View Full Version : Myoskeletal Inverse Dynamics

Scott Tashman, Ph.d.
01-14-2002, 08:19 AM
*This message was transferred with a trial version of CommuniGate(tm) Pro*
I have been following with great interest the recent discussion initiated by
Prof. Hatze. The assessment of neuromuscular disorders (particularly in
children) is probably the dominant clinical application of motion analysis.
The vision Prof. Hatze presented earlier today, of using musculoskeletal
models to design individualized optimal treatment strategies for children
with neuromuscular disorders, has been the "holy grail" of clinical gait
analysis for many years. I remember discussing this vision in the 1980's
with some of the early advocates of automated clinical gait analysis for the
treatment of cerebral palsy in the U.S. (e.g. Jim Gage, David Sutherland).
We acknowledged the limitations of the models and technology available then,
but we were convinced that these problems would be solved over time. In
fact, most of the problems we KNEW about 20 years ago have been solved - the
data is much better (though arguably still far from perfect), the computers
are fast enough, the musculoskeletal models are far more comprehensive, and
better subject-specific anatomy/geometry is available from vastly improved
medical imaging. Some of the results are visible - motion analysis and
musculoskeletal modeling have clearly lead to better understanding and
improved outcome of treatment for children with cerebral palsy. And yet, it
seems that the goal of patient-specific treatment optimization and outcome
prediction for individuals with neuromuscular disorders remains a distant

Prof. Hatze rightly argues that further improvements in all of the areas
listed above are needed. He also addresses our limited understanding of the
performance criteria the nervous system uses to select a motor control
strategy for a specific task. This, in my opinion, is one of the areas
where we are significantly lacking. We are far from fully understanding the
criteria the intact neuromuscular control system utilizes for optimizing
muscle activation during essential motor tasks. Neuromuscular disorders
(such as spastic diplegia) alter essentially all aspects of the motor
control system (sensors, actuators, controllers); this may render many of
our "normal" modeling and control strategy assumptions of limited value.
New experimental and modeling paradigms may be needed to sufficiently expand
our knowledge in this area.

Even as we get closer to addressing these problems, we may run headfirst
into another. Our models for motion analysis have traditionally been
targeted at the anatomy, physiology and mechanics of the body and its
neuromusculoskeletal components. For example, in our modeling world we
typically assume that if we know all of the mechanical properties of a
tendon (geometry, viscoelastic properties, etc), we can predict how it will
behave in a given biomechanical environment. Many of us choose to ignore
the biological reality that no living tissue has static mechanical
properties. We make this choice even though we are generally aware that
bones, tendons and muscles all change their properties with changes in their
mechanical and/or biological environment - perhaps the alternative has been
too complex to consider (or at least to model). This may, however, be a
particularly limiting assumption in children, since a young, growing tissue
is more sensitive to changes in its environment than the corresponding
tissue in an adult. Say, for example, a preoperative gait analysis is
performed on a 7 year old child, who then undergoes a complex surgical
procedure. The goal is to optimize gait performance, as assessed during a
follow-up gait analysis one year later. How much do the geometry and
characteristics of the bones, muscles, tendons, nervous system, etc. of a
young child change in a year? [personal note - As a parent of such a child,
I would say a lot!] How might these changes have been affected/altered by
the procedures that were performed? How would the resulting changes affect
predictions of optimal movement? These are questions for which we currently
have no good answers. Thus, I would argue that a comprehensive model to
predict the optimal treatment for a child would need to somehow incorporate
prediction of growth and tissue remodeling. This would add a whole new
layer of complexity, especially for those who argue for a single,
comprehensive modeling framework capable of addressing a wide variety of

Furthermore, the problem is not limited to neuromuscular disorders of the
young. Biological adaptation is a known issue for natural or biologically
engineered tissue replacements. For example, there is strong evidence
suggesting that the autologous tendon grafts used to replace failed anterior
cruciate ligaments undergo significant long-term biological remodeling after
they are placed in the knee capsule, with resultant changes in mechanical
properties. And yet, models (mathematical/computer and cadaver) used to
study this procedure generally do not (or cannot) account for these changes.
The extent to which this simplification limits the value of these models is
difficult to assess with currently available data (though it would depend to
some extent on the goals/hypotheses of the modeler).

Perhaps when our models were crude and generic, biological adaptations were
not worth considering. But, as the other aspects of musculoskeletal
modeling continue to improve, I believe that the importance of modeling
biological responses and adaptations in musculoskeletal tissues will
continue to emerge. This is hardly a new direction for research - there is
a great deal of active and historical research in tissue growth and
remodeling. But, biological growth/adaptation has not traditionally been
investigated within the framework of in-vivo studies of human movement.
This presents both a tremendous challenge and a real opportunity for
intelligently designed research to investigate biological adaptations and
their impact on in-vivo human movement biomechanics.

Please pardon my rambling on, but I hope these thoughts add fuel to the
current discussion. I am very interested in others' opinions on these

Scott Tashman

__________________________________________________ ____
Scott Tashman, Ph.D.

Head, Motion Analysis Section Assistant Professor
Bone and Joint Center Department of Orthopaedics
Henry Ford Hospital School of Medicine
2799 W. Grand Blvd, ER2015 Case Western Reserve University
Detroit, MI 48202

Voice: (313) 916-8680
FAX: (313) 916-8812
E-Mail: tashman@bjc.hfh.edu
__________________________________________________ ____

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