View Full Version : 3-D upper extremity kinematics

Guy Simoneau
10-26-1993, 03:11 AM
Received: from bugs.bme.ri.ccf.org by vms.csd.mu.edu (PMDF V4.2-14 #3524) id
; Mon, 25 Oct 1993 14:11:59 CDT
Received: from [] (speedy.bme.ri.ccf.org) by bugs.bme.ri.ccf.org
(4.1/5.921208) id AA05442; Mon, 25 Oct 93 15:07:31 EDT
Date: Mon, 25 Oct 1993 15:05:09 -0400 (EDT)
From: Brian Davis
Subject: RE: Upper Extremity Info
To: SIMONEAUG@vms.csd.mu.edu
X-Envelope-to: SIMONEAUG
Content-transfer-encoding: 7BIT
X-Popmail-Charset: English

A sincere thank you to the individuals that answered my request for
info on 3-D kinematic models for the upper extremity. As promised,
here are the answers I received:

We have been tracking the trunk and upper arm using a SELSPOT/TRACk system
for quite some time. To track the trunk we mount an array of markers on
a pilon coming off of the back. The base of the pilon sits on a plate
which is formed of plastic. It sits between the shoulder blades over
the thorasic spine. It is held in place by bands that rap around the
rib cage and expand with breathing. There are also a set of suspenders
which cross over near the neck so that they are not much affected by
shoulder motion. This is certainly not perfect but I think it gives
a better representation of trunk (thorax) motion than markers riding
on the shoulders which are affected as much by arm motion as trunk

We do not track the shoulder blades.

We use an arrays of markers on the upper arm to track the arm. We
do not realy track the forearm or hand. If you want to track the
different arm segments, I would use markers on the shoulder and
a ring of markers that go completely around the arm at the elbow
and wrist.

Good Luck,
Patrick O. Riley

We have by now some experience in 3D kinematic data collection in manual
wheelchair propulsion and the use of these data for 3D modelling.
To start with, I will give you some references:

We previously used a fairly simple 3D model of the upper extremity to
calculate net torques around wrist, elbow and idealized shoulder. You can
find a publication on this in: Journal of Electromyography and Kinesiology
1(4); 270-280, 1991.
Dr Mary Rodgers (Dayton VA Medical Center) has also published something on
3-D modelling in wheelchair propulsion (see: NACOB-II Proceedings, 1992;
pp. 457-460.

Of course, the above models were fairly simple, in the sense that the
shoulder was modelled as a single ball-joint, whereas modelling of the
upper extremity should include the full shoulder mechanism. However, in
that case one immediately experiences difficulties in collecting kinematic
information on the scapula and clavicle. This can be solved by assuming a
relationship between isometric motions and movements, and measuring the
position of the scapula for a series of positions. We have tried to do this
for manual wheelchair propulsion (See Clinical Biomechanics, 8; 81-90,
1993). Runciman from Strathclyde has also done a study on the upper
extremity (push-ups), that uses the same methodology. I do not have his
adress here, or a publication, but I am certain you contact him through dr.
A.C. Nicol (Strathclyde?). Anyway, they will be on the bmech-l list.

Lately we did an experiment in which we collected data for dynamic
modelling (using VICON), but there are not yet any publications available
on that project.

For modelling of the upper extremity itself, and especially the shoulder,
you ought to have a look into the work by Pronk and by Van der Helm, for
instance Journal of Biomechanics 25(2); 129-144, 1992, but I am quite
certain that dr. Van der Helm will send you a reaction too.

To give you some tips on the actual measurements with VICON would be a bit
too much for now, but if you want some more information, just contact me at
the Mayo adress. If you want us to send you some publications on the
wheelchair studies, please contact dr Van der Woude in Amsterdam. Lastly,
for shoulder model publications, you can better contact dr Van der Helm
directly. The adresses are given below.


DirkJan Veeger,
Orthopaedic Biomechanics Lab,
Mayo Clinic,
Rochester MN 55905,
tel: (507)-284 4294

The Amsterdam adress is:
Dr. L.H.V. van der Woude,
Vrije Universiteit Amsterdam,
Faculty of Human Movement Sciences,
v.d. Boechorststraat 9,
1081 BT Amsterdam.
e-mail: L_H_V_van_der_Woude@FBW.VU.NL

A 3-D link segment model of the upper extremities and trunk was
developed by Nicol and Dvir at the University of Strathclyde in the

Nicol AC (1977) Elbow joint prosthesis design: biomechanical aspects.
PhD Thesis, University of Strathclyde, Glasgow.

Dvir Z (1978) Biomechanics of the shoulder joint. PhD Thesis,
University of Strathclyde, Glasgow.

