Forceplates Data Reliability Tested on Children with Neuromuscular

Disabilities

Sarah Rosen

1,2

, Carole A. Tucker

and Rahamim Seliktar

School of Biomedical Engineering, Science and Health Systems, Drexel University,

3141 Chestnut Street, Philadelphia, PA, U.S.A.

Shriners Hospitals for Children, 3551 North Broad Street, Philadelphia, PA, U.S.A.

Department of Physical Therapy, Temple University, 3307 North Broad Street, Philadelphia, PA, U.S.A.

Keywords: Gait Forces, Gait Cycle, Gait Impulse, Gait Consistency, Pediatric Gait.

Abstract: The “gait cycle” is considered the fundamental repetitive element of locomotion and is often used in

comparative analyses between pathological and normal gait. One device that is used to measure ground

reaction forces in the gait laboratory is a six degrees of freedom force transducer (“Force Plate”). Gait labs

use two or more force plates to measure forces during one or two complete strides. Often, the walker “aims”

to step sequentially on the force plates. This aiming may cause significant distortion of the person’s gait

pattern. To validate the measured result as representative of the individual’s gait, a “Gait Consistency Test”

was developed by Seliktar et.al with a group of adult individuals. The present study employed the

consistency test to the ground reaction force records of two groups of children; typically developing

children and children with cerebral palsy. The children’s gait consistency study yielded even better results

than the adult study; the results showed that the consistency test is useful for screening children’s gait data

and should be used to ensure higher fidelity of the measured data. The test applied equally to normally

developing children and to children with gait pathologies.

1 INTRODUCTION

The hypothesis that is being tested in this work is: Is

it legitimate to assume that human being ambulate

by performing repetitive patterns (rhythm) of

movement, often referred to as “Gait Cycles?” These

cycles, if exist, should apply equally to normal gait

as well as pathological gait. If such rhythm does not

exist, then the term “Gait Cycle” is invalidated and

would not be appropriate to be used in comparative

gait analysis.

3D human motion analysis relies on the use of

dynamic data acquired with the aid of motion

tracking and ground reaction forces measurement

instruments. The usefulness of the raw data

characteristics themselves is often apparent in

detection of gait abnormalities. The processing of

the force and motion data to yield joint forces,

moments and power during gait, further enhances its

utility (Gage 2004). The relative simplicity of the

ground reaction forces (GRF) characteristics makes

them attractive as a clinical assessment tool, as

opposed to the abundance of the kinematic data.

Moreover, the joint moment information derived

directly from these forces, further enhance the utility

of these data in understanding the “human gait

engine.”

Considering the importance of the ground

reactions as the main driving force of the overall

human mechanism little has been done to address

the validity and consistency of the measured GRF

data during gait. The awareness of the shortcomings

of the force plates as the primary available

instrument, has led people to develop devices for

continuous measurements of the ground forces.

Nevertheless, all the new developments were

restricted to measurement of the vertical component

only. Seliktar et al. (Seliktar, Yekutiel, Bar 1979)

developed a “Gait Consistency Test” to determine if

due to the person’s aiming towards the force plates

or due to any other environmental distractions, the

steady state nature (rhythm) of the measured gait is

distorted. The test was designed to ensure that the

velocity of the body center of mass is “the same”

(within certain variation limits) at any two

equivalent points of (at least) two consecutive

303

Rosen S., A. Tucker C. and Seliktar R..

Forceplates Data Reliability Tested on Children with Neuromuscular Disabilities.

DOI: 10.5220/0004934303030306

In Proceedings of the International Conference on Biomedical Electronics and Devices (TPDULL-2014), pages 303-306

ISBN: 978-989-758-013-0

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

cycles. The premise of the test was that the

velocities of the CM is governed by the ground

forces and hence the force-time integral (Impulse)

equals the difference in Momenta ∆ (the

Impulse-momentum theorem):







  









(1)

over a full cycle, should be zero, as illustrated in

Figure 1. Since the GRFs are vectors, the antero-

posterior (A-P) force component was selected as the

most adequate to be used for such validation of the

results.

