Mylene Dault

07-16-1997, 12:53 AM

This is the message I sent:

I'm presently working on postural sway with head injured subjects.

I'm having some troubles to find a way to quantify the surface of the sway.

The results shown a multi-form pattern. Subjects have been tested on a AMTI

force plate-form and the acquisition is done by the Peak system.

Here are some responses.

Thank you for the numerous responses !!!

Mylene Dault

************************************************** *************************

I have the following references for you

Dual-task Assessment of Reorganization..

ACH Geurts et al. Arch Phys Med Rehabil 72, 1991,p 1059.

Attention Demands in Balance...

ACH Geurts et al. J of Motor Behavior 26(2), 1994, p 162.

Identification of Static and Dynamic..

ACH Geurts et al. Arch Phys Med Rehabil 77, 1996, p639.

In this artikles they have done postural and balance studies.

May be it will be a help for you.

Rico

"Rico de Visser"

************************************************** ******************

Currently a graduate student I advise is working on postural sway using an

AMTI force plate. To quantify the amount of sway he sums the amount of

displacement of the ground reaction forces in the X and Y directions

(lateral and fore/aft, respectively, according to AMTI's definition) during

a five sec. trial with sampling at 100 Hz. For example if the there are five

consecutive samples of 0.5, 0.10, 0.5, -0.5, -0.10 cm, for the X values he

adds 0.5+0.5+0.10+0.5=2.5 cm (differences between the values). This is also

calculated for the Y direction. The velocity can be determined by dividing

by the amount of time to get cm/sec.

The calculations for the difference between the 500 numbers can easily and

quickly be derived using a spreadsheet like Microsoft Excel. If the X data

is in column A (cells A1:A500) of the worksheet, enter the formula in cell

B2: =ABS(A2-A1) and fill the formula down to B500. In cell B501, calculate

the sum by entering the formula: =SUM(B2:B500) or by clicking on the sum S

button on the toolbar.

This is like an integration of a full-retified EMG signal. A

root-mean-square (RMS) calculation would also work, I believe. If your Peak

system is recording the force plate data, it should be able to export the

files to a spreadsheet. We use an Ariel system which exports data into the

Excel spreadsheet.

I have not verified this method with other people studying postural sway and

would be interested in knowing if this is a valid method. I would also like

to see other responses you get. If you don't post replies, please forward

them to me.

Thank you.

BE

Bruce Etnyre, Ph.D., P.T. _/_/_/_/ _/_/_/_/ _/_/_/ _/_/_/_/ _/ _/

Human Performance and _/ _/ _/ _/ _/ _/ _/ _/

Health Sciences Dept. _/ _/ _/ _/ _/ _/ _/

Rice University _/_/_/_/_/ _/ _/ _/_/_/ _/ _/

6100 Main MS 545 _/ _/ _/ _/ _/ _/ _/

Houston, Texas 77005 _/ _/ _/ _/ _/ _/ _/

USA _/ _/ _/ _/ _/ _/ _/ _/

etnyre@rice.edu _/ _/ _/_/_/_/ _/_/_/ _/_/_/_/ _/_/ _/

(713)527-4058

FAX: (713)285-5329

************************************************** ***********************

We resolved the problem as follow supposing you have about 1000 points

passed by sway. All points have a pair of coordinates (x , y) :

1. taking the average of x and y gives X0 and Y0

2. calculate the distance D and the angle A of each point vs X0 and Y0

3. divide the area in 24 sectors with X0 and Y0 as center

4. sort the data so that you find which point is in wich sector

5. sort the data in each sector to the biggest D

6. calculate the surface of each triangle obtained and make the sum.

The more sectors gives more chance to be accurate but also increase

the chance of finding a area without points...

I found the algorithm in :

Arch. Phys. Med. Rehabil vol 75 September 1994

Sincerely

Pierre

================================================== ========

ing Pierre Van Cleven

Dept of Kinesiology Institute of Physical Education

University of Gent

Watersportlaan 2 B-9000 Gent (Belgium)

tel (32)9.264.63.21 fax (32)9.264.64.97

================================================== ======

************************************************** *****************

>I'm having some troubles to find a way to quantify the surface of the sway.

>The results shown a multi-form pattern.

You might try looking at the data in terms of fractional brownian motion,

as in these papers:

Riley, M. A., Mitra, S., Stoffregen, T. A., & Turvey, M. T. (1997).

Influences of body lean and vision on unperturbed postural sway. Motor

Control, 1, 229-246.

Collins, J.J. & De Luca, C.J. (1995). The effect of visual input on

open-loop and closed-loop postural control mechanisms. Experimental Brain

Research, 103, 151-163.