I have used this model with the 4-camera VICON system at the
University of Strathclyde whilst looking at the biomechanics of standing
up and sitting in paraplegics assisted by functional electrical
stimulation and the 5-camera VICON system at the Dundee Limb Fitting
Centre looking at upper limb intersegmental loadings during various
manoeuvres performed by wheelchair users.

Marker locations are crucial - even with 5 cameras some marker loss is
inevitable. However, it can be minimised by careful positioning of the
cameras and the subject, and by using various marker combinations to
define technical reference frames which can then be used to determine
the positions of the points required by the model to define the
anatomical reference frames.

Another problem, not accounted for in Nicol and Dvir's model, is that of
pronation and supination of the forearm which is apparent in all but the
most simple movements of the upper limb.

If you require any further information on Nicol's, Dvir's or my own work
just let me know and I'll post on any diagrams etc that you may need.

Good luck,

Mike Dolan.

Michael J. Dolan, School of Biomedical Engineering,
University of Dundee, Dundee DD1 4HN, Scotland, U.K.
Tel: ++ 44 (0)382 344642 Fax: ++ 44 (0)382 344644
E-mail: m.j.dolan@dundee.ac.uk

We have some experience in using a VICON system for upper extremity
motions. In a joint research project between Dept. of Human Movement
Science, Vrije Universiteit Amsterdam, and Delft University of
Technology, we have recorded upper extremity motions during manual
wheelchair propulsion. Presently, I am working at Catholic
University of America, Washington D.C. In a project at the National
Rehabilitation Hospital we have recorded motions for normals and
bone cancer patients where parts of the scapula and humerus were
removed and replaced by an endoprosthis. At the NRH we are using
a software package developed by the National Institutes of Health,
called Move3D.
The scheme is pretty much the same for all experiments. The motions
are tracked by marker trees on the forearm, upperarm and trunk.
Typically four markers are used, as widely apart as possible.
Additional markers are placed on bony landmarks. From these bony
landmarks the local coordinate systems of the segments can be
calculated. In a seperate trial the relation between the bony
landmarks and the marker trees is recorded. Subsequently, the
markers on the bony landmarks can be removed.
For the upper extremity we are using the following bony landmarks:

two markers on the pelvic girdle, one markers on each acromion.

Vrije Universiteit Amsterdam:
two markers on the sternum (Incisura Jugularis, Processus
Xiphoideus) and two marker on the spine (C7 and T8). Advantage of
this scheme is that the local coordinate system of the thorax can be
defined using the direction of the sternum. However, the NIH scheme
will do as well.

Upper arm:
One marker at the acromion, two at the medial and lateral
epicondyle. The length axis of the upperarm is defined by the
midpoint of the latter two and the acromion. The other axis is
defined perpendicular to the plane of the three markers.

two markers at the medial and lateral epicondyle, two markers at the
proc. styloideus radiale and ulnare. This will define the length
axis and the other axis perpendicular to the plane fitted to these
four markers.

Always troublesome. At NRH we do not define it explicitly, and
assume that it is rigidly linked to the forearm. In Amsterdam the
second metacarpale is marked to get an estimate where the forces

The NIH-software automates the whole procedure, and incorporates
body segment parameters, which allows for net moment and net power
calculations, in addition to the kinematics. In Amsterdam we obtain
the kinematic variables and use them as input for a musculoskeletal
model of the shoulder mechanism.

With kind regards,

Frans C.T. van der Helm
Dept. of Mechanical Engineering
Catholic University of America
620 Michigan Ave., N.E.
Washington, D.C. 20064
tel. (202)-319-5170
fax. (202)-319-4499
e-mail: helm@pluto.ee.cua.edu

as of November 1, 1993:
Dept. of Mechanical Engineering
Delft University of Technology
Mekelweg 2
2628 CD Delft
The Netherlands
tel. (+31)-15-783812
fax. (+31)-15-784717
e-mail: wbmrasp@dutrex.tudelft.nl
We studied the 3-D motion of the arm while subjects performed
daily-living tasks (the trunk was held stationary, because the
results were later used for orthosis design). Markers were
placed at the shoulder, elbow, wrist, on an extension from
the wrist (to detect pro/supination) and on the MCP3 knuckle.
The joint angles were calculated directly, based on an assumed
model of the arm, rather than calculating general Euler angles
(which produce a less consistent result). To handle the fact
that the markers (we used spherical styrofoam balls) are on
the outside of the arm rather than at the joint centres, the
approximate distance from the marker centroid to the joint
centre was measured on the subject so that they could later
be translated to the joint centres.
Hope this helps some. Best of luck.
Carolyn Anglin
Dept. of Mechanical Engineering
University of British Columbia
email: Carolyn_Anglin@mtsg.ubc.ca