According to Newton’s 2

Law, for a system,

the ground reaction forces and the gravitational force

are driving the center of mass of the body;

∑









(2)

Since the impulse of the driving forces equals the

difference in mass times velocity, the force-time

integral between those two points, in rhythmic gait,

must be zero. Nevertheless, small variations between

cycles are expected and therefore absolute zero is

not a feasible expectation.

The decision to use the A-P component for

validation of the consistency is due to the authors’

observation that human gait is less cyclical in the

medio-lateral (M-L) and the measured vertical force

is inconvenient to use, because of the body weight

superposition. Moreover, because the A-P force is

responsible for the forward progression of the body

it is more likely to reflect deviations from rhythmic

steady state gait due to environmental biases.

Previously reported results of the study were

obtained from 28 adults with normal and

pathological gait. Consistency was defined as the

sum of the A-P impulse over one full gait cycle

divided by the absolute values of impulse for both

the braking and propulsion forces, Figure 1 and

Equation (2) describe the calculation:



























(3)

When looking at Figure 1 and Equation (3) it is clear

that the contribution of impulse 1; the impulse of the

right foot from the time of the left foot heel strike to

when the right foot toes off, is not accounted for in

the calculation.

This is because only 2 force plates were available

and the assumption made was that the areas denoted

as 1 and the dashed part of area D are almost

equivalent. The consistency threshold value was

determined to be 7.3% meaning that values greater

than 7.3% for the gait cycle would be considered

Figure 1: The GRFs in the A-P direction during two force

plate strikes of an able bodied individual. HS stands for

heel strike and TO for toe off. ‘l’ and ‘r’ stand for the left

and right sides. T represents the time of duration of the

gait cycle.As depicted in the illustration of figure (1),

consistency was calculated over two force plate strikes, for

both legs, left and right.

inconsistent and should be removed from further

analyses. Clinical gait analysis is commonly

performed in children with cerebral palsy (CP), and

even though the GRF data is typically relatively

underutilized in comparison to kinematic data it can

provide a useful tool for ensuring the data used is

representative. Currently the method used for

ensuring representability is by averaging a number,

usually three, trials of gait analysis. In the current

study, the above approach was applied to three

groups of children: typically developed children

(TD), children with diplegic CP (DCP) and children

with hemiplegic CP (HCP) to determine the utility

of the gait consistency test.

2 METHODS

Subjects: Retrospective data from a sample of 53

ambulatory children who underwent gait analysis

testing at Shriners Hospital for Children (SHC) –

Philadelphia, Motion Analysis Laboratory from

2001-2006 was used in this analysis. The first group,

TD, consisted of 8 males and 8 females, 7-17 years

of age (Mean=11.2, SD= ±2.1). The second group,

DCP), consisted of 8 males and 8 females, 9-17

years of age (Mean=12.5, SD= ±2.0). The third

group, HCP, with left (n=8) or right (n=13) sides

affected, consisted of 10 males and 11 females, 10-

17 years of age (Mean=12.9, SD= ±2.3). The

analysis included 2 walking trials for each subject

and an average of 2 gait cycles per trial.

Data Collection: Data were obtained from

BIODEVICES2014-InternationalConferenceonBiomedicalElectronicsandDevices

304

motion analysis database and report reviews. Data

were only included for children with CP who did not

use assistive devices during ambulation, and

completed two trials during which consecutive gait

cycles for each of the four force plates was

available.Only trials during which only a single foot

landed on each of the force plates was deemed

acceptable (e.g. trials during which 2 feet or more

landed on a single force plate were discarded). This

retrospective data used in this study was collected

during gait analysis utilizing motion capture and

force plate data collection during level ground

ambulation. The Motion Analysis Laboratory at

SHC–Philadelphia is equipped with an 8-camera

MX-Vicon motion capture system (Vicon Motion

systems, Lake Forest, CA) and 4 AMTI force plates

(Advanced Mechanical Technology Inc.,

Watertown, MA). Analog force plate data were

synchronized and collected through the Vicon

system software. Each subject wore shorts and a T-

shirt throughout the evaluation and gait analysis was

performed with the subject walking barefooted and

walking at his/her freely chosen walking speed along

an 8.4-meter level walkway. Force data were

collected when the subject traversed the middle 5

meters.

