Tweet
Dear Colleagues,
Recently, Dr. Viceconti called for the submission of topics that might
be regarded controversial issues in musculoskeletal biomechanics. I did
respond to his request with some suggestions, and hope that others will
do so as well.
Quite independent from his move, I initiated discussions on UNSOLVED
FUNDAMENTAL PROBLEMS IN BIOMECHANICS already this spring and summer in
numerous seminars, lectures, and conference presentations at Stanford
University, the University of Texas at Austin, in San Francisco,
Cologne, etc. These discussions turned out to be exceptionally inspiring
and involved many internationally renowned colleagues, peers, and
graduate students. My impression was that these talks elicited a
positive echo and great interest in all those who participated.
I initiated these discussion sessions because I feel the time is ripe
for the development of new and innovative approaches to solve intricate
biomechanical problems within a framework that I would call "SECOND
GENERATION BIOMECHANICS". In principle, the current situation shows some
similarity to that prevailing in the late 1950s and during the 1960s. At
that time, researchers gathered tremendous volumes of data because the
technology was available. However, seldom were these experimental data
related to model predictions because there existed practically no
sufficiently complex biomechanical models, especially not of the human
neuromusculoskeletal system. This situation changed when in the period
after 1969 a few colleagues and I began to introduce computational
biomechanics, notably large-scale biomechanical model building and the
development of methods for the determination of subject-specific
biomechanical parameter values. I am grateful for having had the
opportunity and for having received continuous encouragement to do this
pioneering work, which was not always easy.
Today, senseless collection of data is, of course, no longer an issue
and modeling of biomechanical systems has become a popular exercise. The
implementation of such models and their simulation on the computer is as
easy as it never was before, thanks to tremendously increased (and
continuously increasing) computing power and the general availability of
computing facilities. However, I often get the impression that the
number of biomechanical models created (and published) is inversely
related to their quality and to the biomechanico-philosophical content
behind the model building efforts. By this I mean that often quantity is
traded for quality and fundamental issues are largely ignored,
frequently under the pressure of project funding and having to produce
results quickly.
It would therefore seem appropriate to discuss these issues on a
world-wide scale. BIOMCH-L is certainly the most appropriate forum for
this. As a basis for such discussions, I would like to introduce a
selection of problems which, in my opinion, are some of the most
pressing ones, calling for novel approaches and concerted research
efforts to be solved.
1. THE FUNDAMENTAL PROBLEM OF MYOSKELETAL INVERSE DYNAMICS
The problem arises from the discrepancy between the responses of
inadequate and frequently oversimplified models of the human
musculoskeletal system combined with inaccurate collection and
processing of kinematic input data, and fairly accurately measurable
dynamic observables of the real biosystem, such as ground reaction
forces. The problem and its negative implications are described in
detail in a forthcoming publication (January-2002-issue of the Journal
of Biomechanics Nr. 35/1, pp. 109-115).
Closely related to this problem is that of computing first and second
time derivatives from noise-contaminated data sequences which problem,
by definition, belongs to the class of incorrectly posed (or ill-posed)
problems.
A solution of the myoskeletal inverse dynamics problem with all its
facets would be of great practical value for all forms of motion
analysis in orthopaedics, ergonomics, sports, etc.
2. THE MYOSKELETAL INDETERMINACY PROBLEM
There are, in general, more muscles generating torques across joints
than would be necessary to uniquely define all torque components
corresponding to the angular degrees of freedom of a joint. Because the
existence of redundant muscles is highly unlikely from teleological
considerations one has to conclude that some task-dependent optimality
principles are operative in selecting the moment sharing distribution of
a specific muscle group. Current approaches to this problem are
generally inadequate. However, a solution would be most desirable
because it is closely linked to the myoskeletal inverse dynamics problem
as applied in clinical practice.
3. IDENTIFICATION OF NEUROMYOSKELETAL PERFORMANCE CRITERIA
In order to solve problem 2 discussed above, a clear understanding of
the task-relevant performance criterion (also called performance index,
objective function, cost function, etc.) that determines the neural
control inputs to the muscles must exist. Research in this important
area is still in its infancy despite its emminent practical importance.
4. DEVELOPMENT OF ADEQUATE SKELETAL MUSCLE ANALOGS
Latest experimental results obtained in biochemistry and muscle
physiology indicate that our current models of skeletal muscle must
undergo thourough revision with respect to both the properties of
contractile elements and those of the passive elements. A strict
distinction between these components does no longer seem appropriate. It
may be necessary to develop new concepts such as three-dimensional force
fields to account for the phenomena of intra-, inter-, and extramuscular
lateral force transmission via cytoarchitectural and myofascial
structures. In addition, the complete motor unit substructure as well as
the peculiarities of the stochastic control behavior of skeletal muscle
will have to be incorporated, if muscle models are to mimic biological
reality. The practical importance of adequate muscle models for clinical
applications is obvious.