About a dozen additional ways to characterize postural sway are presented in

Riccio, G. E., & Stoffregen, T. A. (1991). An ecological theory of motion

sickness and postural instability. Ecological Psychology, 3, 195-240.

Good luck.

Tom Stoffregen

------------------------------------------------------------------------------

Thomas A. Stoffregen, Ph.D.

Department of Psychology, P.O. Box 210376

University of Cincinnati

Cincinnati, OH 45221-0376

(513) 556-5569

Personal homepage: http://ucaswww.mcm.uc.edu/psychology/Faculty/Stoffreg.htm

Postural stability laboratory: http://ucaswww.mcm.uc.edu/psychology/stoffregen/

------------------------------------------------------------------------------

************************************************** *************************

Mylene:

We are just developing a program for testing postural sway

in knee patients. So far, we have the subjects stand on the

force plate (Kistler) in a single leg stance for 15 seconds

with eyes open and eyes closed. We record the vertical ground

reaction force Fz and the location of the force vector on the

platform in both directions (ax and ay). We sample 600

measurements, i.e. we have a frequency of 40 Hz. We calculate the

standard deviation of ax and ay, the range between min and max

values in both directions and the total distance that the force

vector "traveled" during the whole period. Dividing distance by

time gives you something like the sway velocity.

We have yet to find out which parameter(s) is (are) most reliable

and can help to distinguish between normal and pathologic

conditions. I hope this helps a little.

Ciao,

Dieter...

__________________/\ ________/\ _____________________

\/ \/

Dr. Dieter Rosenbaum

Abt. fuer Orthopaedische Physiologie (Kinesiology Lab)

Klinik und Poliklinik fuer Allgemeine Orthopaedie

Westfaelische Wilhelms-Universitaet Muenster

Albert-Schweitzer-Str. 33

D-48129 Muenster, Germany

Phone: (...49) 251 - 834 7981 or 7995 (fax 7989)

email: diro@uni-muenster.de

************************************************** *************************

Thank you to Stephen J. Kinzey, Ph.D. for his response.

Stephen J. Kinzey, Ph.D.

Assistant Professor / Director of Biomechanics Laboratory

The University of Mississippi

Department of ESLM

University, MS 38677

e-mail: skinzey@olemiss.edu

http://www.olemiss.edu/~skinzey/biomch.htm

office: (601) 232 - 5540

fax: (601) 232 - 5525

I'm presently working on postural sway with head injured subjects.

I'm having some troubles to find a way to quantify the surface of the sway.

The results shown a multi-form pattern. Subjects have been tested on a AMTI

force plate-form and the acquisition is done by the Peak system.

Here are some responses.

Thank you for the numerous responses !!!

Mylene Dault

************************************************** *************************

I have the following references for you

Dual-task Assessment of Reorganization..

ACH Geurts et al. Arch Phys Med Rehabil 72, 1991,p 1059.

Attention Demands in Balance...

ACH Geurts et al. J of Motor Behavior 26(2), 1994, p 162.

Identification of Static and Dynamic..

ACH Geurts et al. Arch Phys Med Rehabil 77, 1996, p639.

In this artikles they have done postural and balance studies.

May be it will be a help for you.

Rico

"Rico de Visser"

************************************************** ******************

Currently a graduate student I advise is working on postural sway using an

AMTI force plate. To quantify the amount of sway he sums the amount of

displacement of the ground reaction forces in the X and Y directions

(lateral and fore/aft, respectively, according to AMTI's definition) during

a five sec. trial with sampling at 100 Hz. For example if the there are five

consecutive samples of 0.5, 0.10, 0.5, -0.5, -0.10 cm, for the X values he

adds 0.5+0.5+0.10+0.5=2.5 cm (differences between the values). This is also

calculated for the Y direction. The velocity can be determined by dividing

by the amount of time to get cm/sec.

The calculations for the difference between the 500 numbers can easily and

quickly be derived using a spreadsheet like Microsoft Excel. If the X data

is in column A (cells A1:A500) of the worksheet, enter the formula in cell

B2: =ABS(A2-A1) and fill the formula down to B500. In cell B501, calculate

the sum by entering the formula: =SUM(B2:B500) or by clicking on the sum S

button on the toolbar.

This is like an integration of a full-retified EMG signal. A

root-mean-square (RMS) calculation would also work, I believe. If your Peak

system is recording the force plate data, it should be able to export the

files to a spreadsheet. We use an Ariel system which exports data into the

Excel spreadsheet.

I have not verified this method with other people studying postural sway and

would be interested in knowing if this is a valid method. I would also like

to see other responses you get. If you don't post replies, please forward

them to me.

Thank you.

