. This system, composed of wearable
inertial sensor units, can complement OGA by providing reliable quantitative measures
of gait symmetry and gait normality. This study evaluates the proposed symmetry and
normality measures both in-lab, against measures derived from MOCAP; and in-situ,
compared to clinical assessments of hip-replacement patients.

2 Related Work

2.1 Observational Gait Analysis

Observational gait analysis (OGA) is frequently aided by video recordings, which pro-
vide certain interesting functions to the observer such as pause and slow motion. In
some cases, quantitative measurements, such as joint angles, can be directly calculated
from the video image [6]. This type of analysis is often accompanied by a form or
questionnaire that facilitates the extraction of relevant information from the video. Two
such forms have been more thoroughly investigated and more widely adopted: the Vi-
sual Gait Assessment Scale (VGAS) [7] and the Edinburgh Visual Gait Score (EVGS)
[8]. Both questionnaires target the assessment of children with cerebral palsy.
OGA can be complemented by other more quantitative measurements, such as av-
erage gait speed, average step length and other gait parameters. These are typically
measured during walking tests, such as the 10-m walking test [9], or the timed up and
go test (TUG) [10]. The TUG is normally employed in studies where balance and risk of
fall are of interest, as it requires that the subject stand up and sit down on a chair without
help. The 10-m walking test is a simple way of determining, average gait speed, stride
length and cadence. Average gait speed, for example, has been identified as an indicator
of: activity of daily living function in geriatric patients [11]; high risk of health-related
outcomes in well-functioning older people [12]; and leg strength in older people [13].
Stride length is another interesting measure that has been associated with, for example,
metabolic cost and impact during walking [14].

2.2 MOCAP Gait Analysis

Gait may be studied in the spatio-temporal, kinetic or kinematic domains. In order to
simplify the analysis of this extremely rich source of information, one may focus on
measures of symmetry and normality. Symmetry refers to the similarity between the
movements of the right and left sides of the body. Normality refers to the similarity
between the movements of one individual compared to average movements of a popu-
lation that is judged healthy or normal.
The symmetry of kinematic gait data is usually evaluated by visual inspection of
superimposed curves from right and left sides. Few quantitative symmetry measures
have been proposed which take into account complete joint angle curves. A measure of
trend symmetry based on the variance around the 1 st principal component of a right-side
vs. left-side plot has been suggested [15]. This trend symmetry measure is insensitive
to scaling, and must be compensated by an additional measure, the range amplitude