G Model

GAIPOS-3652; No. of Pages 3

Gait & Posture xxx (2012) xxx–xxx

Contents lists available at SciVerse ScienceDirect

Gait & Posture

journal homepage: www.elsevier.com/locate/gaitpost

Short Communication

Clinical stabilometry standardization

Basic definitions – Acquisition interval – Sampling frequency
Fabio Scoppa a,b,1, Roberto Capra a,c,1, Michele Gallamini a,d,1,*, Riccardo Shiffer a,b,1
a
International Standardization Committee for Clinical Stabilometry I.S.P.G.R. - International Society for Posture and Gait Research, Italy
b
Faculty of Medicine and Dental Surgery, Sapienza University of Rome, Italy
c
La Colletta Bioengineering Center (DIST Department Genoa University and ASL3 Regional Public Health Service), Italy
d
RGMD SpA R&D Department, Italy

A R T I C L E I N F O A B S T R A C T

Article history: Several statements on the still debated issue of stabilometry standardization were agreed upon by the
Received 18 July 2011 ISPGR Standardization Committee.
Received in revised form 2 July 2012  A set of metrological characteristics for stabilometric platforms was defined.
Accepted 7 July 2012
 Relying both on practice and experimental verification it was agreed that, to obtain appropriate
accuracy and sensitivity in the Romberg Test:
Keywords:
Clinical Static Stabilometry * The acquisition interval should not be less than 25 s.
Force Platform * The sampling frequency should be at least 50 Hz.
Romberg Test  After careful consideration it was decided that the recommendations made in the previous
Standardization proposal in 1983 regarding environmental conditions should be maintained.
ß 2012 Elsevier B.V. All rights reserved.


1. Posturographic standards X ¼ xðCOPÞ ðtÞ
COPðtÞ !
Y ¼ yðCOPÞ ðtÞ
During the last Meeting of the International Society for Posture
and Gait Research (ISPGR) (Bologna, Italy 2009) a new Interna- This topic has been widely discussed [2–4].
tional Standardization Committee was set up to address the still The conventional definition whereby
unresolved issue of standardization in posturographic or static
stabilometry.  the X axis is the horizontal trace of the Latero-Lateral plane
Agreement was reached on the following topics. aimed towards the right side of the patient,
 the Y axis is the horizontal trace of the Antero-Posterior plane
aimed ahead of the patient
2. Metrological characteristics is recognized and accepted.
The origin of the coordinates is placed on the posterior left
It is strongly felt that there should be an agreement on standard corner of the platform with reference to the subject’s position
requirements for instrumental measurement performance during the test.
(Technical Performance) rather than on device design. According to the stated principle, the time functions
Since Force Platforms, regardless of the number, type and
characteristics of the chosen sensors, provide the COP Sway Signal X ¼ xðtÞ
[1], it has been agreed that Technical Performance parameters for
stabilometric measurement instruments should be based on the Y ¼ yðtÞ
COP Sway Signal measurement.
should be produced by the measuring device with the following
The ‘‘COP Sway Signal’’, consists of the X, Y time plot of the COP
Technical Performances:
during the test.
 Accuracy: better than 0.1 mm.
 Precision: better than 0.05 mm.
* Corresponding author at: RGMD SpA R&D Department, Italy.
Tel.: +39 10 8594434; fax: +39 10 8601304.
 Resolution: higher than 0.05 mm.
E-mail address: [email protected] (M. Gallamini).  Linearity: better than 90% over the whole range of measure-
1
These authors contributed equally to this work. ment parameters.

0966-6362/$ – see front matter ß 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.gaitpost.2012.07.009

Please cite this article in press as: Scoppa F, et al. Clinical stabilometry standardization. Gait Posture (2012), http://dx.doi.org/10.1016/
j.gaitpost.2012.07.009
G Model
GAIPOS-3652; No. of Pages 3

2 F. Scoppa et al. / Gait & Posture xxx (2012) xxx–xxx

 Anthropometric range:
* Weight: 20–200 kg,
* Height: 80–250 cm,
* Foot size: up to 35 cm.
 Frequency bandwidth: 0.01–10 Hz.

3. Clinical parameters

The above requirements should be verified as the goal must be

the accuracy of the clinical parameters. The process should be in
fact inverted to set Technical Performances requirements capable
of ensuring such accuracy.
It was however recognized that parameters are dependent: Fig. 1. Parameters stability vs. sampling rate.