BE

Bruce Etnyre, Ph.D., P.T. _/_/_/_/ _/_/_/_/ _/_/_/ _/_/_/_/ _/ _/

Human Performance and _/ _/ _/ _/ _/ _/ _/ _/

Health Sciences Dept. _/ _/ _/ _/ _/ _/ _/

Rice University _/_/_/_/_/ _/ _/ _/_/_/ _/ _/

6100 Main MS 545 _/ _/ _/ _/ _/ _/ _/

Houston, Texas 77005 _/ _/ _/ _/ _/ _/ _/

USA _/ _/ _/ _/ _/ _/ _/ _/

etnyre@rice.edu _/ _/ _/_/_/_/ _/_/_/ _/_/_/_/ _/_/ _/

(713)527-4058

FAX: (713)285-5329

************************************************** ***********************

We resolved the problem as follow supposing you have about 1000 points

passed by sway. All points have a pair of coordinates (x , y) :

1. taking the average of x and y gives X0 and Y0

2. calculate the distance D and the angle A of each point vs X0 and Y0

3. divide the area in 24 sectors with X0 and Y0 as center

4. sort the data so that you find which point is in wich sector

5. sort the data in each sector to the biggest D

6. calculate the surface of each triangle obtained and make the sum.

The more sectors gives more chance to be accurate but also increase

the chance of finding a area without points...

I found the algorithm in :

Arch. Phys. Med. Rehabil vol 75 September 1994

Sincerely

Pierre

================================================== ========

ing Pierre Van Cleven

Dept of Kinesiology Institute of Physical Education

University of Gent

Watersportlaan 2 B-9000 Gent (Belgium)

tel (32)9.264.63.21 fax (32)9.264.64.97

================================================== ======

************************************************** *****************

>I'm having some troubles to find a way to quantify the surface of the sway.

>The results shown a multi-form pattern.

You might try looking at the data in terms of fractional brownian motion,

as in these papers:

Riley, M. A., Mitra, S., Stoffregen, T. A., & Turvey, M. T. (1997).

Influences of body lean and vision on unperturbed postural sway. Motor

Control, 1, 229-246.

Collins, J.J. & De Luca, C.J. (1995). The effect of visual input on

open-loop and closed-loop postural control mechanisms. Experimental Brain

Research, 103, 151-163.

About a dozen additional ways to characterize postural sway are presented in

Riccio, G. E., & Stoffregen, T. A. (1991). An ecological theory of motion

sickness and postural instability. Ecological Psychology, 3, 195-240.

Good luck.

Tom Stoffregen

------------------------------------------------------------------------------

Thomas A. Stoffregen, Ph.D.

Department of Psychology, P.O. Box 210376

University of Cincinnati

Cincinnati, OH 45221-0376

(513) 556-5569

Personal homepage: http://ucaswww.mcm.uc.edu/psychology/Faculty/Stoffreg.htm

Postural stability laboratory: http://ucaswww.mcm.uc.edu/psychology/stoffregen/

------------------------------------------------------------------------------

************************************************** *************************

Mylene:

We are just developing a program for testing postural sway

in knee patients. So far, we have the subjects stand on the

force plate (Kistler) in a single leg stance for 15 seconds

with eyes open and eyes closed. We record the vertical ground

reaction force Fz and the location of the force vector on the

platform in both directions (ax and ay). We sample 600

measurements, i.e. we have a frequency of 40 Hz. We calculate the

standard deviation of ax and ay, the range between min and max

values in both directions and the total distance that the force

vector "traveled" during the whole period. Dividing distance by

time gives you something like the sway velocity.

We have yet to find out which parameter(s) is (are) most reliable

and can help to distinguish between normal and pathologic

conditions. I hope this helps a little.

Ciao,

Dieter...

__________________/\ ________/\ _____________________

\/ \/

Dr. Dieter Rosenbaum

Abt. fuer Orthopaedische Physiologie (Kinesiology Lab)

Klinik und Poliklinik fuer Allgemeine Orthopaedie

Westfaelische Wilhelms-Universitaet Muenster

Albert-Schweitzer-Str. 33

D-48129 Muenster, Germany

Phone: (...49) 251 - 834 7981 or 7995 (fax 7989)

email: diro@uni-muenster.de

************************************************** *************************

Thank you to Stephen J. Kinzey, Ph.D. for his response.

Stephen J. Kinzey, Ph.D.

Assistant Professor / Director of Biomechanics Laboratory

The University of Mississippi

Department of ESLM

University, MS 38677

e-mail: skinzey@olemiss.edu

http://www.olemiss.edu/~skinzey/biomch.htm

office: (601) 232 - 5540

fax: (601) 232 - 5525