View Full Version : Re: Thanks re Angular Velocity Data

12-28-2006, 01:30 PM
Dear Joanna/others interested,

Indeed, Santa in Heidelberg did send one set of normative angular velocity
data (see below).

I've added it to the CGA database: http://www.univie.ac.at/cga/data


----- Forwarded Message ----

Dear Chris,

We have some normative data for temporal gradients as an approximation for
joint velocities computed from a data set of 22 children (age 8.3
+-1.8years) walking at self-selected pace (see attachment). The data
set was
recorded in the Gait lab of the Orthopädische Universitätsklinik Heidelberg
in Germany by Sebastian Wolf and Leonhard Doederlein.

The methodology is described in Eq. (1) from

Wolf, S.; Loose, T.; Schablowski, M.; Döderlein, L.; Rupp, R.; Gerner, H.
J.; Bretthauer, G.; Mikut, R.:

Automated feature assessment in instrumented gait
Gait & Posture, 23 (3), pp. 331-338; 2006

The resulting units for the temporal gradients are degree per percent gait

The data are processed using the open source MATLAB toolbox Gait-CAD (see

At the moment, this toolbox is only available in German language – but an
English version is under development.

If you want, you can provide this mail via the CGA list.

Best regards



Dr.-Ing. Ralf Mikut

Forschungszentrum Karlsruhe GmbH

Institut fuer Angewandte Informatik (IAI)

Postadresse: 76021 Karlsruhe, Postfach 3640

Lieferadresse: 76344 Eggenstein-Leopoldshafen, Hermann-von-Helmholtz-Platz 1

Tel.: ++49-(0)7247-82-5731

Fax: ++49-(0)7247-82-5702

E-Mail: ralf.mikut@iai.fzk.de



GMA-FA 5.14 Computational Intelligence: <


Von: cga@info.curtin.edu.au [mailto:cga@info.curtin.edu.au] Im Auftrag von
Chris Kirtley

Gesendet: Mittwoch, 20. Dezember 2006 08:27

An: Multiple recipients of list CGA

Betreff: Re: angular velocity data

On 12/27/06, Johanna Robertson wrote:
> Thankyou to everybody who replied to my search for normative angular
> velocity data. It appears that no-body has any such data!! Chris Kirtley
> suggests why:
> No - I have never seen any normative data on joint velocity (and I have
> looked many times!). There are some methodological problems which I
> suspect
> may be the reason why. To get velocity you need to run a derivative
> algorithm - usually a difference equation followed by a filter, because
> differentiation inevitably amplifies any high-frequency noise in the
> signal.
> The end result will be very dependant on what cutoff frequency you use for
> the filter - a lower cutoff will reduce the velocity recorded, while a
> higher cutoff will let through high frequency noise. Choosing the cutoff
> is a bit subjective and empirical.
> Thanks again.
> Johanna Robertson
> Physiotherapist, MSc
> Movement Analysis Lab
> Hôpital Poincaré
> Garches, France.
> ---------------------------------------------------------------
> Information about BIOMCH-L: http://www.Biomch-L.org
> Archives: http://listserv.surfnet.nl/archives/Biomch-L.html
> ---------------------------------------------------------------

Dr. Chris Kirtley MB ChB, PhD

608 Dockside
44 Ferry St.
Kangaroo Point
Queensland 4169
Tel. (07) 3891 6644 x 1608
Fax 3891 6900

Clinical Gait Analysis: http://www.univie.ac.at/cga