View Full Version : antagonism or optimisation

Ben Stansfield
05-20-2003, 11:49 PM
Dear All,
I am currently attempting to explore the distribution of muscle force in the
muscles of the lower limb. I have a 3D model mathematical model of the
lower limb with muscle locations, lever arms, etc. Fortunately I also have
a set of hip joint contact forces measured in vivo (Georg Bergmann's group)
to provide validation for my attempts. I am, however, finding it very
difficult to reproduce a suitable level of antagonism in the system to
ensure that my calculated results are the same as the measured values.

Peak values of hip joint contact forces seem to be in reasonable agreement
with those measured, but it appears that the strategy I am using does not
produce the correct form of antagonism during swing. The body appears to
over stiffen the joint during early swing and before foot contact.

I have tried the following muscle force distribution strategies:
-Minimisation of maximum stress in any one muscle
-Minimisation of sum of muscle forces
-Minimisation of sum of squared muscle forces
-Minimisation of sum of cubed muscle forces

In addition I have tried the following forced antagonism:
-Forced antagonism at varying levels (ensures that the muscles over
constrain a joint by xNm)
-Look forward to level of joint moment 0.x of a second in front and attempt
to move smoothly towards that value by imposing antagonism at a level
dependent on how long to the peak value and how much bigger than the current
value of moment it is.

I have also constrained the muscle forces to have a maximum change of stress
with time.

I would like to hear of any suggestions people have for distribution of
force to the muscles. I believe I have read about most of the options, but
would welcome all views.

I will summarise responses in due course.


Ben Stansfield

PhD, CEng, MIMechE
Bioengineering Unit
University of Strathclyde
Wolfson Centre
106 Rottenrow
G4 0NW
Tel: ++44 (0)141 548 3031
Fax: ++44 (0)141 552 6098

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