Data Analysis: Force plate data in the A-P

direction that had been collected at 1200Hz

sampling rate, was extracted from the Vicon’s c3d

files and imported into Matlab 7.1 (The Mathworks

Inc. Natick MA) for further analysis. A custom-

written Matlab program was used to calculate gait

consistency, as defined in Equation (2). A one-way

ANOVA test was used to determine significant

differences in the consistency values between the

three populations (Portney & Watkins 2009).

3 RESULTS

Data extraction resulted in a total of 2 trials for each

of the 53 subjects. The ANOVA test revealed the

values for the three populations were not

significantly different (F=1.18, p=0.31). An average

consistency value was calculated for the three

populations separately, and for the entire population

as a whole. Results for the pediatric population

demonstrate that children can consistently ambulate

within a motion analysis laboratory setting. The

consistency test values can be found in Table 1 and

Figure 2. These results were extremely close to the

findings of Seliktar et al. (Seliktar, Yekutiel, Bar

1979) and show consistency is achieved with a ratio

below 6.33%.

Table 1: Results of the consistency test for typically

developed children, children with diplegic CP and children

with hemiparetic CP.

Population

Average Consistency

value [%]

Children with Typical

Development

3.44 ± 2.48

Children with Diplegic

3.59 ± 2.47

Children with

Hemiparetic CP

4.09 ± 2.58

Entire population with

no regards to condition

3.77 ± 2.56

Figure 2: Results of the consistency test for typically

developed children and children with diplegic and

hemiplegic CP. Results are presented as the mean with the

standard deviation.

4 CONCLUSIONS

The results of the study support the application of a

gait consistency test to children with typical

development and with CP. Our findings are

consistent with those found by Seliktar (Seliktar,

Yekutiel, Bar 1979) supports the definition of an

inconsistent gait in children to be one in which the

consistency value is above 6.33%. Given no

significant differences in the consistency values for

each of the three populations, it appears gait

consistency is not dependent on the presence, or

absence, of pathological conditions. The consistency

test reflects ambulation at a constant speed of

progression (velocity averaged per cycle), and many

gait parameters are affected by gait velocity – or

relative changes in gait velocity over consecutive

cycles (e.g. speeding up or slowing down during gait

analysis trials). The application of such a

ForceplatesDataReliabilityTestedonChildrenwithNeuromuscularDisabilities

305

consistency test to ensure gait trials are compared

across steady state trials, this measure provides a

simple means to assess whether a specific gait cycle,

chosen for analysis is in fact a representative gait of

the child.

ACKNOWLEDGEMENTS

The authors would like to thank the staff at Shriners

Hospitals for Children - Philadelphia Motion

Analysis Laboratory for its assistance in this study

and the support obtained from the Calhoun

Fellowship Fund of the School of Biomedical

Engineering Science and Health Systems at Drexel

University.

REFERENCES

Gage, James R. The Treatment of Gait Problems in

Cerebral Palsy. London: Mac Keith, 2004. Print.

Portney, Leslie Gross., and Mary P. Watkins. Foundations

of Clinical Research: Applications to Practice. Upper

Saddle River, NJ: Pearson/Prentice Hall, 2009. Print.

Seliktar, R., Yekutiel, M., Bar A. (1979). Gait Consistency

Test Based on the Impulse-Momentum Theorem.

Prosthet Orthot Int. 3 (2), 91-98.

Winter, David A. Biomechanics and Motor Control of

Human Movement. Hoboken, NJ: Wiley, 2009. Print.

BIODEVICES2014-InternationalConferenceonBiomedicalElectronicsandDevices

306