View Full Version : A discussion on JCS

08-03-1998, 03:06 AM
Hello all,

On July 28, 1998, I received an e-mail message from Mr. Juan Vicente Dura
Gil, Universidad Politécnica de Valencia, who posed a question about why
the ISB Standardization Committee is recommending the use of JCS based on
Grood and Suntay's work rather than Woltring's work. In the message, Mr.
Juan Vicente Dura Gil discussed the disadvantages of Grood & Suntay's
approach and the advantages of Woltring's approach. Since then, we have
had some discussions on this topic.

In relation to our discussion, I was forwarded a message from Dr. van den
Bogert (thanks to Dr. Brian Davis), who reminded us the "excellent
Woltring-Grood polemic on Biomch-L, which I summarized in

The purpose of this message is to inform you about the concern of Mr. Juan
Vicente Dura Gil (I included his original e-mail below), and to invite you
to visit the above web page if interested. If you would like to further
contribute to the discussion, I believe Mr. Juan Vicente Dura Gil, the ISB
Standardization Committee, and Biomech-L'ers would like to hear from you.


Ge Wu

Mr. Juan Vicente Dura Gil's original message:

>Date: Tue, 28 Jul 1998 18:26:21 +0200
>From: Juan Vicente Dura Gil
>Organization: Universidad Politécnica de Valencia
>X-Mailer: Mozilla 4.03 [es] (Win95; I)
>To: Ge Wu
>Subject: Re: Invitation to test use the joint coordinate systems
>Dear Mr. Wu,
>I would like to express some ideas and to know the opinion of people
involved in the
>standardisation committee. Perhaps these ideas have been discussed before.
>excuse me if my english is not good.
>I believe that it is necessary to separate the problem in two levels.
>Level 1 is the selection of the axis in the segment (local reference
system). The
>selection of the axis is an anatomical problem. Normally this axis are
>looking principal inertia moments, trying to adapt the body segments to
>solids (cylinders, pyramids,..). The relative attitude matrix express the
>position between two segments, and this matrix has one unique solution.
>Level 2 is expressing the attitude matrix in three angles. The problem is
to find
>three angles with a reasonable anatomical meaning. It seems that most of
>prefer Euler/Cardan angles. Euler/Cardan angles expresses the attitude of
body segment
>like three consecutive rotations around the axes of local reference system
and it is
>possible to select six different orders of rotations. The attitude vector
>the attitude of the body segment like only one rotation around finite
helical axis,
>and the three angles are the three components of the vector.
>My opinion is that attitude vector has mathematical advantages:
>a) It is less sensible to errors than Euler/Cardan,
>b) It has less problems of continuity (gimbal-lock effect do not exist),
>c) the absolute value of the angles are the same describing the position
of distal
>segment referred to
> proximal or proximal referred to distal.
>My short experience is that for the hip, knee and ankle attitude vector
has the
>similar anatomical interpretation as the better Euler rotation order. But,
>vector is not considered in the draft proposals. ¿Why do you prefer
>angles? ¿Have you considered to test attitude vector?
>I have read the Woltring discussion with other researches, and I do not
know what are
>the reasons for preferring Euler/Cardan angles. Perhaps this point has
been discussed
>before, but if attitude vector has mathematical advantages, the only
reason for
>preferring Euler/Cardan must be anatomical. ¿There is any work that probe
>attitude vector has less anatomical meaning in any joint?
>Attachment Converted: "c:\internet\eudora\attach\vcard4.vcf"

To unsubscribe send UNSUBSCRIBE BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://www.bme.ccf.org/isb/biomch-l