View Full Version : Subject: The future of biomechanics

Ilse Jonkers
01-10-2002, 09:33 PM
Dear all,

Following the remarks made by Paul Ostic, on the establishment
of a generic super-model of the locomotor system, I would like to
point out some of my concerns on this approach (nevertheless a
very attractive idea and a great challenge) . Personally, I would
welcome all initiatives streiving to integrate existing software tools
so that medical image parametrisation, personalised
musculoskeletal modelling, inverse and forward dynamics
(optimisation) and if possible FEA can be used within one platform.
In order to have a tool available which has the potential to answer
all potential research questions.

However, the complexity of the different parameter settings and the
inherent difficulty to control their effect on the possible outcome of
the analysis (especially the interaction effect)is worrying. We tried
to apply response surface modelling techniques to get an
estimation of this effet on the behavior of a 7segment model with 22
parameters describing the muscle activation profiles. Since most
mathematical routines were unable to perform this analysis, I find it
difficult to imagin that the individual researcher can keep track of
the interaction effects he introdudes.

I think that it is very important to carefully adopt the level of
complexity of the system to the specified working hypothesis,
taking into account the accuracy with which the proposed solution
can be obtained. Especially in the field of gait analysis, I am often
confronted with questions in which it is difficult to find this balance.
(How important is it to define the degree of osteotomies or muscle
lengthening procedures?). I think that the real art is the ability to
balance the results of analysis in the light of the inherent boundaris
imposed by the selected complexity of the system.

>From a clinical point, I am often confronted with questions related
to hypothesis testing and the need for simulation of surgical
treatment options (ie at the level of motor functioning of the patient
rather structural loading, What would happen if I would /would not
lengthen psoas, derotate). I am always surprised by the gap
between the expectations of the medical world and the present
state of the art especifically in forward simulation. I would like to
start the discusison on this aspect.

I see two aspects that could help in advancing the construction of
such a simulator:
1. Integration of Motor control issues on top of the musculoskeletal
model to help the optimisation processe establishing the activation
profiles (or muscle coordination pattern) to replicate the registered
movement of the patient. If applicable to the control of normal gait,
it would allow an interesting testbed for understanding altered
control mechanisms in patients with CND.
2. More realistic approaches towards modelling the mechanical
response of the muscle after activation. (How to define the
mechanical action line of an individual muscle?) Furthermore, more
advanced studies to describe muscle mechanics in pathology (eg.
modelling of spasticity/hypertonia, structural muscle shortening)
are required, as well as the definition of methodologies applicable
for individualised parametrisation of patient specific parameters.

I realize that by definition, this approach will further increase the
complexity of the system, so I end up contradicting myself. I
presume that the level of complexity of each these items will
depend on the individual research question.

I would welcome any reations on this topic.

Kind Regards


Ilse Jonkers
Research Assistant
Ergonomics Lab (Prof. Spaepen)-Faculty of Physical Education and Physiotherapy
Gait lab-University Hospital Pellenberg
Tervuursevest 101
3001 Leuven

Voice: ++ 32 16 329105/329100
Fax: ++ 32 16 329196

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