5. DEVELOPMENT OF RIGIDO-ELASTIC HYBRID ANALOGS OF THE HUMAN LIMB SYSTEM
Currently used segmented human body models almost exclusively use rigid
body segments. This does not conform to reality and is probably a source
of large errors in inverse dynamical investigations, at least in motions
containing acceleration transients such as impacts. The practical value
of appropriate body models containing elastic structures modeled as
continua and not as wobbling mass-spring-
damper combinations is obvious.
6. MICROSTRUCTURAL REPRESENTATION OF TERMINAL SEGMENTS
Terminal segments of the human body such as hands and feet exhibit a
predominantly bony structure of extreme complexity. This complexity is
also present as far as the representation of ligamentous, tendinous,
myodynamic, and soft-tissue structures is concerned. Because the
terminal segments are of cardinal importance as manipulators, the
detailed modeling of their microstructure should be considered a task of
high priority. (To my knowledge, no really satisfactory FE-model exists
of terminal segments that really incorporates ALL functional facets of
these intricate structures).
7. DEVELOPMENT OF NEW AND INNOVATIVE METHODS FOR SUBJECT-SPECIFIC
PARAMETER IDENTIFICATION
For any model of the segmental, muscular, articular, or neural subsystem
to be implemented in practice, the values of the respective
subject-specific parameter sets must be available. Comparatively little
effort has been devoted to this extremly important field of
biomechanical research despite its obvious practical relevance.
I hope to have provided some impetus for fruitful discussions of, in my
opinion, some of the most pressing issues of biomechanical research.
Other colleagues may have different views. Their comments would be most
welcome in this forum and also with respect to the upcoming European
BIONET EVENT. This would help all of us to identify on a global scale
those problems of biomechanical basic research which should be tackled
in the near future with high priority, in a decade of research on the
human neuromusculoskeletal system.
I would like to thank all of you who invested time and effort by
entering this discussion forum and would also like to take this
opportunity to wish you a happy Christmas and prosperous Year 2002.
Herbert Hatze
************************************************** ******
Prof. Dr. Herbert Hatze
Head, Department and Laboratory of Biomechanics, ISW,
University of Vienna
Auf der Schmelz 6 Tel: + 43 1 4277 48880
A-1150 WIEN Fax: + 43 1 4277 48889
AUSTRIA e-mail: herbert.hatze@univie.ac.at
************************************************** ******
---------------------------------------------------------------
To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://isb.ri.ccf.org/biomch-l
---------------------------------------------------------------
Recently, Dr. Viceconti called for the submission of topics that might
be regarded controversial issues in musculoskeletal biomechanics. I did
respond to his request with some suggestions, and hope that others will
do so as well.
Quite independent from his move, I initiated discussions on UNSOLVED
FUNDAMENTAL PROBLEMS IN BIOMECHANICS already this spring and summer in
numerous seminars, lectures, and conference presentations at Stanford
University, the University of Texas at Austin, in San Francisco,
Cologne, etc. These discussions turned out to be exceptionally inspiring
and involved many internationally renowned colleagues, peers, and
graduate students. My impression was that these talks elicited a
positive echo and great interest in all those who participated.
I initiated these discussion sessions because I feel the time is ripe
for the development of new and innovative approaches to solve intricate
biomechanical problems within a framework that I would call "SECOND
GENERATION BIOMECHANICS". In principle, the current situation shows some
similarity to that prevailing in the late 1950s and during the 1960s. At
that time, researchers gathered tremendous volumes of data because the
technology was available. However, seldom were these experimental data
related to model predictions because there existed practically no
sufficiently complex biomechanical models, especially not of the human
neuromusculoskeletal system. This situation changed when in the period
after 1969 a few colleagues and I began to introduce computational
biomechanics, notably large-scale biomechanical model building and the
development of methods for the determination of subject-specific
biomechanical parameter values. I am grateful for having had the
opportunity and for having received continuous encouragement to do this
pioneering work, which was not always easy.
Today, senseless collection of data is, of course, no longer an issue
and modeling of biomechanical systems has become a popular exercise. The
implementation of such models and their simulation on the computer is as
easy as it never was before, thanks to tremendously increased (and
continuously increasing) computing power and the general availability of
computing facilities. However, I often get the impression that the
number of biomechanical models created (and published) is inversely
related to their quality and to the biomechanico-philosophical content
behind the model building efforts. By this I mean that often quantity is
traded for quality and fundamental issues are largely ignored,
frequently under the pressure of project funding and having to produce
results quickly.
It would therefore seem appropriate to discuss these issues on a
world-wide scale. BIOMCH-L is certainly the most appropriate forum for
this. As a basis for such discussions, I would like to introduce a
selection of problems which, in my opinion, are some of the most
pressing ones, calling for novel approaches and concerted research
efforts to be solved.
