View Full Version : Re: Surface markers in 3D analysis

Paolo De Leva - Sport Biomechanics, Rome,italy
12-09-1995, 03:08 AM
> Using joint center locations is just one method of calculating in vivo
> human body kinematics. Another method is to use 3 (or more) markers
> attached to each body segment, and assuming each segment is a rigid body,
> the 6 DOF between each segment can be calculated. If the subject is
> filmed in a known position, in which all joint angles are
> defined, all joint angles can be calculated with respect to their defined
> reference positions. This method allows for the use of surface markers
> which can be positioned without reference to any bony landmarks or joint
> centers, and often these marker positions can be adjusted to allow for
> better viewing by multiple cameras.
> Chris Sherwood
> Department of Bioengineering
> University of Pittsburgh

you might be interested to know that a similar method was described by

Cappozzo et al. (1995) Position and orientation in space of bones during
movement: anatomical frame definition and
determination. Clinical Biomechanics,
Vol. 10, 4, 171-178.

Contrary to Cappozzo, Chris suggests using landmarks positioned
without reference to any bony landmarks or joint centers. I would want to
point out that this method is useful when you are inerested only in joint angles,
angular positions and velocities of segments, and a few other kinematic
parameters. However, it makes it difficult to locate the segment CMs,
and to estimate the segment tensors of inertia, which are often needed,
particularly when you want to study the motion of either the whole human body,
or a chain of linked segments.

In fact, you cannot use mean segment inertia parameters from the
literature, that are always referenced to bony landmarks or joint centers,
unless you measure the 3D position of these reference points relative to
your surface markers (as suggested by Cappozzo et al.).

Of course, you can also apply a personalized inertia model to estimate
the individual parameters for each of your subjects. In that case, you might
use the positions of the surface markers placed on the subject as reference
points to locate the segment CMs. This means that you have to use the same
markers both for building the inertia model and for performing the kinematic
analysis. The practical application of this second method either is very complex
or requires expensive automatic systems for estimating the individual inertia

I would like to add that I am not sure the use of surface landmarks
increases measurement accuracy. In fact, it is even possible that skin artifacts
may prove in some cases to be a larger source of error than the use of
invisible (and purely theoretical) joint centers as reference points.

Of course, the use of joint centers is not indicate when you need to
study a 6DOF motion. For that, you need at least three points on each
assumed-rigid segment. Also, you have to use surface landmarks when you want
to use automatic data acquisition, as Tochi wrote.

It is important, however, to be aware that many situations exist
which make impossible the use of surface landmarks, and that in those
situations the manual and somewhat subjective estimate of joint center
positions is broadly utilized. Although joint centers are invisible
reference points, and although they are assumed to have a fixed position
relative to both the segments forming the respective joints (which is not
true because human joints do not behave as simple mechanical joints),
they are widely used as reference points for kinematic analyses.

I believe that many people will keep using joint centers in the
future. Although the use of joint centers involves measurement errors,
no method is errorless.

Surface landmarks may introduce other errors and problems, as other
subscribers have pointed out. Skin artifacts and interpolation for
estimating hidden landmark positions are also important sources of
measurement error.

I have been told that the use of fixed joint centers as reference
points should be rejected, because it gives unaccurate results. Also,
somebody maintains that joint centers belong to the past. This
is true only for some fields of study, tipically performed inside a
lab setting, for which more accurate techniques have been developed.
Observing the thorough landskape, we cannot fail to notice that joint
centers are still used for high level research all over the world, as
Jesus Dapena pointed out, and it is difficult to imagine that in the
near future technical progress will offer us a way to do without them.

__________ _________ ___________~___ ________ _________________~___
/ ~ ~ ~ ~ \
/______________~______~__________ _______~_____~______________~_____~_____\
| Paolo de Leva ~ \ Tel.+ FAX: (39-6) 575.40.81 |
| Istituto Superiore di Educazione Fisica > other FAX: (39-6) 361.30.65 |
| Biomechanics Lab / |
| Via di Villa Pepoli, 4 < INTERNET e-mail address: |
| 00153 ROME - ITALY \ deLEVA@RISCcics.Ing.UniRoma1.IT |
|_____________________~________~__________________ __________________ _____|
challenging entropy :-)