(a) on the quality of measured coordinates,

(b) on the type and performance of the signal processing From a theoretical standpoint, the maximum detectable
algorithms applied, frequency depends on the choice of sampling rate with a ratio
(c) on the adopted test method, mainly with respect to feet of at least 2, although general practice suggests a ratio ranging
position and postural requirements. from 5 to 10. This means that, if the maximum detectable
frequency is 10 Hz, the sampling rate should not be less than 50 Hz.
It has been suggested that an increase in the sampling rate
4. Processing algorithms
could add ‘‘noise’’ rather than ‘‘information’’ to the detected signal,
which is conventionally cut off below 3 Hz, but more recent
Starting from the COP Sway Signal (COP(t) as defined) the
observations [12] suggest to consider an effective frequency range
following will be extracted:
of up to 10 Hz (i.e. a sampling rate higher than 50 Hz).
An experimental verification was performed to check whether
(a) The two basic Stabilometric Graphs, namely the ‘‘STABILO-
this assumption is correct. From an archive of 40-s recordings
GRAM’’, or the time plot of the two coordinates X, Y, and the
sampled at 100 Hz from a Force Platform (RGMD ARGO1), 14
‘‘STATOKINESIOGRAM’’, or the Y vs. X plot as a function of time.
recordings were randomly extracted. The recordings (timed series
(b) A set of parameters describing the COP Sway Signal (COP(t)).
of x, y coordinates) were down-sampled at 50, 25 and 20 Hz using
‘‘Classical’’ and new balance parameters, including Spectral
Matlab1 software package and all the Parameters then calculated
Harmonic Analysis, ‘‘random-walk’’ [5], Sway Density [6],
for a sampling rate of 100, 50, 25 and 20 Hz.
Fractal Analysis [7], Chaotic and Stochastic Analysis, etc. will be
Calculated parameters were then normalized to the value
examined.
obtained at the 100 Hz sampling rate.
As shown in the graph of Fig. 1, Classical Parameters (Sway Path,
5. Test methods Sway Area and Confidence Ellipse Area) are reasonably steady and
a sampling rate of 50 Hz seems to be acceptable to get reliable
The criteria proposed by Kapteyn et al. [8] are generally values.
confirmed: some objections still need to be addressed with regard Both Oscillations and Sway Density Parameters are instead
to Sampling Rate, Test Duration, Foot position and the patient’s requiring a higher sampling frequency and the 100 Hz sampling
posture. rate is recommended.
Some experiments addressed the first two issues.
The feet position was discussed by a number of qualified 7.2. Test duration
authors [9,10] but a conclusive recommendation on the subject has
not been reached yet [11]. It has been argued that

6. Environmental test conditions (1) any task requires an ‘‘adaptation phase’’ and will be affected by
fatigue or lack of attention;
It is recommended that the test conditions as defined in (2) the constantly changing conditions of the living body affect
paragraph 8 of the aforesaid work on standardization [8] (except performances that can never be thought of as ‘‘steady’’
for item a) (feet position), which will be examined further, should whatever the recording time used. This would cast doubt on
be accepted. the applicability of the Fourier Transform.

7. Experimental activities According to literature however, the most common recording

time values range from 20 to 60 s. A previous study suggests a
The topics of: recording interval of around 40 s [13].
To check the effect of recording time, 25 randomly chosen series
 Sampling rate, of 40 s recordings (5 s offset at start) sampled at 100 Hz from a
 Test duration (acquisition time) Force Platform (RGMD ARGO1) were analyzed and processed to
were experimentally investigated, as follows. obtain parameters at different duration times (5, 10, 15, 20, 30, 35
and 40 s).
7.1. Sampling rate The following observations may be made (see graphs in Fig. 2):

There is no technical advantage in reducing either the sampling (1) Ellipse Area – although the area is constantly increasing, there
rate or the digital resolution (a 2 GB USB memory stick could store is an indication of convergence towards a ‘‘stable’’ value for
more than 15,000 COP acquisition tests (60 s, 100 Hz, 16 bits). recording intervals equal to or greater than 30 s.

Please cite this article in press as: Scoppa F, et al. Clinical stabilometry standardization. Gait Posture (2012), http://dx.doi.org/10.1016/
j.gaitpost.2012.07.009
G Model
GAIPOS-3652; No. of Pages 3

F. Scoppa et al. / Gait & Posture xxx (2012) xxx–xxx 3

MG performed the data acquisition and processing while FS and RS

were mainly involved in the clinical contribution.

