View Full Version : Predicting Co-contraction with Optimization Methods

Morris Alan R
12-18-1995, 01:12 PM
Hello biomch-L readers:

I am reviewing optimization methods that have been developed to predict
muscle forces from either movement data or goal directed forward dynamic
simulation. In reviewing the static optimization approach (inverse
dynamic method with iterative solution for force distribution based on an
optimizing criteria, ie. min. muscle stress, min. muscle force, etc.) and
dynamic optimization approach (forward dynamic simulation, with optimal
control based on maximizing a performance goal, ie. min. time, maximum
jumping height).

In reviewing the methods, both seem to be unable to accurately predict
muscle phasic activity of antagonistic muscles durng gait, primarily the
quadriceps-hamstrings action at the knee and hip around heelstrike. Is
there any optimization criteria that has been tried out there (yet not
published) that can predict this activity. Or will it be impossible to
find an optimizing criteria, since co-contraction is not an optimal
response but rather a safety measure to increase joint stiffness at times
of peak loading. These methods at this juncture, would be inadequate to
predict muscle forces in pathalogical cases, where the presence of
co-contraction is alot greater throughout.

The most recent and informative review of the static optimization method
has been:
Collins, J.J. (1995) The redundant nature of locomotor optimization
laws. J. Biomechanics 28, 3, pp.251-267.

I will post responses.

Alan R. Morris, M.A.Sc, P.Eng.
Research Engineer - Gait Laboratory
The Hugh MacMillan Rehabilitation Centre
350 Rumsey Road
Toronto, Ont.
M4G 1R8

(416) 425-6220 x508