1. THE FUNDAMENTAL PROBLEM OF MYOSKELETAL INVERSE DYNAMICS
The problem arises from the discrepancy between the responses of
inadequate and frequently oversimplified models of the human
musculoskeletal system combined with inaccurate collection and
processing of kinematic input data, and fairly accurately measurable
dynamic observables of the real biosystem, such as ground reaction
forces. The problem and its negative implications are described in
detail in a forthcoming publication (January-2002-issue of the Journal
of Biomechanics Nr. 35/1, pp. 109-115).
Closely related to this problem is that of computing first and second
time derivatives from noise-contaminated data sequences which problem,
by definition, belongs to the class of incorrectly posed (or ill-posed)
problems.
A solution of the myoskeletal inverse dynamics problem with all its
facets would be of great practical value for all forms of motion
analysis in orthopaedics, ergonomics, sports, etc.
2. THE MYOSKELETAL INDETERMINACY PROBLEM
There are, in general, more muscles generating torques across joints
than would be necessary to uniquely define all torque components
corresponding to the angular degrees of freedom of a joint. Because the
existence of redundant muscles is highly unlikely from teleological
considerations one has to conclude that some task-dependent optimality
principles are operative in selecting the moment sharing distribution of
a specific muscle group. Current approaches to this problem are
generally inadequate. However, a solution would be most desirable
because it is closely linked to the myoskeletal inverse dynamics problem
as applied in clinical practice.
3. IDENTIFICATION OF NEUROMYOSKELETAL PERFORMANCE CRITERIA
In order to solve problem 2 discussed above, a clear understanding of
the task-relevant performance criterion (also called performance index,
objective function, cost function, etc.) that determines the neural
control inputs to the muscles must exist. Research in this important
area is still in its infancy despite its emminent practical importance.
4. DEVELOPMENT OF ADEQUATE SKELETAL MUSCLE ANALOGS
Latest experimental results obtained in biochemistry and muscle
physiology indicate that our current models of skeletal muscle must
undergo thourough revision with respect to both the properties of
contractile elements and those of the passive elements. A strict
distinction between these components does no longer seem appropriate. It
may be necessary to develop new concepts such as three-dimensional force
fields to account for the phenomena of intra-, inter-, and extramuscular
lateral force transmission via cytoarchitectural and myofascial
structures. In addition, the complete motor unit substructure as well as
the peculiarities of the stochastic control behavior of skeletal muscle
will have to be incorporated, if muscle models are to mimic biological
reality. The practical importance of adequate muscle models for clinical
applications is obvious.
5. DEVELOPMENT OF RIGIDO-ELASTIC HYBRID ANALOGS OF THE HUMAN LIMB SYSTEM
Currently used segmented human body models almost exclusively use rigid
body segments. This does not conform to reality and is probably a source
of large errors in inverse dynamical investigations, at least in motions
containing acceleration transients such as impacts. The practical value
of appropriate body models containing elastic structures modeled as
continua and not as wobbling mass-spring-
damper combinations is obvious.
6. MICROSTRUCTURAL REPRESENTATION OF TERMINAL SEGMENTS
Terminal segments of the human body such as hands and feet exhibit a
predominantly bony structure of extreme complexity. This complexity is
also present as far as the representation of ligamentous, tendinous,
myodynamic, and soft-tissue structures is concerned. Because the
terminal segments are of cardinal importance as manipulators, the
detailed modeling of their microstructure should be considered a task of
high priority. (To my knowledge, no really satisfactory FE-model exists
of terminal segments that really incorporates ALL functional facets of
these intricate structures).
7. DEVELOPMENT OF NEW AND INNOVATIVE METHODS FOR SUBJECT-SPECIFIC
PARAMETER IDENTIFICATION
For any model of the segmental, muscular, articular, or neural subsystem
to be implemented in practice, the values of the respective
subject-specific parameter sets must be available. Comparatively little
effort has been devoted to this extremly important field of
biomechanical research despite its obvious practical relevance.
I hope to have provided some impetus for fruitful discussions of, in my
opinion, some of the most pressing issues of biomechanical research.
Other colleagues may have different views. Their comments would be most
welcome in this forum and also with respect to the upcoming European
BIONET EVENT. This would help all of us to identify on a global scale
those problems of biomechanical basic research which should be tackled
in the near future with high priority, in a decade of research on the
human neuromusculoskeletal system.
I would like to thank all of you who invested time and effort by
entering this discussion forum and would also like to take this
opportunity to wish you a happy Christmas and prosperous Year 2002.
Herbert Hatze
************************************************** ******
Prof. Dr. Herbert Hatze
Head, Department and Laboratory of Biomechanics, ISW,
University of Vienna
Auf der Schmelz 6 Tel: + 43 1 4277 48880
A-1150 WIEN Fax: + 43 1 4277 48889
AUSTRIA e-mail: herbert.hatze@univie.ac.at
************************************************** ******
---------------------------------------------------------------
To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://isb.ri.ccf.org/biomch-l
---------------------------------------------------------------