Conflict of interest statement

The authors Fabio Scoppa, Roberto Capra, Roberto Schiffer

declare that they have no competing interests. Michele Gallamini is
the R&D Director of RGMD SpA an Italian Company designing and
producing Medical Devices and in particular a range of Force
Platforms.

References

[1] Gagey PM, Bizzo G. La mesure en Posturologie (4/1/2001) [http://pierremarie.

Fig. 2. Normalized parameters values vs. recording time. gagey.perso.sfr.fr/MesureEnPosturologie.htm].
[2] Bizzo G, Guillet N, Patat A, Gagey PM. Specifications for building a vertical force
platform designed for clinical stabilometry. Medical and Biological Engineer-
ing and Computing 1985;23:474–6.
(2) Sway Path and Sway Density (Mean Stay Time and Mean [3] Browne J, O’Hare N. Development of a novel method for assessing balance: the
Spatial Distance) – stable values are achieved almost quantitative posturography system. Physiological Measurement 2000;21:525–34.
immediately but good ‘‘quality’’ parameters are obtained with [4] Browne J, O’Hare N. A quality control procedure for force platforms. Physio-
logical Measurement 2000;21:515–24.
at least 30 s.
[5] Collins JJ, De Luca CJ. Upright, correlated random walks: a statistical-biomechanics
(3) Sway Area and Oscillations – there is a strong indication of approach to the human postural control system. Chaos 1995;5(1):57–63.
convergence towards a ‘‘stable’’ value for recording intervals [6] Baratto L, Morasso PG, Re C, Spada G. A new look at posturographic analysis in
equal to or greater than 30 s as for the Confidence Ellipse Area. the clinical context: sway-density versus other parameterization techniques.
Motor Control 2002;6(3):246–70.
[7] Błaszczyk JW, Klonowski W. Postural stability and fractal dynamics. Acta
In general terms it can be concluded that from a recording time Neurobiologiae Experimentals (Wars) 2001;61(2):105–12. Erratum in: Acta
of 25–40 s the Sway Parameters are steady and reliable and a Neurobiologiae Experimentals (Wars) 2001;61(4):327.
[8] Kapteyn TS, Bles W, Njiokiktjien Ch, Kodde L, Massen CH, Mol JMF. Standardi-
reasonable compromise could well be 30 s with some 5 s of zation in platform stabilometry being a part of posturography. Agressologie
adjustment time before starting the recording. 1983;24:321–3.
Clinical observations nevertheless suggest that, especially with [9] Kirby RL, Price NA, MacLeod DA. The influence of foot position on standing
balance. Journal of Biomechanics 1987;20:423–7.
impaired patients, maintaining an upright unperturbed stance is a [10] Mouzat A, Dabonneville M, Bertrand P. The effect of feet position on orthostatic
demanding task and in many cases extending the recording time posture in a female sample group. Neuroscience Letters 2004;365:79–82.
would just add ‘‘noise’’ due to fatigue or diminished attention. It [11] Ruhe A, Fejer R, Walker B. The test–retest reliability of centre of pressure
measures in bipedal static task conditions – a systematic review of the
has instead been suggested that random temporary effects
literature. Gait and Posture 2010;32:436–45.
possibly give rise to irregular responses and hence it would be [12] Bottaro A, Casadio M, Morasso PG, Sanguineti V. Body sway during quiet
advisable to calculate parameters as the average of those obtained standing: is it the residual chattering of an intermittent stabilization process?
Human Movement Science 2005;24:588–615.
in three successive recordings [14].
[13] Baratto L, Jacono M, Morasso P, Simonini M, Navarra S, Gallamini M. La durata
di registrazione nel test stabilometrico statico su piattaforma di forza [Acqui-
Authors’ contributions sition time in the stabilometric test on force platform]. Italian Journal of
Rehabilitation Medicine-MR 2006;20:103–8.
[14] Pinsault N, Vuillerme N. Test–retest reliability of centre of foot pressure
All the mentioned Authors took part in the whole process, measures to assess postural control during unperturbed stance. Medical
sharing opinions and experiences under the Co-Chair of FS. RC and Engineering and Physics 2009;31:276–86.

Please cite this article in press as: Scoppa F, et al. Clinical stabilometry standardization. Gait Posture (2012), http://dx.doi.org/10.1016/
j.gaitpost.2012.07.009